PAGV2

PSS®E 33.
4
PROGRAM APPLICATION GUIDE
VOLUME 2
March, 2013
Siemens Industry, Inc.

Siemens Power Technologies International
400 State Street, PO Box 1058
Schenectady, NY 12301-1058 USA
+1 518-395-5000
www.siemens.com/power-technologies
© Copyright 1990-2013 Siemens Industry, Inc., Siemens Power Technologies International
Information in this manual and any software described herein is confidential and subject to change without notice and does not
represent a commitment on the part of Siemens Industry, Inc., Siemens Power Technologies International. The software
described in this manual is furnished under a license agreement or nondisclosure agreement and may be used or copied only
in accordance with the terms of the agreement. No part of this manual may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, for any purpose other than the purchaser’s personal use, without the
express written permission of Siemens Industry, Inc., Siemens Power Technologies International.
PSS®E high-performance transmission planning software is a registered trademark of Siemens Industry, Inc., Siemens Power
Technologies International in the United States and other countries.
The Windows® 2000 operating system, the Windows XP® operating system, the Windows Vista® operating system, the
Windows 7® operating system, the Visual C++® development system, Microsoft Office Excel® and Microsoft Visual Studio® are
registered trademarks of Microsoft Corporation in the United States and other countries.
Intel® Visual Fortran Compiler for Windows is a trademark of Intel Corporation in the United States and other countries.
The Python™ programming language is a trademark of the Python Software Foundation.
Other names may be trademarks of their respective owners.
Table of Contents
Dynamic Simulation Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1

12.1 Basic Dynamic Simulation .......................................................................................12-1
12.2 Simulation Model Library .........................................................................................12-7
12.3 Numerical Integration Stability .................................................................................12-8
12.4 Classification of Variables .....................................................................................12-11
12.4.1 General Classification ...............................................................................12-11
12.4.2 PSS®E Parameter and Variable Classification ........................................12-12
12.4.3 Data Space Allocation ..............................................................................12-14
12.5 PSS®E Dynamic Simulation Sequence .................................................................12-17
12.5.1 Overall Flow Chart ....................................................................................12-17
12.5.2 Subroutines TBLCNC and CONEC ..........................................................12-18
12.5.3 Subroutines TBLCNT and CONET ...........................................................12-18
12.5.4 Program Control .......................................................................................12-19
12.5.5 Simulation Output Channels .....................................................................12-23
12.6 Overview of Simulation Procedure ........................................................................12-24
12.6.1 Model Setup and Use ...............................................................................12-24
12.6.2 Simulation Model Setup ............................................................................12-24
12.6.3 Execution of Simulation Runs ...................................................................12-27
12.6.4 Preservation of Simulation Setup – Snapshots ........................................12-28
Dynamic Simulation Setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-1

13.1 Overview ..................................................................................................................13-1
13.2 Basic Setup Example ..............................................................................................13-1
13.2.1 Files ............................................................................................................13-1
13.2.2 Example System .........................................................................................13-2
13.2.3 Power Flow Model ......................................................................................13-2
13.2.4 Equipment Model Selection ........................................................................13-8
13.2.5 PSS®E Dynamics Data File .......................................................................13-8
13.2.6 Dynamics Data Entry and CONEC, CONET Construction .......................13-23
13.2.7 Output Channel Selection and CONET Construction ...............................13-23
13.2.8 Initial Condition Snapshot .........................................................................13-27
13.2.9 End of Skeleton PSS®E Execution ..........................................................13-28
13.2.10 Refinement of CONEC and CONET .........................................................13-28
13.2.11 Compilation and Linking of CONEC/CONET ............................................13-28
13.2.12 Start PSS®E .............................................................................................13-28
13.2.13 Retrieval of Network Data and Initial Conditions ......................................13-33
13.2.14 Entry of Additional Data ............................................................................13-34
13.2.15 Updated Snapshot ....................................................................................13-34
13.2.16 Initialization ...............................................................................................13-34
13.2.17 Initial Condition Error Checking ................................................................13-35
All material contained in this documentation is proprietary to Siemens Industry, Inc., Siemens Power Technologies International.
i
PSS®E 33.4
13.2.18 Initial Condition Snapshot .........................................................................13-36

13.2.19 Dynamic Simulation Documentation .........................................................13-36
13.3 Error Checking ......................................................................................................13-37
13.3.1 Gross Parameter Errors ...........................................................................13-37
13.3.2 Initialization Errors ....................................................................................13-39
13.4 Running Simulations .............................................................................................13-43
13.4.1 Basic Steps ..............................................................................................13-43
13.4.2 Simulation Procedure ...............................................................................13-55
13.5 Use of CONET in System Modeling ......................................................................13-69
13.5.1 Calling Models via CONET .......................................................................13-69
13.5.2 Example Using CONET Models ...............................................................13-69
Generator Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-1

14.1 Introduction ..............................................................................................................14-1
14.2 Standards on Equipment Modeling and Data Exchange .........................................14-1
14.3 Rules for Using PSS®E Equipment Models ............................................................14-2
14.3.1 Memory Reference .....................................................................................14-2
14.3.2 Model Calling Sequence .............................................................................14-2
14.4 Generator Models ....................................................................................................14-3
14.4.1 Source Impedance .....................................................................................14-3
14.4.2 Generator Rotor Modeling ..........................................................................14-4
14.4.3 Magnetic Saturation ..................................................................................14-11
14.4.4 Generator Rotor Speed Damping .............................................................14-12
14.4.5 Generator Model Details ...........................................................................14-13
14.5 Models CIMTR1 and CIMTR3 (Induction Generators) ..........................................14-28
Static var Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1

15.1 Principles .................................................................................................................15-1
15.2 Dynamic Characteristics ..........................................................................................15-3
15.2.1 Representation ...........................................................................................15-3
15.2.2 Key Time Constant .....................................................................................15-5
15.3 Static var Device Models .........................................................................................15-6
15.3.1 System Interface .........................................................................................15-6
15.3.2 Handling SVGs in Power Flow Solutions ....................................................15-8
15.3.3 Model CSVGN1 ..........................................................................................15-9
15.3.4 Model CSVGN3 ..........................................................................................15-9
15.3.5 Model CSVGN4 ..........................................................................................15-9
15.3.6 Models CSVGN5 and CSVGN6 ...............................................................15-10
15.3.7 Model CSSCS1 ........................................................................................15-10
15.4 Example Case .......................................................................................................15-10
15.4.1 Example Case 2 .......................................................................................15-19
15.5 Control Tuning of an SVG .....................................................................................15-33
Excitation System and Controller Models . . . . . . . . . . . . . . . . . . . . . . . . . . 16-1

16.1 Overview .................................................................................................................16-1
ii
PSS®E 33.4
16.2 The dc Exciters ........................................................................................................16-1

16.2.1 Separately Excited Exciter ..........................................................................16-5
16.2.2 Shunt Excited Exciter .................................................................................16-6
16.2.3 Exciter Operating Practice ..........................................................................16-6
16.2.4 Automatic Calculation of dc Exciter Parameters in PSS®E Models ...........16-8
16.3 The ac Excitation Systems ....................................................................................16-10
16.3.1 Components .............................................................................................16-10
16.3.2 The ac Exciters .........................................................................................16-10
16.3.3 Compounded Shunt ac Excitation ............................................................16-12
16.3.4 Field Current Rectifiers .............................................................................16-14
16.4 System Voltage Regulators and Control Elements ...............................................16-17
16.4.1 Transfer Functions ....................................................................................16-17
16.4.2 Voltage Regulator Limits ..........................................................................16-20
16.5 Specific Excitation System Models ........................................................................16-21
16.5.1 Models IEEET1, IEET1A, IEEEX1, ESDC1A ...........................................16-21
16.5.2 Models IEEET2, IEEEX2, IEEX2A ...........................................................16-21
16.5.3 Models IEEET3, IEEEX3 ..........................................................................16-21
16.5.4 Models IEEET4, IEEEX4, IEEET5, IEET5A .............................................16-22
16.5.5 Models SEXS and REXSYS .....................................................................16-22
16.5.6 Model SCRX .............................................................................................16-23
16.5.7 Models EXDC2 and ESDC2A ...................................................................16-24
16.5.8 Models EXAC1 and ESAC1A ...................................................................16-25
16.5.12 Models EXST1 and ESST1A ....................................................................16-26
16.5.13 Models EXST2, EXST2A and ESST2A ....................................................16-28
16.5.14 Models EXST3 and ESST3A ....................................................................16-28
16.5.15 Model ESAC5A .........................................................................................16-28
16.5.16 Model IEET1B ..........................................................................................16-28
16.5.17 Model EXPIC1 ..........................................................................................16-29
16.5.18 Type ESAC6A Excitation System .............................................................16-29
16.5.19 Model EXBAS ...........................................................................................16-29
16.5.20 Model ESAC8B .........................................................................................16-29
16.5.21 Model EXELI .............................................................................................16-29
16.5.22 Model ESST4B .........................................................................................16-30
16.5.23 Models REXSYS and REXSY1 ................................................................16-30
16.6 Alternate Voltage Regulator Inputs ........................................................................16-30
16.6.1 Principle ....................................................................................................16-30
16.6.2 Model IEEEVC ..........................................................................................16-30
16.6.3 Model COMP ............................................................................................16-31
16.6.4 Model REMCMP .......................................................................................16-31
16.6.5 Model COMPCC .......................................................................................16-32
16.7 Supplementary Excitation Controller Models .........................................................16-32
16.7.1 General Principles ....................................................................................16-32
16.7.2 Model IEEEST ..........................................................................................16-33
16.7.3 Model STAB1 ...........................................................................................16-34
16.7.4 Models STAB2A, STAB3, and STAB4 .....................................................16-35
16.7.5 Model IEE2ST ..........................................................................................16-35
16.7.6 Model ST2CUT .........................................................................................16-35
iii
PSS®E 33.4
16.7.7 Model PTIST1 ...........................................................................................16-35

16.7.8 Model PTIST3 ...........................................................................................16-35
16.7.9 Model STBSVC ........................................................................................16-38
16.7.10 PSS2A ......................................................................................................16-38
16.8 Models MNLEX1, MNLEX2 and MNLEX3 .............................................................16-38
Speed Governor System Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-1

17.1 Overview .................................................................................................................17-1
17.2 Turbine-Governor Data ...........................................................................................17-2
17.3 Model TGOV1 .........................................................................................................17-2
17.4 Gas Turbine-Governors ...........................................................................................17-2
17.4.1 Model GAST ...............................................................................................17-2
17.4.2 Model GAST2A ...........................................................................................17-4
17.4.3 Model GASTWD .........................................................................................17-4
17.4.4 Model WESGOV .........................................................................................17-4
17.5 Hydro Governor Models ..........................................................................................17-4
17.5.1 Linear Models .............................................................................................17-4
17.5.2 Model HYGOV ............................................................................................17-6
17.5.3 Model HYGOVM .........................................................................................17-8
17.5.4 Model WEHGOV ......................................................................................17-13
17.5.5 Model HYGOVT ........................................................................................17-16
17.5.6 Model PIDGOV .........................................................................................17-18
17.6 Model IEESGO ......................................................................................................17-18
17.7 Model HYGOV2 .....................................................................................................17-18
17.8 Model TGOV2 .......................................................................................................17-20
17.9 Model IEEEG1 .......................................................................................................17-21
17.10 Models IEEEG2 and IEEEG3 ................................................................................17-21
17.11 Model TGOV3 .......................................................................................................17-22
17.12 Model CRCMGV ....................................................................................................17-22
17.13 Models DEGOV and DEGOV1 ..............................................................................17-22
17.14 Model SHAF25 ......................................................................................................17-23
17.14.1 Data Preparation for SHAF25 ...................................................................17-23
17.14.2 Calculation of Inertias for Masses ............................................................17-24
17.14.3 Shaft Stiffness Calculation ........................................................................17-24
17.14.4 Mass Damping ..........................................................................................17-25
17.14.5 Data Calculation Check ............................................................................17-25
17.14.6 Time Step when Using SHAF25 ...............................................................17-25
17.15 Model TGOV4 .......................................................................................................17-26
17.15.1 Governor Model ........................................................................................17-26
17.15.2 Boiler Model ..............................................................................................17-27
17.15.3 Valve Model ..............................................................................................17-29
17.15.4 Control of Fast Valving .............................................................................17-33
17.15.5 PLU Relay ................................................................................................17-33
17.15.6 EVA Relay ................................................................................................17-34
17.16 Model WPIDHY .....................................................................................................17-34
iv
PSS®E 33.4
17.17 Model GGOV1 .......................................................................................................17-35

17.18 WSCC Governor Models .......................................................................................17-36
17.19 Turbine Load Controllers .......................................................................................17-36
Modeling dc Transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-1

18.1 General Considerations ...........................................................................................18-1
18.1.1 Modeling Asymmetries in the ac System ....................................................18-6
18.1.2 Blocking, Bypassing, and Commutation Failures .......................................18-6
18.1.3 Capacitor Commutated Converters ............................................................18-6
18.2 Model CDC4 ............................................................................................................18-7
18.2.1 Normal Operation .......................................................................................18-7
18.2.2 Operation During Transients ......................................................................18-9
18.2.3 Modulating Control ...................................................................................18-13
18.2.4 Use of CDC4 when Bridge Firing Angle is Initialized on Power Flow Limit ..18-
13
18.2.5 Calling Sequence .....................................................................................18-14
18.2.6 Example Cases ........................................................................................18-15
18.3 Model CDC6 ..........................................................................................................18-31
18.4 Model CDC6A ........................................................................................................18-32
18.5 CDC7T Model for the HVDC Cable and/or Overhead Line ..................................18-34
18.5.1 Introduction ...............................................................................................18-34
18.5.2 DC Circuit arrangement simulated by the CDC7T model ........................18-34
18.5.3 Controls Algorithm ....................................................................................18-36
18.5.4 Parameters of the DC circuit ....................................................................18-44
18.6 Models PAUX1, PAUX2, SQBAUX, CPAAUX, HVDCAU .....................................18-45
18.7 Models MTDC01 and MTDC03 .............................................................................18-45
18.7.1 General Description ..................................................................................18-45
18.7.2 Example ....................................................................................................18-47
18.8 Models CASEA1 and CDCRL ...............................................................................18-57
18.8.1 Internal Integration Used by Model ...........................................................18-58
18.8.2 Initializing Thevenin Impedances .............................................................18-59
18.8.3 Simulating dc Faults .................................................................................18-59
18.8.4 Simulating ac System Faults at the Converter Terminals .........................18-60
18.9 Model CDCVUP .....................................................................................................18-60
18.9.1 Simulating ac System Faults at the Converter Terminals .........................18-60
18.10 Model DCPOW ......................................................................................................18-60
18.11 Model CDCAB1 .....................................................................................................18-65
18.11.1 Master Controller ......................................................................................18-66
18.11.2 Frequency Control ....................................................................................18-68
18.11.3 Damping Control .......................................................................................18-68
18.11.4 Emergency Power Control ........................................................................18-68
18.11.5 Voltage Dependent Current Order Limiter (VDCOL) ................................18-69
18.11.6 Current Control Amplifier (CCA) ...............................................................18-70
18.11.7 Alpha-Max Segment .................................................................................18-73
18.11.8 Converger Firing Control (CFC) ...............................................................18-74
18.11.9 Converters and Converter Transformers ..................................................18-77
18.11.10Cable Model and Current Return Path ....................................................18-80
v
PSS®E 33.4
18.11.11Special Control Functions .......................................................................18-81

18.12 CDCAB1 Example Case .......................................................................................18-86
18.12.1 Power Flow Data (RAWD File) for CDCAB1 Example Case ....................18-86
18.12.2 Dynamics Data (DYRE File) for CDCAB1 Example Case ........................18-86
18.12.3 DOCU Report ...........................................................................................18-87
Transmission Line and Other Relay Models . . . . . . . . . . . . . . . . . . . . . . . . 19-1

19.1 Principles .................................................................................................................19-1
19.2 Relay Elements .......................................................................................................19-1
19.2.1 Impedance Detection Unit ..........................................................................19-1
19.2.2 Zone Timer .................................................................................................19-1
19.2.3 Circuit Breaker Timer ..................................................................................19-5
19.2.4 Circuit Breaker ............................................................................................19-5
19.2.5 Reclosure Timer .........................................................................................19-5
19.2.6 Minimum Current Detector .........................................................................19-5
19.2.7 Supervisory Input - Permissive Flag ...........................................................19-6
19.2.8 Blinders ......................................................................................................19-6
19.2.9 Relay Time-Gating Unit ..............................................................................19-6
19.3 Bases for Relay Data ..............................................................................................19-6
19.4 Model DISTR1 .........................................................................................................19-9
19.4.1 General Characteristics ..............................................................................19-9
19.4.2 Specific Characteristics ............................................................................19-11
19.4.3 Data ..........................................................................................................19-11
19.4.4 Example Application - DISTR1 .................................................................19-12
19.5 Model RXR1 ..........................................................................................................19-30
19.6 Model CIROS1 ......................................................................................................19-30
19.6.1 Operation ..................................................................................................19-30
19.7 Model SLNOS1 .....................................................................................................19-35
19.8 Model DPDTR1 .....................................................................................................19-39
19.9 Model TIOCR1 ......................................................................................................19-39
19.10 Advanced Uses of Branch Relay Models ..............................................................19-41
19.10.1 DPDTR1 Flags .........................................................................................19-41
19.10.2 TIOCR1 Flags ...........................................................................................19-41
19.10.3 DISTR1 Flags ...........................................................................................19-41
19.10.4 RXR1 Flags ..............................................................................................19-41
19.10.5 CIROS1 Flags ..........................................................................................19-41
19.10.6 SLNOS1 Flags ..........................................................................................19-41
19.10.7 SLLP1 Flags .............................................................................................19-41
19.10.8 LOEXR1 Flags ..........................................................................................19-42
19.11 Model SLLP1 .........................................................................................................19-42
19.12 Model LOEXR1 .....................................................................................................19-43
19.13 Model SLYPN1 ......................................................................................................19-45
19.14 Model SCGAP2 .....................................................................................................19-47
vi
PSS®E 33.4
Load Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1

20.1 Introduction – Application of Load Models ...............................................................20-1
20.2 Load Characteristic Models .....................................................................................20-2
20.2.1 Basic Considerations ..................................................................................20-2
20.2.2 Algebraic Characteristic Models .................................................................20-3
20.2.3 Complex Models .........................................................................................20-4
20.2.4 Induction Motor Models ..............................................................................20-7
20.2.5 Single-Phase AC Motor Model .................................................................20-22
20.2.6 Composite Load Models ...........................................................................20-27
20.3 Load Relay Models ................................................................................................20-57
20.3.1 Basic Considerations ................................................................................20-57
20.3.2 Underfrequency Load Shedding Models ..................................................20-57
20.3.3 Undervoltage Load Shedding Models ......................................................20-60
20.3.4 Under voltage and underfrequency load shedding models ......................20-60
Generic Wind Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-1

21.1 Generic WT1 (Type 1) Model ..................................................................................21-1
21.1.1 Introduction .................................................................................................21-1
21.1.2 Power Flow Setup ......................................................................................21-2
21.1.3 Dynamic Setup ...........................................................................................21-3
21.2 Generic WT2 (Type 2) Model ..................................................................................21-6
21.2.1 Introduction .................................................................................................21-6
21.2.2 Power Flow Setup ......................................................................................21-7
21.2.3 Dynamic Setup ...........................................................................................21-8
21.3 Generic WT3 (Type 3) Model ................................................................................21-12
21.3.1 Introduction ...............................................................................................21-12
21.3.2 Power Flow Setup ....................................................................................21-12
21.3.3 Dynamic Setup .........................................................................................21-13
21.4 Generic WT4 (Type 4) Model ................................................................................21-22
21.4.1 Introduction ...............................................................................................21-22
21.4.2 Power Flow Setup ....................................................................................21-23
21.4.3 Dynamic Setup .........................................................................................21-23
Photovoltaic (PV) System Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1

22.1 Introduction ..............................................................................................................22-1
22.1.1 Load flow Representation ...........................................................................22-1
22.1.2 Dynamic Setup ...........................................................................................22-1
22.2 Example Dynamic Input Data ..................................................................................22-3
Elementary Blocks for Handling Transfer Functions in Dynamic Models 23-1

23.1 General ....................................................................................................................23-1
23.2 Integrator .................................................................................................................23-3
23.3 Non-windup Integrator .............................................................................................23-4
23.4 First Order (Lag) Block ............................................................................................23-6
23.5 First Order (Lag) Block with Non-Windup Limits .....................................................23-8
vii
PSS®E 33.4
23.6 Wash-Out Block ....................................................................................................23-10

23.7 Lead-Lag Block .....................................................................................................23-11
23.8 Lead-Lag Block with Non-Windup Limits ...............................................................23-13
23.9 Proportional-Integral (PI) Block .............................................................................23-15
23.10 Proportional-Integral (PI) Block with Non-Windup Limits ......................................23-17
23.11 Proportional-Integral-Derivative (PID) Block .........................................................23-19
23.12 Proportional-Integral-Derivative (PID) Block with Non-Windup Limits ...................23-21
23.13 Second Order Block ..............................................................................................23-24
Other Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-1

24.1 Saturating Reactive Load ........................................................................................24-1
24.1.1 Model SAT1 ................................................................................................24-1
24.1.2 Model SAT2 ................................................................................................24-3
24.2 Models Requiring User-Supplied Logic ...................................................................24-3
24.2.1 Model LINTRP ............................................................................................24-4
24.2.2 Model BSDSCN ..........................................................................................24-4
24.2.3 Model GENTRP ..........................................................................................24-4
24.3 Model NETFRQ .......................................................................................................24-4
24.4 System-Wide Monitoring Models .............................................................................24-5
24.4.1 Simulation Option Monitoring Models .........................................................24-5
24.4.2 Model SYSANG Called from CHSB ...........................................................24-8
24.5 Playback Model 'TSTGOV1' ....................................................................................24-8
24.5.1 Use of TSTGOV1 to Simulate a Grid Code Frequency Shape ...................24-9
24.5.2 Use of TSTGOV1 to Simulate a Measured Frequency Response ...........24-11
24.5.3 Example of the Use of TSTGOV1 to Simulate a Measured Frequency
Response 24-12
24.6 Equipment Monitoring Models ...............................................................................24-13
24.6.1 Models Set Up by Activities CHAN and CHSB .........................................24-13
24.6.2 Additional Monitoring Models ...................................................................24-14
24.7 Reactive Switching Devices ..................................................................................24-14
24.7.1 Model SWCAP ..........................................................................................24-14
24.7.2 Model SWSHN1 .......................................................................................24-15
24.8 FACTS Devices .....................................................................................................24-15
24.8.1 CRANI (also known as TCSC) .................................................................24-16
24.8.2 Static Condenser Model - CSTATT and CSTCNT ...................................24-16
24.8.3 CSMEST ...................................................................................................24-17
24.8.4 CBEST (EPRI Battery Energy Storage) ...................................................24-21
24.8.5 CDSMS1 ...................................................................................................24-21
24.8.6 SVSMO1U1 - Model of continuously controlled SVC ...............................24-25
24.8.7 SVSM02U1 - Model of discretely controlled SVC .....................................24-29
24.8.8 SVSM03U1 - Model of VSC based SVC ..................................................24-33
24.9 Under/Over Voltage/Frequency Relay Models ......................................................24-36
24.10 Voltage Source Converter DC Line Dynamic Model .............................................24-37
24.10.1 VSC Dynamic Model Overview ................................................................24-37
24.10.2 VSCDYN Module ......................................................................................24-38
viii
PSS®E 33.4
24.10.3 DCLINE Module ........................................................................................24-40

24.10.4 VSC DC Power Flow Representation .......................................................24-41
24.11 Variable Frequency Transformer (VFT) Model ......................................................24-42
24.11.1 Limitations ................................................................................................24-43
24.11.2 Example ....................................................................................................24-43
24.12 Modeling for Stability Analysis ...............................................................................24-45
Data Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-1

25.1 General Approach ...................................................................................................25-1
25.1.1 Stages of Verification ..................................................................................25-1
25.1.2 Parameter Range Checking .......................................................................25-1
25.1.3 Performance Verification ............................................................................25-4
25.2 Generator Data ........................................................................................................25-5
25.2.1 Reactances and Saturation Function .........................................................25-5
25.2.2 Generator V-Curves ...................................................................................25-5
25.2.3 Use of Typical Machine Data ......................................................................25-8
25.3 Excitation System Data .........................................................................................25-14
25.3.1 General Considerations ............................................................................25-14
25.3.2 Exciter Ceiling ...........................................................................................25-14
25.3.3 Checking of Automatically Initialized Parameters .....................................25-20
25.3.4 Excitation Loop Tuning .............................................................................25-23
25.3.5 General Notes on Excitation System Data Review ..................................25-29
25.4 Turbine Governor Data Verification .......................................................................25-29
25.4.1 Governor Response Test .........................................................................25-29
25.4.2 Governor Data Verification Example ........................................................25-30
25.5 Parameter Ranges for Activity DOCU ...................................................................25-36
25.6 Other Data Assumptions Made By Models ............................................................25-59
Linear Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26-1

26.1 Overview ..................................................................................................................26-1
26.1.1 State Space Formulation of a Linear System .............................................26-1
26.1.2 Eigenvalues ................................................................................................26-2
26.1.3 Eigenvectors ...............................................................................................26-3
26.1.4 Linear Independence of Eigenvectors ........................................................26-3
26.1.5 Modal Formulation of a Linear System .......................................................26-3
26.1.6 Participation Factors ...................................................................................26-4
26.2 PSS®E Interface - Inferring Linear System Matrices ...............................................26-5
26.3 Accessing Linear System Activities/Picking Up the Matrices ................................26-13
26.4 Listing the A, B, H, and F Matrices ........................................................................26-14
26.5 Calculating Eigenvalues and Eigenvectors ...........................................................26-15
26.6 Documenting the Eigensystem ..............................................................................26-15
26.7 Modal Form Transformation of Activity DIAG ........................................................26-21
26.8 Calculating and Listing Participation Factors .........................................................26-22
26.9 Frequency Response Calculations ........................................................................26-24
ix
PSS®E 33.4
26.10 Plotting Results .....................................................................................................26-28

26.11 Small Perturbation Simulations .............................................................................26-32
26.12 Stopping the Program ...........................................................................................26-36
26.13 Timing Statistics ....................................................................................................26-36
26.14 The Null Activity .....................................................................................................26-36
26.15 Report Output Devices ..........................................................................................26-37
26.16 Activity ECHO ........................................................................................................26-37
26.17 Activity HELP .........................................................................................................26-38
26.18 Directory Listing .....................................................................................................26-38
26.19 Dialog Input Selection ...........................................................................................26-38
26.20 Changing Elements of the Matrices ......................................................................26-39
Extended Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-1

27.1 Introduction ..............................................................................................................27-1
27.2 Explicit and Implicit Integration Algorithms ..............................................................27-2
27.2.1 Explicit Integration Algorithm ......................................................................27-2
27.2.2 Implicit or Trapezoidal Algorithm ................................................................27-5
27.3 Modes of Extended Term Simulation ......................................................................27-8
27.4 Additional Models for Extended Term Simulations ..................................................27-8
27.4.1 Online Tap Changer, OLTC1 ......................................................................27-8
27.4.2 Online Phase Shifter, OLPS1 ...................................................................27-11
27.4.3 The dc Line Tap Changer, DCTC1 ...........................................................27-11
27.4.4 Maximum Excitation Limiters, MAXEX1 and MAXEX2 .............................27-11
27.4.5 EXTL Type Models ...................................................................................27-13
27.5 TGOV5 ..................................................................................................................27-14
27.5.1 Introduction ...............................................................................................27-14
27.5.2 Governor Model ........................................................................................27-14
27.5.3 Fuel and Boiler Models .............................................................................27-14
27.5.4 Pressure Set Point and Load Reference Control .....................................27-14
27.5.5 Model Constants .......................................................................................27-15
27.5.6 Typical Data ..............................................................................................27-16
27.6 Running Extended Term Simulations ....................................................................27-19
27.7 The dc Line Auxiliary Signals ................................................................................27-24
27.8 Converting a State Space Model to Extended Term .............................................27-24
27.8.1 Background, Prerequisites, and Requirements ........................................27-24
27.8.2 Writing the Basic Model ............................................................................27-26
27.8.3 Models With Feedback and Limits ............................................................27-34
x
List of Figures
12-1 Differential Equations Describing Behavior of Simple Planetary System . . . . . . . . . .12-2

12-2 Graphical Illustration of Simulation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-3
12-3 Simple Simulation of Planetary Motion Modeled as in Figure 12-1 . . . . . . . . . . . . . .12-4
12-4 Simple Program to Simulate Motion of Rocket About a Heavenly Body . . . . . . . . . .12-5
12-5 Rewritten Dynamic Simulation Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-7
12-6 Basic Program Structure of PSS®E Dynamic Simulation Section . . . . . . . . . . . . . . .12-8
12-7 Comparison of Numerical Integrations of Equation (1.4) . . . . . . . . . . . . . . . . . . . . .12-10
12-8 Differential Equations for Generator Connected to Infinite Bus . . . . . . . . . . . . . . . .12-11
12-9 Use of PSS®E Array Elements to Accommodate System Parameters and Variables 12-
14
12-10 Assignment of PSS®E Array Space to Equipment Model Subroutines
Called from Subroutine CONEC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-16
12-11 Dynamic Simulation Basic Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-18
12-12 Dynamic Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-20
12-13 PSS®E Output Channel Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-23
13-1 Small Example System for Dynamic Simulation Setup Examples . . . . . . . . . . . . . . .13-2
13-2 Input Data for Small Example System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-3
13-3 Preparation of Power Flow Case for Use in Dynamic Simulation . . . . . . . . . . . . . . .13-5
13-4 Data for Small Example System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-7
13-5 Sample Dynamic Data Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-9
13-6 Dynamic Data Used for Representation of One Hydro Unit . . . . . . . . . . . . . . . . . . .13-21
13-7 Basic Dynamics Data File for Setup of Simulation Model . . . . . . . . . . . . . . . . . . . .13-22
13-8 Dialog of Basic PSS®E Dynamic Simulation Model Setup . . . . . . . . . . . . . . . . . . .13-24
13-9 Basic CONEC and CONET Subroutines as Built by Activity DYRE . . . . . . . . . . . . .13-27
13-10 Checkout of Dynamics Simulation Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-29
13-11 Phasor Diagram Used in Initializing Synchronous Machines . . . . . . . . . . . . . . . . . .13-35
13-12 Use of Activity DOCU to Review CON Values for Reasonableness . . . . . . . . . . . .13-38
13-13 Error Conditions Detected by STRT and Use of DOCU to Track Down Errors . . . .13-42
13-14 Dialog for Dynamic Simulation Run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-45
13-15 Dialog for Plotting Program PSSPLT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-49
13-16 Representative Plot of the Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13-50
13-17 Distortion of Simulation Display by Adjustment of Plotting . . . . . . . . . . . . . . . . . . . .13-57
xi
List of Figures PSS®E 33.4
Program Application Guide: Volume I
13-18 Implied Linkage via Activities CHNG and FACT from Activity ALTR . . . . . . . . . . . 13-58
13-19 Use of NEW SAVE CASE Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-60
13-20 Simple Continuation of Dynamic Simulation Run . . . . . . . . . . . . . . . . . . . . . . . . . . 13-64
13-21 Restarting Simulation Run with Immediate Application of New Perturbation . . . . . 13-65
13-22 Use of Snapshot, Saved Case, and Dynamics Output Files . . . . . . . . . . . . . . . . . . 13-67
13-23 Management of Dynamics Output Files in Switching Operation Runs . . . . . . . . . . 13-68
13-24 Network Frequency Dependence Model (NETFRQ) Model Data Sheet . . . . . . . . . 13-72
13-25 Frequency Sensitive Load Model (LOADF) Model Data Sheet . . . . . . . . . . . . . . . 13-73
13-26 Underfrequency Load Shedding Relay Model (LDSHD) Model Data Sheet . . . . . . 13-75
13-27 Additional Dynamics Data File for Adding to Basic Model . . . . . . . . . . . . . . . . . . . 13-75
13-28 Dialog for Adding More Raw Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-76
13-29 Files Created by DYRE and CHAN in Making New Setup . . . . . . . . . . . . . . . . . . . 13-77
13-30 Editing CONETs to Combine Old and New Setups . . . . . . . . . . . . . . . . . . . . . . . . 13-78
13-31 DOCU of New Models Only . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-79
13-32 Listing of New Data by Activity DLST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-80
13-33 Simulation Run with New Setup (Section 13.5) . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-82
13-34 Simulation Run with New Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-84
14-1 Generator Model Equivalent Current Source and Norton Equivalent Circuit . . . . . . 14-3
14-2 Electromagnetic Model of Salient Pole Generator . . . . . . . . . . . . . . . . . . . . . . . . . . 14-7
14-3 Electromagnetic Model of Round Rotor Generator . . . . . . . . . . . . . . . . . . . . . . . . . . 14-9
14-4 Electromagnetic Model of Salient Pole Generator . . . . . . . . . . . . . . . . . . . . . . . . . 14-10
14-5 Typical Generator Open-Circuit Saturation Curves . . . . . . . . . . . . . . . . . . . . . . . . 14-10
14-6 Definition of Saturation Factor, S, for Entry as Generator Data . . . . . . . . . . . . . . . 14-11
14-7 Composition of Generator Transient Torque in Response to Switching . . . . . . . . . 14-16
14-8 CGEN1 Equivalent Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-18
14-9 CGEN1 Data Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-20
14-10 Equivalent Circuit for Machine Model Used in the Simulation . . . . . . . . . . . . . . . . 14-23
14-11 Comparison Between GENROU and CGEN1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-24
14-12 Comparison Between CGEN1 With and Without Subtransient Saliency . . . . . . . . 14-26
15-1 Control of Reactor Current by Thyristors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-2
15-2 Static var Source Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-3
15-3 Controlled Reactor Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-4
15-4 Form of Static var System Representation Used in PSS®E . . . . . . . . . . . . . . . . . . . 15-5
15-5 Simple Static var Device Control Arrangement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-5
15-6 Sample System for Application of Model CSVGN1 . . . . . . . . . . . . . . . . . . . . . . . . 15-11
15-7 Power Flow Data for CSVGN1 Application Example . . . . . . . . . . . . . . . . . . . . . . . 15-12
15-8 Base Case for CSVGN1 Application Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-14
xii
PSS®E 33.4 List of Figures
15-9 Dynamic Simulation Setup Main Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-16

15-10 Initialization of Dynamic Simulation Run Using CSVGN1 . . . . . . . . . . . . . . . . . . . .15-16
15-11 System Response to Fault at Bus 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-17
15-12 System Response to Fault at Bus 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-18
15-13 One-Line Impedance Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-20
15-14 Solved Power Flow Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-21
15-15 Dynamic Equipment Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-22
15-16 L-G Fault, No Supplementary Control, Hydro Unit . . . . . . . . . . . . . . . . . . . . . . . . . .15-25
15-17 L-G Fault, No Supplementary Control, Steam Unit . . . . . . . . . . . . . . . . . . . . . . . . .15-26
15-18 L-G Fault, No Supplementary Control, SVC-Bus 12 . . . . . . . . . . . . . . . . . . . . . . . .15-27
15-19 L-G Fault, Supplementary Control, Hydro Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-28
15-20 L-G Fault, Supplementary Control, Steam Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-29
15-21 L-G Fault, Supplementary Control, Steam Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-30
15-22 Rotor Angle Difference for the Two Runs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-31
15-23 PSS®E Model of CSVGN1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-33
15-24 Bode Plot of Simplified SVG Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-34
16-1 Excitation Power Source Alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16-2
16-2 Separately and Shunt Excited Arrangements for Main Exciters . . . . . . . . . . . . . . . .16-3
16-3 Nonlinear Transfer Function Model for Rotating Exciter . . . . . . . . . . . . . . . . . . . . . .16-4
16-4 Effect of Field Resistance on Output of Shunt-Excited Exciter . . . . . . . . . . . . . . . . .16-7
16-5 Transfer Function for ac Exciter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16-11
16-6 Current-Compounded Excitation Transformer Arrangement . . . . . . . . . . . . . . . . . .16-13
16-7 Characterization of Excitation Rectifier Commutation Voltage Drop . . . . . . . . . . . .16-15
16-8 Combined Alternator-Rectifier Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16-16
16-9 Combined Compounding Transformer-Rectifier Model . . . . . . . . . . . . . . . . . . . . . .16-16
16-10 Excitation System Overall General Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16-17
16-11 General Required Form of Voltage Regulator Transfer Function . . . . . . . . . . . . . .16-19
16-12 Use of Rate Feedback to Produce Transient Gain Reduction in Excitation System 16-19
16-13 Simplified Excitation System With Rate Feedback . . . . . . . . . . . . . . . . . . . . . . . . .16-20
16-14 Controlled Rectifier Excitation System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16-24
16-15 Overall Synchronous Machine Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16-33
17-1 Speed Governor and Turbine in Relationship to Generator . . . . . . . . . . . . . . . . . . . .17-1
17-2 Hydraulic and Governor Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17-6
17-3 Penstock Dynamics Loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17-10
17-4 HYGOVM Governor System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17-11
17-5 Short-Term versus Midterm GSTR/GRUN Hydro Governor Test . . . . . . . . . . . . . .17-14
17-6 Traveling-Wave Model HYGOVT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17-17
xiii
17-7 Elastic versus Inelastic Hydro Governor Models Comparison . . . . . . . . . . . . . . . . 17-19

17-8 Elastic versus Inelastic Hydro Governor Models Comparison . . . . . . . . . . . . . . . . 17-20
17-9 Sample Shaft System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-23
17-10 Numbered Shaft Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-23
17-11 Model TGOV4 Logic Diagram for PLU and EVA . . . . . . . . . . . . . . . . . . . . . . . . . . 17-28
17-12 Model TGOV4 Block Diagram for PLU and EVA . . . . . . . . . . . . . . . . . . . . . . . . . . 17-29
17-13 Per Unit Valve Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-31
17-14 Control Signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-32
18-1 Arrangement of dc Transmission Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-2
18-2 Forced Control of dc Converter Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-4
18-3 Ranges of Alpha and Gamma Angles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-8
18-4 CDC4 dc Transmission Control Arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-8
18-5 Use of Bypass Switch to Allow Shutdown of Inverter . . . . . . . . . . . . . . . . . . . . . . . 18-10
18-6 Examples of dc Voltage and Current Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-11
18-7 Voltage-Dependent Current Limit for dc Converters . . . . . . . . . . . . . . . . . . . . . . . . 18-12
18-8 Power Flow Data for dc Lines and Synchronous Condensers . . . . . . . . . . . . . . . . 18-16
18-9 Permissible dc Line Operating Domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-19
18-10 dc Line Dynamics Data to Complement Power Flow Data of Figure 18-8 . . . . . . . 18-20
18-11 Use of DYRE to Set Up CONEC and CONET Calling CDC4 . . . . . . . . . . . . . . . . . 18-21
18-14 Response to High Impedance L-G Fault at Bus 1200 . . . . . . . . . . . . . . . . . . . . . . 18-29
18-15 CDC6 dc Transmission Control Arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-33
18-16 CDC6 Blocking and Unblocking Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-33
18-17 A DC Circuit Arrangement Simulated by the CDC7T Model . . . . . . . . . . . . . . . . . 18-35
18-18 Basic Converter Control Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-38
18-19 VDCOL Controller for CDC7T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-39
18-20 Typical VDCOL Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-39
18-21 VDCOL DC Current Order for Rectifier and Inverter . . . . . . . . . . . . . . . . . . . . . . . . 18-41
18-22 Example of Complex Simulation for a dc Line with the Long Cable . . . . . . . . . . . . 18-43
18-23 Another Example of Complex Simulation for a dc Line with the Long Cable . . . . . 18-43
18-24 Converter Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-46
18-25 Voltage-Dependent Current Limit for dc Converters . . . . . . . . . . . . . . . . . . . . . . . . 18-46
18-26 Voltage-Controlling Converter dc Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-47
18-27 Example DOCU for Four-Terminal Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-48
18-28 Example Case Data Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-55
18-29 Example Code for Automatic Blocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-55
xiv
18-30 Example Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-57

18-31 Example DOCU for CDCVUP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-61
18-32 Response to Inverter Fault With Commutation Failures . . . . . . . . . . . . . . . . . . . . .18-63
18-33 Block Diagram Showing Power Control for dc System (DCPOW) . . . . . . . . . . . . . .18-65
18-34 Diagram of the HVDC System with Connections to the ac Systems . . . . . . . . . . . .18-67
18-35 Master Controller in Power Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-67
18-36 Frequency Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-69
18-37 Converted Block Diagram for the Damping Controller . . . . . . . . . . . . . . . . . . . . . . .18-69
18-38 Voltage Dependent Current Order Limiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-71
18-39 VDCOL Characteristic (Current Order as a Function of the dc Voltage) . . . . . . . . .18-71
18-40 Ud/Id Characteristic for the dc Model without the VDCOL . . . . . . . . . . . . . . . . . . . .18-72
18-41 Current Control Amplifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-73
18-42 max Segment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-74
18-43 CFC with Dynamic Current Compound and Phase Shift Correction . . . . . . . . . . . .18-75
18-44 Limits on the Change in  per Commutation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-76
18-45 CDCAB1 Algebraic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-78
18-46 Transient Controller Rectifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-82
18-47 Transient Controller Inverter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-83
18-48 Inverter Gamma Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-85
18-49 Voltage Control for Long Cables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-86
19-1 Setting and Resetting of Zone Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-3
19-2 Operation of Zone Timer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-4
19-3 Logic for Breaker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-5
19-4 Straight-Line Relay Blinders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-8
19-5 Operation Time Gating Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-9
19-6 Overall Logic Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-10
19-7 Application of DISTR1 to Circuits 1 and 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-14
19-8 DOCU Output Corresponding to Figure 19-7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-18
19-9 Dialog for Simulation with DISTR1 on Circuits 1 and 2 . . . . . . . . . . . . . . . . . . . . . .19-19
19-10 Locus of Apparent Z Seen by DISTR1 on Circuit 2 . . . . . . . . . . . . . . . . . . . . . . . . .19-21
19-11 Time Plots of Simulation Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-22
19-12 Relay Messages for Fault at Bus 1 End of Circuit 1 . . . . . . . . . . . . . . . . . . . . . . . . .19-24
19-13 Locus of Apparent Z Seen by DISTR1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-25
19-14 Time Domain Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-26
19-15 Relay Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-27
19-16 Locus of Apparent Z Seen by DISTR1 on Circuit 2 . . . . . . . . . . . . . . . . . . . . . . . . .19-28
19-17 Time Domain Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19-29
xv
19-18 Relay Model RXR1 Logic Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-31

19-19 Definition of Circular Zones in CIROS1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-32
19-20 Conditions for Setting/Resetting Timers in CIROS1 . . . . . . . . . . . . . . . . . . . . . . . . 19-33
19-21 CIROS1 Logic Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-34
19-22 SLNOS1 Single-Line Blinders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-36
19-23 Zone Definition for Single Blinder of SLNOS1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-36
19-24 SLNOS1 Relay Logic After Receiving Out-of-Step Signal . . . . . . . . . . . . . . . . . . . 19-37
19-25 SLNOS1 Double Line Blinders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-37
19-26 Zone Definition for Double Blinder SLNOS1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-38
19-27 DPDTR1 Logic Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-39
19-28 Time-Inverse Operating Characteristic of TIOCR1 . . . . . . . . . . . . . . . . . . . . . . . . . 19-40
19-29 Tripping on Way In . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-42
19-30 Tripping on Way Out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-43
19-31 Loss of Field Trajectory and Protective Relay . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-44
19-32 Directional Comparison Scheme Assumed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-46
20-1 Large Motor Performance Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-5
20-2 Small Motor Performance Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-6
20-3 Voltage versus Magnetizing Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-7
20-4 Circuit for Type 2 Specification in Program IMD . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-10
20-5 Use of IMD to Estimate Induction Motor Equivalent Circuit Parameters . . . . . . . . . 20-12
20-6 Plotted Induction Motor Performance from IMD . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-15
20-7 Induction Model With Rotor Flux Linkages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-18
20-8 CIM5 Type Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-19
20-9 CIMW Type Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-20
20-10 Performance Model Characteristics of Compressor Motors . . . . . . . . . . . . . . . . . . 20-23
20-11 Compressor Motor Thermal Relay Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-24
20-12 Model Structure of ACMTBL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-26
20-13 Composite Load Model Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-27
20-14 Induction Motor Equivalent Circuit for Type 1 Specification in Program IMD . . . . . 20-31
20-15 Adjustment of Bus Loads when Model CMOTOR is Initialized as Online . . . . . . . . 20-35
20-16 Setup of Power Flow Case to Accommodate CMOTOR Load . . . . . . . . . . . . . . . . 20-39
20-17 CMOTOR Model Data Sheet as Completed From Figure 20-15 . . . . . . . . . . . . . . 20-41
20-18 Initialization of Simulation Using Model CMOTOR . . . . . . . . . . . . . . . . . . . . . . . . . 20-42
20-19 Motor Behavior Following Fault at Bus 150 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-45
20-20 Expansion of Load Bus 151 into Separate Buses for SCR Load . . . . . . . . . . . . . . 20-46
20-21 CMOTOR Data for Off-Line Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-48
20-22 Setup of Dynamic Simulation for Motor Starting using Model CMOTOR . . . . . . . . 20-49
xvi
20-23 Behavior of Motor 1 – Direct Online Starting Against Full Load . . . . . . . . . . . . . . . .20-53
20-24 Generator Behavior Direct Online Starting Against Full Load . . . . . . . . . . . . . . . . .20-54
20-25 Motor Starting Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-55
20-26 Motor Starting Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20-55
20-27 Underfrequency Detection and Load Shedding in LDSH Type Models . . . . . . . . . .20-58
20-28 Time Inverse Operating Characteristic of LDST Type Models . . . . . . . . . . . . . . . . .20-59
21-1 Conventional Directly-Connected Induction Generator . . . . . . . . . . . . . . . . . . . . . . .21-2
21-2 WT1 Connectivity Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-3
21-3 WT12T Two-Mass Shaft model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-4
21-4 Pseudo-Governor Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-5
21-5 Example Dynamic Data Input File, WT1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-5
21-6 Example of Documentation, WT1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-6
21-7 Wound Rotor Induction Generator with Variable Rotor Resistance Control . . . . . . .21-7
21-8 WT2 Connectivity Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-8
21-9 WT2E Electrical Control model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-9
21-10 Example Dynamic Data Input File, WT2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-10
21-11 Example of Documentation, WT2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-11
21-12 Doubly Fed Induction Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-12
21-13 Interaction among Generic Wind Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-13
21-14 WT3G1 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-14
21-15 WT3G2 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-15
21-16 WT3E1 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-16
21-17 WT3T1 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-17
21-18 WT3P1 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-18
21-19 Example Dynamic Data Input File (Set 1) with Reference to GE 1.5 MW Wind Turbine .
21-18
21-20 Example Dynamic Data Input File (Set 2) with Reference to GE 1.5 MW Wind Turbine .
21-19
21-21 Example of Documentation (Set 1) with Reference to GE 1.5 MW Wind Turbine . .21-19
21-22 Example of Documentation (Set 2) with Reference to GE 1.5 MW Wind Turbine . .21-21
21-23 A Generator Connected to the Grid via the Power Converter . . . . . . . . . . . . . . . . .21-23
21-24 WT4 Connectivity Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-24
21-25 WT4 Generator/Converter Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-25
21-26 WT4E Electrical Control Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-27
21-27 Example of Dynamic Data Input File for WT4 with reference to the GE 2.5 MW wind
turbine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21-28
21-28 Example of Documentation for WT4 with reference to the GE 2.5 MW wind turbine . .21-
29
xvii
21-29 Example of Dynamic Data Input File for WT4 with reference to the Siemens 2.3 MW wind
turbine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-30
21-30 Example of Documentation for WT4 with reference to the Siemens 2.3 MW wind turbine
21-31
22-1 PV Model Connectivity Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-2
23-1 Integrator Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-3
23-2 Variable Declaration for Integrator Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-3
23-3 FORTRAN Code for Initialization of Integrator Block (MODE=1) . . . . . . . . . . . . . . . 23-3
23-4 FORTRAN Code for Calculation of Derivatives of Integrator Block (MODE=2) . . . . 23-4
23-5 FORTRAN Code for Calculation of Output Integrator Block (MODE=3) . . . . . . . . . 23-4
23-6 Integrator Block with Non-Windup Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-4
23-7 Variable Declaration for Integrator Block with Non-Windup Limits . . . . . . . . . . . . . . 23-4
23-8 FORTRAN Code for Initialization of Integrator Block . . . . . . . . . . . . . . . . . . . . . . . . 23-5
23-9 FORTRAN Code for Calculation of Derivatives of Integrator Block . . . . . . . . . . . . . 23-5
23-10 FORTRAN Code for Calculation of Output Integrator Block . . . . . . . . . . . . . . . . . . . 23-5
23-11 First Order Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-6
23-12 PSS®E Implementation of the First Order Block . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-6
23-13 PSS®E Implementation of the First Order Block when T = 0 . . . . . . . . . . . . . . . . . . 23-6
23-14 Variable Declaration for First Order Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-6
23-15 FORTRAN Code for Initialization of First Order Block (MODE=1) . . . . . . . . . . . . . . 23-7
23-16 FORTRAN Code for Calculation of Derivatives of First Order Block (MODE=2) . . . 23-7
23-17 FORTRAN Code for Calculation of Output of First Order Block (MODE=3) . . . . . . . 23-7
23-18 First Order Block with Non-Windup Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-8
23-21 Variable Declaration for First Order Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-8
23-22 FORTRAN Code for Initialization of First Order Block . . . . . . . . . . . . . . . . . . . . . . . 23-9
23-23 FORTRAN Code for Calculation of Derivatives of First Order Block . . . . . . . . . . . . 23-9
23-24 FORTRAN Code for Calculation of Output of First Order Block . . . . . . . . . . . . . . . . 23-9
23-25 Wash-Out Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-10
23-26 PSS®E Implementation of the Wash-Out Block . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-10
23-27 Variable Declaration for Wash-Out Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-10
23-28 FORTRAN Code for Initialization of Wash-Out Block (MODE=1) . . . . . . . . . . . . . . 23-10
23-29 FORTRAN Code for Calculation of Derivatives of Wash-Out Block (MODE=2) . . . 23-11
23-30 FORTRAN Code for Calculation of Output of Wash-Out Block (MODE=3) . . . . . . 23-11
23-31 Lead-Lag Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-11
23-32 PSS®E Implementation of the Lead-Lag Block . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-12
23-33 Variable Declaration for Lead-Lag Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-12
xviii
23-34 FORTRAN Code for Initialization of Lead-Lag Block (MODE=1) . . . . . . . . . . . . . . .23-12

23-35 FORTRAN Code for Calculation of Derivatives of Lead-Lag Block (MODE=2) . . . .23-13
23-36 FORTRAN Code for Calculation of Output of Lead-Lag Block (MODE=3) . . . . . . .23-13
23-37 Lead-Lag Block with Non-Windup Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-13
23-38 PSS®E Implementation of the Lead-Lag Block . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-14
23-39 Variable Declaration for Lead-Lag Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-14
23-40 FORTRAN Code for Initialization of Lead-Lag Block . . . . . . . . . . . . . . . . . . . . . . . .23-14
23-41 FORTRAN Code for Calculation of Derivatives of Lead-Lag Block . . . . . . . . . . . . .23-15
23-42 FORTRAN Code for Calculation of Output of Lead-Lag Block . . . . . . . . . . . . . . . .23-15
23-43 Proportional-Integral (PI) Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-15
23-44 PSS®E Implementation of the Proportional-Integral (PI) Block . . . . . . . . . . . . . . . .23-16
23-45 Variable Declaration for Proportional-Integral (PI) Block . . . . . . . . . . . . . . . . . . . .23-16
23-46 FORTRAN Code for Initialization of Proportional-Integral (PI) Block (MODE=1) . . .23-16
23-47 FORTRAN Code for Calculation of Derivatives of Proportional-Integral (PI) Block
(MODE=2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-16
23-48 FORTRAN Code for Calculation of Output of Proportional-Integral (PI) Block (MODE=3)
23-17
23-49 Proportional-Integral (PI) Block with Non-Windup Limits . . . . . . . . . . . . . . . . . . . . .23-17
23-50 PSS®E Implementation of the Proportional-Integral (PI) Block . . . . . . . . . . . . . . . .23-17
23-51 Variable Declaration for Proportional-Integral (PI) Block . . . . . . . . . . . . . . . . . . . . .23-18
23-52 FORTRAN Code for Initialization of Proportional-Integral (PI) Block . . . . . . . . . . . .23-18
23-53 FORTRAN Code for Calculation of Derivatives of Proportional-Integral (PI) . . . . . .23-18
23-54 FORTRAN Code for Calculation of Output of Proportional-Integral (PI) Block . . . .23-19
23-55 Proportional-Integral-Derivative (PID) Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-19
23-56 PSS®E Implementation of the Proportional-Integral-Derivative (PID) Block . . . . . .23-19
23-57 Variable Declaration for Proportional-Integral-Derivative (PID) Block . . . . . . . . . . .23-20
23-58 FORTRAN Code for Initialization of Proportional-Integral-Derivative (PID) Block
(MODE=1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-20
23-59 FORTRAN Code for Calculation of Derivatives of Proportional-Integral-Derivative (PID)
Block (MODE=2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-20
23-60 FORTRAN Code for Calculation of Output of Proportional-Integral-Derivative (PID) Block
(MODE=3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23-21
23-61 Proportional-Integral-Derivative (PID) Block with Non-Windup Limits . . . . . . . . . . .23-21
23-62 PSS®E Implementation of the Proportional-Integral-Derivative (PID) Block . . . . . .23-22
23-63 Variable Declaration for Proportional-Integral-Derivative (PID) Block . . . . . . . . . . .23-22
23-64 FORTRAN Code for Initialization of Proportional-Integral-Derivative (PID) . . . . . . .23-23
23-65 FORTRAN Code for Calculation of Derivatives of Proportional-Integral- . . . . . . . . .23-23
23-66 FORTRAN Code for Calculation of Output of Proportional-Integral-Derivative . . . .23-24
xix
23-67 Second Order Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-24

23-68 PSS®E Implementation of the Second Order Block . . . . . . . . . . . . . . . . . . . . . . . . 23-24
23-69 Variable Declaration for Second Order Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-25
23-70 FORTRAN Code for Initialization of Second Order Block (MODE=1) . . . . . . . . . . . 23-25
23-71 FORTRAN Code for Calculation of Derivatives of Second Order Block (MODE=2) 23-25
23-72 FORTRAN Code for Calculation of Output of Second Order Block (MODE=3) . . . 23-26
24-1 Initialization of SAT1 Model Knee Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-2
24-2 Standard RELAY1 and RELAY2 Relay Circles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-6
24-3 Voltage Violation Check . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-7
24-4 One Bus Test System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-9
24-5 Defined Frequency Excursion for Primary Response Calculation . . . . . . . . . . . . . 24-10
24-6 Secondary Response Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-10
24-7 SVSMO1 Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-25
24-8 SVC arrangement used for testing SVSMO2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-30
24-9 SVSMO2U1 Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-32
24-10 Block Diagram of SVSMO3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-34
24-11 Modeling of short-term rating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-34
24-12 Example Dynamic Data Input File, SVSMO3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-36
24-13 VSCDCT PSS®E Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-38
24-14 Characteristic of Current Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-39
24-15 VSC Steady-State Reactive Capability Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-42
24-16 Single Line Diagram of the VFT and Associated GSU and Capacitor Bank . . . . . . 24-44
24-17 Data Record in PSS®E Power Flow Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-44
24-18 Block Diagram of the VFT Drive System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-45
24-19 Block Diagram of the Main Controller of the VFT . . . . . . . . . . . . . . . . . . . . . . . . . . 24-46
24-20 Block Diagram of the Torque Limiter of the VFT . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-46
24-21 Block Diagram of the VFT Governor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-47
25-1 Typical Output from Activity DOCU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-2
25-2 Typical Output from Data Checking Mode of Activity DOCU . . . . . . . . . . . . . . . . . . 25-3
25-3 Typical Generator V-Curves as Supplied by Manufacturer Corresponds to Table 1 25-6
25-4 Saturation Curve as Supplied by Manufacturer; Corresponds to Table 1 . . . . . . . . . 25-6
25-5 VCV Graphical Output for Case Where Full Machine Data is Known . . . . . . . . . . . . 25-7
25-6 Excitation Voltages at Rated MVA Using Typical Generator Data . . . . . . . . . . . . . . 25-9
25-7 Estimation of SE(1.2) for Use with Typical Generator Data . . . . . . . . . . . . . . . . . . 25-10
25-8 Typical Salient Pole Generator Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-11
25-9 Typical Solid Rotor Generator Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-12
25-10 Typical Excitation System Model and Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-13
xx
25-11 Sample Data for Excitation System Data Verification . . . . . . . . . . . . . . . . . . . . . . .25-17

25-12 Saturation Curves of Initial dc Exciter Model, (A) . . . . . . . . . . . . . . . . . . . . . . . . . . .25-17
25-13 Evaluation of Response Ratio by Activity ERUN . . . . . . . . . . . . . . . . . . . . . . . . . . .25-18
25-14 Exciter Responses in Response Ratio and Ceiling Tests for IEEEX1 . . . . . . . . . . .25-19
25-15 Initial Data Estimate for Compounded Transformer-Type Excitation System . . . . .25-22
25-16 Use of ESTR to Determine Value of EXST2 KI Parameter . . . . . . . . . . . . . . . . . . .25-24
25-17 Presetting of KI = 2.5 and Checking for Reasonableness . . . . . . . . . . . . . . . . . . . .25-25
25-18 Change of KC Parameter of EXST2 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25-26
25-19 Plotted Response from Open-Circuit Step Test of Model EXST2 . . . . . . . . . . . . . .25-27
25-20 Plotted Response of Open-Circuit Step Test for Model EXST2 . . . . . . . . . . . . . . . .25-28
25-21 Initial Trial Data for Hydro Turbine Governor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25-32
25-22 Use of Activity GSTR/GRUN to Check Hydro Turbine Governor Data . . . . . . . . . .25-33
25-23 Plotted Response of Governor Step Load Test . . . . . . . . . . . . . . . . . . . . . . . . . . . .25-34
25-24 Plotted Response of Governor Step Load Test . . . . . . . . . . . . . . . . . . . . . . . . . . . .25-35
26-1 Tree Structure Showing Flow of Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-5
26-2 Small Example System for Linear Analysis Examples . . . . . . . . . . . . . . . . . . . . . . . .26-7
26-3 Dynamics Raw Data File for Setup of Simulation Model of Small Example System .26-7
26-4 Flow Chart of Activity ASTR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-9
26-5 Creating the Linear Analysis Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-10
26-6 Linking to Linear Analysis Activities Picking-Up and Listing Matrices . . . . . . . . . . .26-16
26-7 Sample System Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-17
26-8 Obtaining and Listing Eigenvalues and Eigenvectors . . . . . . . . . . . . . . . . . . . . . . .26-18
26-9 Plot of Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-20
26-10 Plot of an Eigenvector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-21
26-11 Obtaining Modal Equivalents for Sample System . . . . . . . . . . . . . . . . . . . . . . . . . .26-23
26-12 Obtaining and Listing Participation Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-24
26-13 The Matrices of the Linear System Used in the Frequency Calculations . . . . . . . . .26-26
26-14 The Eigenvalues and Eigenvectors of the Linear System . . . . . . . . . . . . . . . . . . . .26-27
26-15 Frequency Response Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-29
26-16 Dialog to Create Bode Plot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-30
26-17 Bode Plot of Output 5 Varying with Input 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26-31
26-18 Natural Scale Polar Plot of Output 5 Varying with Input 3 . . . . . . . . . . . . . . . . . . . .26-33
26-19 Logarithmic Scale Polar Plot of Output 5 Varying with Input 3 . . . . . . . . . . . . . . . . .26-34
27-1 Modified Euler Integrator Z-Form Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27-2
27-2 Time-Domain Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27-3
27-3 Simulation of an Excitation System Controlling Generator Terminal Voltage . . . . . .27-4
27-4 Trapezoidal Integrator Z-Form Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27-5
xxi
27-5 Time-Domain Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-6

27-6 Z-Form Block Diagram Giving Discrete Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-6
27-7 Simulation of an Excitation System Controlling Generator Terminal Voltage . . . . . . 27-7
27-8 Example of Integration Timer for the OLTC1 and OLPS1 Models . . . . . . . . . . . . . . 27-9
27-9 Inverse Time Characteristics of MAXEX1 and MAXEX2 . . . . . . . . . . . . . . . . . . . . 27-12
27-10 Block Diagrams of MAXEX1 (Top) and MAXEX2 (Bottom) . . . . . . . . . . . . . . . . . . 27-12
27-11 Three Simulations of Artificial Damping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-22
27-12 Three Simulations of Artificial Damping Introduced by Increasing DELT . . . . . . . . 27-23
27-13 Z-Form Expression of Trapezoidal Integration Algorithm . . . . . . . . . . . . . . . . . . . . 27-26
27-14 Z-Form Code for Integrator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-27
27-15 Z-Form Code for Lag with Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-28
27-16 Z-Form for Lead/Lag or Lag/Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-29
27-17 Z-Form Code for Rate Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-30
27-18 Sample Exciter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-31
27-19 Z-Form Block Diagram of Sample Exciter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-31
27-20 Extended Term Code for Sample Exciter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-33
27-21 State Space Block Diagram of Exciter with Feedback and Nonwindup Limits . . . . 27-34
27-22 Z-Form Block Diagram of Exciter with Feedback and Nonwindup Limits . . . . . . . . 27-34
27-23 Mode 2 of Feedback Exciter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-36
xxii
Chapter 12
Dynamic Simulation
Principles
Chapter 12 - Dynamic Simulation Principles
12.1 Basic Dynamic Simulation

The dynamic simulation of a physical process has three basic steps:
1. Construction of a set of differential equations describing the behavior of the physical
system in general.
2. Determination of a set of values of constant and variable parameters describing, in

detail, the condition of the physical system at some instant.
3. Integration of the differential equations with the values determined in Step 2 as initial
conditions.
The overall process is best illustrated by an example.

A rocket, coasting after acceleration to a high velocity, would move around a heavenly
body according to Newton’s inverse square law as given by the four differential equations
(1.1) and Figure 12-1.
dvx -G x
m =
dt x2 + y2  x2 + y2
dvy -G y
m =
dt x2 + y2  x2 + y2
dx
= vx
dt
dy
= vy
dt
(1.1)
12-1
Dynamic Simulation Principles PSS®E 33.4
Basic Dynamic Simulation Program Application Guide: Volume II
vx
v
y
• m1
•
x
Figure 12-1. Differential Equations Describing Behavior of Simple Planetary System
If the values of the four variables x, y, vx, vy are known at some instant, t, their values at a
later instant, t + t, can be computed to first-order accuracy from (1.2).
dx
xnew = xold + t
dt
dy
ynew = yold + t
dt
dvx
v xnew = vxold + t
dt
dvy
v ynew = vyold + t
dt (1.2)
This process is illustrated in Figure 12-2. Numerical calculations are very simple because
the four variables where instantaneous values are known (vx, vy, x, y) are all that is needed
to complete the evaluation of the right-hand sides of (1.1).
12-2
PSS®E 33.4 Dynamic Simulation Principles
Program Application Guide: Volume II Basic Dynamic Simulation
new x
x
x x
old • t
•
t
t t+h
new dy
y old y ------
y • dt
•
t
old
vx new
v vx
x • dv x
• ---------
dt
v
y
old
vy new
• vy
• t
dv y
---------
dt
Figure 12-2. Graphical Illustration of Simulation Process
The flow chart of a digital computer program to execute the simulation of the rocket’s
motion is shown in Figure 12-3. Figure 12-4 shows the Fortran listing of a very simple pro-
gram to apply this calculation.
12-3
Initialization:
specify initial values of
x, y, v x , v y
Calculate time derivatives,

dx , dy , dvx , dvy
dt dt dt dt
by substituting values of x, y, vx, vy into ((1.1))
Calculate outputs and

print as required
Update values of x, y, vx, vy using (1.2):

dx
x = x+h
dt
dy
y = y+h
dt
dv
vx = vx + h x
dt
dvy
vy = vy + h
dt
Update t; t = t + h
Figure 12-3. Simple Simulation of Planetary Motion Modeled as in Figure 12-1
12-4
Program Application Guide: Volume II Basic Dynamic Simulation
REAL KG
C
C ALLOW USER TO ENTER INITIAL VALUES OF POSITION AND VELOCITY
C
WRITE(1,998)
998 FORMAT(1X,’ENTER X Y VX VY KG H’)
READ(1,*)X,Y,VX,VY,KG,H
C
C INITIALIZE VALUE OF TIME
C
TIME=0.
C
C START OF TIME ADVANCEMENT CALCULATION
C
1 CONTINUE
C
C CALCULATION OF VALUES OF ALL TIME DERIVATIVES
C
DVXDT=-KG*X/((X**2+Y**2)**1.5)
DVYDT=-KG*Y/((X**2+Y**2)**1.5)
DXDT=VX Process model
DYDT=VY
C
C OUTPUT PROCESSING. CALCULATE VALUES OF ANY REQUIRED
C VARIABLES AND PRINT RESULTS. HERE WE CALCULATE RADIAL
C DISTANCE AND ANGLE OF RADIUS ABOVE HORIZONTAL AXIS
C
R=SQRT(X**2+Y**2)
TH=ATAN2(Y,X)*57.2958
WRITE(1,999)TIME,R,TH,VX,VY Auxiliary output
999 FORMAT(1X,5E14.4) calculations
C
C CALCULATION OF NEW VALUES OF ALL VARIABLES
C
VX=VX+DVXDT*H
VY=VY+DVYDT*H Numerical
X=X+DXDT*H integration
Y=Y+DYDT*H calculation
C
C INCREMENT TIME
C
TIME=TIME+H
C
C RECYCLE THE PROCESS AS LONG AS IT IS OF INTEREST
C
IF(TIME.LT.1E6) GO TO 1
STOP
END
Figure 12-4. Simple Program to Simulate Motion of Rocket About a Heavenly Body
12-5
This example, and particularly the computer code in Figure 12-4, shows the three separate
major elements of the dynamic simulation calculation:
• Calculation of time derivatives.
• Output processing.
• Numerical integration calculations.
It is apparent that no intelligence on the process being simulated other than numerical val-
ues of the four time derivatives is needed to apply the numerical integration calculations.
If variables were renamed as follows, the program shown in Figure 12-4 could be written as
shown in Figure 12-5:
vx = s(1)
vy = s(2)
x = s(3)
y = s(4)
r = v(1)
 = v(2)
The program shown in Figure 12-5 consists of two sections:

• A main program to handle data input, output, and numerical integration without specific
information on the process being simulated
• A subroutine, ROCKET, to calculate numerical values of time derivatives from the pro-
cess differential equations, and to calculate numerical values of auxiliary output
variables
This example shows that the modeling of a specific process can be accomplished in a sub-
routine that is independent of the main dynamic simulation subroutine.
12-6
Program Application Guide: Volume II Simulation Model Library
REAL KG
REAL S(4),DS(4),V(2)
C
C ALLOW USER TO ENTER INITIAL VALUES OF POSITION AND VELOCITY
C
WRITE(1,998)
998 FORMAT(1X,’ENTER X Y VX VY KG H’)
READ(1,*)X,Y,VX,VY,KG,H
C
C INITIALIZE VALUE OF ALL VARIABLES AND TIME
C
TIME=0.
S(1)=VX Initialization
S(2)=VY
S(3)=X
S(4)=Y
C
C START OF TIME ADVANCEMENT CALCULATION Dynamic
C
1 CONTINUE Model
C Subroutine
C CALCULATION OF VALUES OF ALL TIME DERIVATIVES
C
CALL ROCKET(KG,S,DS,V) Call to process model
C
C WRITE OUT OUTPUT VARIABLES
C
WRITE(1,999)TIME,V(1),V(2) Output
999 FORMAT(1X,5E14,4)
C
C CALCULATION OF NEW VALUES OF ALL VARIABLES
C
DO 2 I=1,4 Numerical integration
2 S(I)=S(I)+DS(I)*H calculation
C
C INCREMENT TIME
C
TIME=TIME+H
C
C RECYCLE THE PROCESS AS LONG AS IT IS OF INTEREST
C
IF(TIME.LT.1E6) GO TO 1
STOP
END
SUBROUTINE ROCKET(KG,S,DS,V)
REAL KG
REAL S(4),DS(4),V(2)
C
C CALCULATE TIME DERIVATIVES
C Process
DS(1)=-KG*S(3)/((S(3)**2+S(4)**2)**1.5)
DS(2)=-KG*S(4)/((S(3)**2+S(4)**2)**1.5) Model
DS(3)=S(1) Subroutine
DS(4)=S(2)
C
C CALCULATE AUXILIARY OUTPUT VARIABLES
C V(1) IS RADIAL DISTANCE
C V(2) IS ANGLE ABOVE HORIZONTAL AXIS
C
V(1)=SQRT(S(3)**2+S(4)**2)
V(2)=ATAN2(S(4),S(3))*57.2958
RETURN
END
Figure 12-5. Rewritten Dynamic Simulation Program

with Process Model Isolated in Subroutine
12.2 Simulation Model Library

The dynamic simulation section of PSS®E has the basic structure illustrated by the forego-
ing example and shown in Figure 12-6. The main skeleton of PSS®E contains logic for data
12-7
Numerical Integration Stability Program Application Guide: Volume II
input, output, numerical integration, and electric network solution, but contains no logic
related to differential equations of specific equipment. The differential equation intelli-
gence on the equipment to be simulated is contained in a library of subroutines; each sub-
routine contains logic to calculate time derivatives for one specific type of equipment. The
model subroutines are called whenever the main skeleton logic needs numerical values of
time derivatives. Most models are called directly by PSS®E. Others that may require user
input or are called infrequently, such as switched shunt models, are called via linking sub-
routines CONEC and CONET.
Figure 12-6. Basic Program Structure of PSS®E Dynamic Simulation Section
Each new dynamic simulation requires that new linkage subroutines, CONEC and
CONET, be constructed and incorporated into the skeleton code. The construction and
loading on CONEC and CONET is automated by the PSS®E skeleton, which is able to
bootstrap itself into fully loaded condition. This process is covered in Section 13.2.6.
12.3 Numerical Integration Stability

A numerical integration process such as the one described by equations (1.2) will function
satisfactorily only if the time increment, t, is sufficiently small in relation to the time
constants and natural frequencies of the process being simulated. Too long a step width,
t, will lead to both inaccurate results and eventually to numerical instability of the calcu-
lation. To illustrate the phenomenon of numerical instability, consider the application of
the simple first-order integration formula:
dq
qnew = qold + t
dt (1.3)
to the very simple differential equation:
12-8
Program Application Guide: Volume II Numerical Integration Stability
dq -q
=
dt T (1.4)
Equation (1.4) describes the discharge of a capacitor through a resistor; its exact solution
is:
q(t) = q(o)e-t/T (1.5)
The numerical integration process of Figure 12-2 is equivalent to rewriting (1.4) as:
dq - qold
=
dt T (1.6)
and substituting this into (1.3), with the result:
(1.7)
which shows that:

• The sequence of q values will decrease monotonically if t is less than the time con-
stant, T.
• The sequence of q values will have alternating sign and increasing magnitude if t is
greater than 2 T.
It is apparent, then, that this numerical integration process can be accurate only when t is
small in relation to the time constant, T, and that it can exhibit an error that grows unstably
if t is excessive in relation to the process time constant. This behavior is shown graphi-
cally in Figure 12-7.
12-9
Numerical Integration Stability Program Application Guide: Volume II
Exact
0.2
0.5 1.5
0.75 2.1
Figure 12-7. Comparison of Numerical Integrations of Equation (1.4)

for Varying Values of t/T
PSS®E does not use the simple first-order integration scheme analyzed above; it uses a
second-order Euler numerical integration algorithm where accuracy is better than that of
the simple first-order scheme. Long experience with this method indicates that numerical
instability problems will normally be avoided, and accuracy will be adequate, if the inte-
gration time step, t, is kept smaller than about 1/5 to 1/4 of the shortest time constant in
the process being simulated.
Numerical instability of the integration process usually results in a growing sequence of
values in at least one system variable, with the sign alternating each time step. Numerical
instability is most readily detected, therefore, by examining a plot of the variable at each
time step. Conversely, a numerical instability can be present but may not be detected if the
12-10
Program Application Guide: Volume II Classification of Variables
affected variable is plotted only every n time steps, where n is an even number. Hence, it is
advisable always to use an odd number of time steps between each plotting of results from
a dynamic simulation.
12.4 Classification of Variables
12.4.1 General Classification

The dynamic simulation example in the previous section was particularly simple in that
each of the expressions for a time derivative (as shown in Figure 12-1) involved only the
quantities where derivatives were being evaluated. Because the values of all of these quan-
tities are known when the derivatives are to be evaluated, the derivatives can be calculated
directly by numerical evaluation of the right-hand sides of the differential equations. This
convenient state of affairs does not exist in power system dynamic simulations. The realis-
tic situation is illustrated by the following example: a simulation of the motion of the rotor
of a single generator connected to an infinite bus in accordance with the per unit differen-
tial equations shown in Figure 12-8.
i gen
Vb
jX R + jX
Internal Voltage = E  e = 1 + j0
Inertia = H jB
Damping = D
dn
2H ------ = P m – P e – Dn
dt
d
------ =  o n
dt
Figure 12-8. Differential Equations for Generator Connected to Infinite Bus
The differential equations in Figure 12-8 involve the constants H, D, and o and, in addi-
tion, the first involves the generator electrical power, Pe. In the first equation in Figure 12-
8, the assumption is that the variation in speed is small. In view of this, power (instead of
torque) is used in this equation. The evaluation of Pe at the instant for which the variables
n, speed deviation, and , rotor angle, are given requires solution of the electric network
for the generator current, igen. While it would be possible to write out an algebraic solution
for the current and, subsequently, the power, it is more practical to handle the network
solution and power calculation numerically. The evaluation of the time derivatives is then
handled by the following sequence of calculations:
1. Calculate vb from:
12-11
Classification of Variables Program Application Guide: Volume II
Calculate igen from:

igen = ( E  - vb ) / jx
2. Calculate Pgen from:
Pgen = Real (E  × igen)

3. Calculate time derivatives:
dn
= (Pm - Pgen - Dn) / 2H
dt
d
= on
dt
The calculation of the time derivatives in this example involved the prior calculation of a
set of quantities, vb, igen, and Pgen that are related algebraically to the variables, n and 
for which values are given by the numerical integration process. Variables such as vb, igen,
and Pgen are needed in the majority of dynamic simulations. It is convenient, therefore, to
classify the quantities involved in a dynamic simulation as:
• Constants: Parameters that do not vary during the period to be simulated.
• State Variables: Variables for which instantaneous values are determined by
differential equations.
• Algebraic Variables: Variables for which values can be determined if the values of all
state variables, constants, and input variables are given.
• Input Variables: Quantities for which values are specified at any instant by logic out-
side the dynamic simulation.
12.4.2 PSS®E Parameter and Variable Classification

PSS®E is a general-purpose dynamic simulation tool that can handle any simulation task
where formulation can be reduced to terms of constants, state variables, algebraic vari-
ables, and input variables as outlined in Section 12.4.1. Because it is designed for power
system work, PSS®E, has assigned named memory arrays to accommodate several spe-
cific groups of variables: constants, parameters, and variables, which are listed in Table 1.
This table lists four general-purpose arrays, CON, ICON, STATE, and VAR, together with
the special arrays such as MBASE, VOLT, and VREF. The only distinction between the
general-purpose and special-purpose arrays is that the latter accommodate quantities that
are frequently wanted as input and output quantities in setting up and displaying the
results of dynamic simulations. All arrays are used as constants, state variables, or alge-
braic variables in the same way.
12-12
Table 1: Dynamic Simulation Arrays
Array Contents Indexed By

Constants
CON General constants (real) CON number
ICON General constants (integer) ICON number
CHRICN General constants (character) ICON number
MBASE Machine base MVA Machine index
ZSORCE Machine impedance (complex) Machine index
XTRAN Step-up transformer impedance (complex) Machine index
GENTAP Step-up transformer tap ratio Machine index
State Variables
STATE General state variable array STATE number
Algebraic Variables
VAR General algebraic variable array VAR number
VOLT Bus per unit voltages (complex) Bus sequence number
BSFREQ Bus per unit frequency deviations Bus sequence number
ANGLE Machine relative rotor angle (degrees) Machine index
PELEC Machine electrical power (pu on SBASE) Machine index
QELEC Machine reactive power (pu on SBASE) Machine index
ETERM Machine terminal voltage (pu) Machine index
EFD Generator main field voltage (pu) Machine index
PMECH Turbine mechanical power (pu on MBASE) Machine index
SPEED Machine speed deviation from nominal (pu) Machine index
XADIFD Machine field current (pu) Machine index
ECOMP Voltage regulator compensated voltage (pu) Machine index
VOTHSG Stabilizer signal (pu) Machine index
VUEL Minimum excitation limiter signal (pu) Machine index
VOEL Maximum excitation limiter signal (pu) Machine index
Input Variables
VREF Voltage regulator voltage setpoint (pu) Machine index
Internal Arrays
DSTATE General state variable time derivatives STATE number
STORE General state variable integrator memory STATE number
BSFMEM Memory for frequency calculation Bus sequence number
12-13
Table 1: Dynamic Simulation Arrays (Cont.)
Array Contents Indexed By

STRTIN Starting array indices for plant-related models Array allocation index table
NUMTRM Pointer to bus sequence number Machine index
NUMBUS External bus number Bus sequence number
MACHID Machine identifier Machine index
INTICN Integer memory array ICON number
12.4.3 Data Space Allocation

The arrays summarized in Table 1 are used by the dynamic simulation models to accom-
modate all input and output signals, constant parameters, and internal variables. The spe-
cial-purpose arrays are indexed by either machine number or bus sequence number.
Locations in the general-purpose arrays, CON, VAR, STATE, ICON, and DSTATE are
allocated to equipment models on a first-come, first-served basis. The use of these arrays
is illustrated by Figure 12-9.
DSTATE(L) PMECH(I) PELEC(I) STATE(L)
dn 1
= (P - Pelec - Dn)
dt 2H mech
CON(J) CON(J+1)
DSTATE(L+1) STATE(L)
d
=  on
dt
SYSTEM PARAMETER
DSTATE(L) = (PMECH(I) - PELEC(I) - CON(J+1)  STATE(L)) / (2.  CON(J))

DSTATE(L+1) = OMEGAf  STATE(L)
SPEED(I) = STATE(L)
ANGLE(I) = STATE(L+1)
Figure 12-9. Use of PSS®E Array Elements to Accommodate System

Parameters and Variables
12-14
Figure 12-9 shows the code implementing the two rotor dynamics differential equations
shown in Figure 12-8 within a subroutine named MODEL. The statement defining the
model subroutine would be:
SUBROUTINE MODEL (I, J, L)
The values of J and L would be assigned to the next available locations in arrays CON and
STATE as the MODEL subroutine calls are entered into subroutine CONEC. The argu-
ments I, J, L are, then, pointers to the start of the data for the item of equipment being
modeled in the PSS®E data arrays.
The use of the data arrays by a CONEC subroutine is indicated by Figure 12-10.
Construction of subroutines CONEC and CONET require counting-off of the CONs,
VARs, etc. used each time a model is called. The numbers of array elements used by each
model are shown on the model data sheets. This process and the construction of CONEC
and CONET can be handled manually, or can be automated by activities DYRE and CHAN,
as shown in Sections 13.2.6 and 13.2.7, respectively.
12-15
CALL MODEL (I b ,J b ,L b )
CALL MODEL (I a ,J a ,L a )
Figure 12-10. Assignment of PSS®E Array Space to Equipment Model Subroutines Called from Subroutine CONEC
SUBROUTINE CONEC
RETURN
END
L b +1
L a +1
Lb
La
DSTATE
STATE
J b +1
J a +1
Jb
Ja
CON
Ib
Ia
ANGLE
SPEED
PELEC
PMECH
12-16
Program Application Guide: Volume II PSS®E Dynamic Simulation Sequence
12.5 PSS®E Dynamic Simulation Sequence
12.5.1 Overall Flow Chart

The overall flow of the dynamic simulation in PSS®E is shown in Figure 12-11. The
dynamic simulation functions are handled by activities DYRE, RSTR, STRT, RUN, and
ALTR. These subroutines include logic to accept constant and parameter values, solve the
electric network for its bus voltages, implement the numerical integration calculation, and
display results. They do not, however, include logic relating to the algebraic and differen-
tial equations of any item of power system equipment. The logic required to inform
PSS®E of the differential equations and other characteristics of the power system is pro-
vided by two libraries of subroutines:
• A library of subroutines handling models that involve state variables and differential
equations.
• A library of subroutines handling models that involve only algebraic relationships
between input and output signals, without reference to differential equations.
The linkage of the library subroutines into PSS®E is accomplished by four subroutines
called TBLCNC, TBLCNT, CONEC, and CONET, which have the responsibilities out-
lined below. TBLCNC and TBLCNT are supplied by PSS®E and are never seen by the
user.
12-17
PSS®E Dynamic Simulation Sequence Program Application Guide: Volume II
DYRE
Data Assimilation RSTR
Power Flow Activities
Initialization STRT-CONET-CONEC
Solution of Network for

Machine Currents, RUN-CONET
Bus Voltages
Output of Instantaneous
System Conditions RUN
Data Changes and

ALTR
Input Signals
Calculation of Rate
of Change of RUN-CONEC
All State Variables
Numerical Integration RUN
Figure 12-11. Dynamic Simulation Basic Flow
12.5.2 Subroutines TBLCNC and CONEC

Subroutines TBLCNC and CONEC are responsible for equipment models involving state
variables and differential equations. TBLCNC is responsible for machines and their con-
trol systems; CONEC is responsible for all other models. Both have the responsibility of
calculating the time derivative of every state variable used in modeling equipment, given
the present values of all state variables and of all generator stator currents. In addition,
both have the responsibility of calculating the values of all algebraic variables needed in
the course of obtaining numerical values of the state variable time derivatives.
Subroutine CONEC could be written in Fortran by the user of PSS®E, much in the style of
the modeling subroutine, ROCKET, shown in Figure 12-5. This is rarely done, however.
12.5.3 Subroutines TBLCNT and CONET

Subroutines TBLCNT and CONET are responsible for equipment models in which there
is a purely algebraic relationship between the voltage at a bus and the current drawn by or
key variable seen by a device. The principal equipment modeled in CONET are shunt load
devices (such as reactors), relays, and meters.
12-18
As with subroutine CONEC, the user will rarely write the modeling code directly into
CONET, but will normally establish a CONET subroutine that calls specialized equipment
modeling subroutines from the PSS®E model subroutine library.
12.5.4 Program Control

The role of subroutines CONEC and CONET is shown by Figure 12-12. CONEC and
CONET are called at several points in the simulation process by several different activi-
ties. The specific actions of CONEC, CONET, and the model subroutines that they call are
directed by a set of flags as defined in Table 2.
12-19
Time = –2 time steps

Tpause = –2 time steps
Kpause = 0
All STATE and DSTATE = 0
MODE = 4
Call CONEC for induction motor and dc lines;
set maximum number of integrators
MODE = 1 Subroutine
Call SITER for initial network solution; CONET
calls CONET to calculate current injections
Call CONEC to initialize models; output

initial machine conditions
MODE = 3
Set Pmech, Efd, VOTHSG, VUEL, VOEL,
machine source currents (ISOURCE)
Network solution; calls CONET to calculate Subroutine

current injections CONET
MODE = 2
Call CONEC to calculate all state variable time
derivatives and set VOTHSG, VUEL, and VOEL
Loop through all state variables and check

for initial condition suspect.
Criterion:
If ABS(STATE).GE.0.001
A = ABS(DSTATE)/ABS(STATE)
ELSE
A = ABS(DSTATE)
Then
Initial condition suspect if
A.GT.0.01
a. CONEC and CONET Calls on Initialization
Figure 12-12. Dynamic Simulation

(Sheet 1 of 2)
12-20
Read TPAUSE, NPRT, NPLT, CRTPLT
MODE = 3
CALL CONEC to calculate generator source currents,
exciter field voltages, turbine mechanical powers,
stabilizer outputs, minimum excitation limiter outputs,
and maximum excitation limiter outputs
CALL SITER and CONET to solve network
MODE = 2
CALL CONEC to calculate all state variable time
derivatives and set stabilizer, minimum excitation
limiter, and maximum excitation limiter outputs
Output channel values
KPAUSE = 0
Numerical integration;
update state variables
TIME = TIME + STEP
No
TIME  TPAUSE?
KPAUSE = 1
MODE = 3
CALL CONEC
CALL SITER
MODE = 2
CALL CONEC
Output
KPAUSE = 2
Link to activity, select to Snapshot, make

changes, etc.
b. CONEC and CONET Calls During Activity RUN
Figure 0-0 (Cont). Dynamic Simulation (Sheet 2 of 2)
12-21
Table 2: PSS®E Dynamic Simulation Control Flags
Flag Name Value Significance

Initialization: calculate initial condition values of all state variables and
1 algebraic variables on the assumption that the system is in the steady
state.
The model must make all computations needed to place time derivatives
into the DSTATE array; each stabilizer model must compute the present
value of its output signal and place it in the appropriate entry in the
VOTHSG array; each minimum excitation limiter model must compute the
2
present value of its output signal and place it in the appropriate entry in the
VUEL array; each maximum excitation limiter model must compute the
present value of its output signal and place it in the appropriate entry in the
VOEL array.
Each stabilizer model must compute the present value of its output signal
and place it in the appropriate entry in the VOTHSG array; each minimum
MODE excitation limiter model must compute the present value of its output signal
and place it in the appropriate entry in the VUEL array; each maximum
3 excitation limiter model must compute the present value of its output signal
and place it in the appropriate entry in the VOEL array; a turbine-governor
model must compute the present value of turbine mechanical power and
place it in the PMECH array; other models written by users normally have
no requirements in MODE three.
Apply special calculations in initialization of induction motor and dc trans-
4
mission models.
5 Write model documentation when activity DOCU is run.
6 Write model input data when activity DYDA is run.
Write model documentation when activity DOCU is run in its data checking
7
mode.
Network iterations in progress; current injection models should update
False
IFLAG their contribution.
True Network solution is completed.
1 Time = t–: preswitching solution.
KPAUSE 2 Time = t+: switching has been made, new solution needed.
0 Time = t: routine step, no pause at time, t.
0 Standard simulation is being run.
1 Excitation system response ratio test is being run.
2 Excitation system open-circuit test is being run.
MSTATE 3 Governor test is being run.
4 Extended term dynamic simulation is being run.
Dynamics data is present but an initialization activity has not been suc-
5
cessfully run.
12-22
12.5.5 Simulation Output Channels

Any power system variable that can be represented by an element of an array appearing in
Table 1 may be selected as an output of a PSS®E dynamic simulation run. The handling of
dynamic simulation output is shown in Figure 12-13. PSS®E places a set of output channel
values in a Channel Output File at specified intervals of simulation time, building up a file
containing a comprehensive history of the event being simulated.
PSS®E Working Channel

PSS®E Output
Arrays per Selection
Channel Array
Table 1 Logic
Channel
Output
File
PSS®E Channel
Specification Arrays
per Table 3
Plotting PSSPLT
Device Plotting
Program
Figure 12-13. PSS®E Output Channel Control
The Channel Output File is a binary variable length record file for which the name is spec-
ified by the user when it is created in activity STRT. The PSS®E simulation activities do
not plot or tabulate the output channels; these functions are handled by program PSSPLT.
The output channel values are obtained by copying designated variables from the dynamic
simulation arrays into the output channel array, CHAN. This copying process is done
every NPLT time steps, where NPLT is an integer specified by the user in the dialog of
activity RUN. The selection of simulation variables for output via channels is controlled by
the arrays summarized in Table 3. Array IPRINT contains the addresses of the output vari-
ables in the main dynamic simulation array space of PSS®E. Each output channel is given
a 32-character alphanumeric identification. These are contained in the array, IDENT.
12-23
Overview of Simulation Procedure Program Application Guide: Volume II
Table 3: Channel Specification Arrays
Array Name Size Function

Address of quantity to be displayed, relative to base of PSS®E array
IPRINT n
memory.
CHAN n Value of output channel.
IDENT n 32-character channel identifier.
Numbers of main channels in subset to be plotted during simulation
CRTCHN 6
on CRT.
CMAX 6 Upper limit for progress plot scaling.
CMIN 6 Lower limit for progress plot scaling.
PSS®E allows a subset of six of the output channels to be displayed by a progress output
graph while activity RUN is being run. These six channels and their upper and lower scal-
ing limits are specified by arrays CRTCHN, CMAX, and CMIN. The channel control
arrays are established by the user through a dialog with activity CHAN. Any channel con-
trol parameter may be changed prior to a simulation run, via activity ALTR.
12.6 Overview of Simulation Procedure
12.6.1 Model Setup and Use

A dynamic simulation study involves two general phases:
1. Setup of a simulation model consisting of the following:
a. Valid CONEC and CONET subroutines linking the required equipment dynamic
models into PSS®E.
b. Valid parameters and operating conditions for all items of equipment.
2. Execution of simulation runs using the model as setup in the first phase to show the
effects of proposed events such as short circuit faults, generator trips, or motor starting.
PSS®E is used in both the setup phase and in the execution of simulation runs. The model
setup phase is the more critical and involves the greater amount of work. A valid model
setup with well-planned CONEC and CONET subroutines allows extensive series of sim-
ulation runs to be made with relatively little incremental work. A less-than-ideal model
setup with unchecked data will not only make simulation runs more difficult, it will usu-
ally result in inaccurate results.
12.6.2 Simulation Model Setup

The principal steps in setting up a PSS®E dynamic simulation model are summarized in
Table 4 and detailed in later sections. The mechanics of setting up a system model are han-
dled by PSS®E itself, including the construction of the CONEC and CONET subroutines
12-24
Program Application Guide: Volume II Overview of Simulation Procedure
and the establishment of output channel control tables. The bulk of the system model setup
is automated, with handwork being reserved for special situations. A user with special
requirements may, however, refine the CONEC and CONET subroutines to implement
logic that is not provided by the standard model library subroutines.
12-25
Table 4: Overview of Steps in PSS®E Dynamic Simulation Model Setup
PSS®E
Step Files Used
Activities Used
1. Allocate a set of files for the simulation setup. — —
2. Establish base case power flow with generators Create a converted
See Sections 5.2
and load models converted to their dynamic power flow Saved
and 13.3
forms. Case File.
3. Select the appropriate model for each item of
— —
equipment.
4. Create a PSS®E Dynamics Data File. — Dynamics Data File.
5. Use PSS®E in setup mode to create CONEC LOFL, CASE,
Read Dynamics Data
and CONET subroutines. ORDR, FACT,
File.
RTRN, DYRE
6. Use PSS®E in setup mode to create additional
CONET model references to monitor transmis- CHAN —
sion network via output channels.
7. Make a Snapshot recording the parameter Initial condition Snap-
SNAP
values and channel assignments. shot File.
8. Stop PSS®E. — —
9. Edit manual refinements into CONEC and Use text editor to alter
CONET as required. CONEC and CONET
—
Fortran source code
files.
10. Compile CONEC and CONET and link into CONEC and CONET
—
PSS®E by executing CLOAD4 command file. relocatable code files.
11. Start PSS®E and recover the initial condition Initial condition Snap-
RSTR
Snapshot File. shot File.
12. Retrieve power flow case corresponding to initial LOFL, CASE, Converted power flow
condition Snapshot File. FACT, RTRN Saved Case File.
13. Enter any additional data values needed as a
ALTR —
result to refinements to CONEC or CONET.
14. Make an updated Snapshot. SNAP —
15. Set all VAR and STATE variables to their initial Dynamics Channel
STRT
condition values. Output File.
16. Check out all reported initial condition errors. — —
17. Make an updated initial condition Snapshot. Initial condition Snap-
SNAP
shot File.
18. List all dynamic simulation data. DOCU, DLST —
19. Stop PSS®E. — —
20. Review all data for correctness. — —
12-26
Program Application Guide: Volume II Overview of Simulation Procedure
Also, users may make their own special refinements to CONEC and CONET, because
well-planned control logic in these subroutines often gives great simplifications in the
execution of simulation runs. Examples of user-tailored logic in CONEC and CONET are
given in Section 13.5.
12.6.3 Execution of Simulation Runs

The steps involved in executing dynamic simulation runs vary depending on the details of
the event being simulated. A sequence representative of a basic short circuit fault simula-
tion is summarized in Table 5.
Table 5: Representative Steps in a Dynamic Simulation Run
PSS®E
Step Files Used
Activities Used
1. Recover initial condition Snapshot. Initial condition Snap-
RSTR
shot File.
2. Recover initial condition converted power flow LOFL, CASE, Initial condition power
case. FACT, RTRN flow Saved Case File.
3. Establish and confirm initial conditions in
d y n a m i c m o d e l s ( t i m e i n i t i a l i z e d t o STRT —
t = -2 × DELT.
4. Select and initialize dynamics output file. STRT Channel Output File.
5. Advance simulation with no disturbance to t =
0–, or some other appropriate time for first dis- RUN Channel Output File.
turbing action.
6. Apply first action of disturbance, e.g.,
ALTR —
application of fault.
7. Advance simulation to time of next change of
RUN Channel Output File.
applied disturbance.
8. Apply next change of applied disturbance, e.g.,
ALTR —
clear fault by opening a line.
9. Advance simulation to time at which no addi-
RUN Channel Output File.
tional results are currently needed.
10. Make a new Snapshot of conditions at the final
SNAP New Snapshot File.
instant of the run.
11. Make a new power flow Saved Case recording
New power flow Saved
the network condition at the final instant of the LOFL, SAVE
Case File.
run.
12. Stop PSS®E and plot, tabulate, or otherwise
STOP —
analyze the results.
12-27
12.6.4 Preservation of Simulation Setup – Snapshots

A complete dynamic simulation setup may be preserved at any time by making a Snap-
shot, which is a copy in a disk file of the current contents of dynamic simulation working
memory. A Snapshot File contains a complete copy of the dynamic simulation arrays
noted in Table 1 and of all other arrays used internally by the dynamic simulation of
PSS®E.
A Snapshot may be used to preserve a dynamic simulation setup at any point from the
early stages of data entry, through initialization, to any point during the integration of the
differential equations. When made during the integration process, the Snapshot gives a
complete statement of the instantaneous condition of all equipment where modeling is
accommodated in the arrays listed in Table 1.
A Snapshot contains instantaneous values of bus voltages but does not contain transmis-
sion network impedance, status, load, and related data. This data must be preserved in
power flow Saved Cases. The user may either make a new power flow Saved Case File to
accompany each Snapshot or may rely on a single power flow Saved Case File that is used
as the basis for all network data throughout a dynamic simulation (see Sections 13.4.2.4
and 13.4.2.5).
Snapshots and associated power flow Saved Cases allow dynamic simulation work to be
frozen at any point and to be restarted at the convenience of the user. They also allow in-
progress conditions at key points in a simulation run to be preserved so that the effects of
alternative switchings and perturbations can be investigated without having to rerun the
simulation up to the point of the switching.
12-28
Chapter 13
Dynamic Simulation Setup
Chapter 13 - Dynamic Simulation Setup
13.1 Overview
The dynamic simulation portion of the PSS®E program requires following a specific pro-
cedure to set up the system model. Chapter 12 covered the theory behind dynamic simula-
tion and provided an outline for setting up a system. This section covers a specific
procedure for taking a sample system, doing a study, expanding the modeling of the sys-
tem, and doing another study. The description of specific modeling equipment is covered
in Chapter 14.
13.2 Basic Setup Example
13.2.1 Files
The simulation setup procedure is outlined in Table 4. Table 2 shows the files to be used in
this example. Files will be needed for raw data records, power flow Saved Cases, and
dynamics Snapshots, as well as for the CONEC, CONET subroutines.
Table 1: Files to be Used in First Dynamic Simulation Setup Example
Power Flow Raw Data SMALL

Solved Initial Condition (IC) Power Flow SME1
Solved Converted IC Power Flow SME2
Dynamics Raw Data SMEDD
Initial Condition Snapshot SSN1
First Plotting Output File GOP1
Subroutine CONEC Source CC1
Subroutine CONET Source CT1
Response File for Channel Assignment RC1
Progress Snapshot SSN2
Progress Power Flow Case SLF2
13-1
Dynamic Simulation Setup PSS®E 33.4
Basic Setup Example Program Application Guide: Volume II
13.2.2 Example System

The setup of a dynamic simulation is illustrated in Figure 13-1 using a simple example in
which a small system is modeled for stability studies.
3 × 5 MVA 1 × 10 MVA
Hydro Unit Steam Unit
kV
0.9 pf
kV
0.85 pf
SMALL EXAMPLE SYSTEM
BOONDOCKS POWER AND LIGHT
SMALL SYSTEM EXAMPLE WED FEB 23, 1994 09:28
Figure 13-1. Small Example System for Dynamic Simulation Setup Examples
13.2.3 Power Flow Model

Every dynamic simulation is based on a power flow Saved Case that provides the required
transmission network data, load data, and generator positive sequence model. The power
flow Saved Case must have its generator models in the dynamic simulation form as out-
lined in Section 5.6.2, and its loads represented by a composite of the characteristics from
Section 5.5.1 that is acceptable during excursions of bus voltages.
Figure 13-2 shows the basic power flow raw data file, SMALL, used to set up the power
flow case.
13-2
PSS®E 33.4 Dynamic Simulation Setup
Program Application Guide: Volume II Basic Setup Example
0 100
SMALL EXAMPLE SYSTEM File SMALL
100 3 0 0 0 0 1 1.05 0 'HYDRO' 33
150 1 5 1.56 0 0 2 1 0 'LOAD' 33
151 1 15 7.9 0 0 2 1 0 'LOAD' 3.3 Bus Data
200 1 0 0 0 0 1 1 0 'STEAM' 33
201 2 0 0 0 0 1 1 0 'STEAM' 3.3
0
100 1 4 0 2 0 1.05 0 5 0 .25 0 .09 1.025 Generator Data
100 2 4 0 2 0 1.05 0 5 0 .25 0 .09 1.025
100 3 4 0 2 0 1.05 0 5 0 .25 0 .09 1.025
201 1 8 0 4 0 1.05 200 10 0 .2
0
100 200 1 .02 .2 .03 15 17.5 20
100 150 1 .01 .1 .018 15 17.5 20 Branch Data
150 200 1 .01 .1 .018 15 17.5 20
200 201 1 0 .8 0 10 12 15 1.05
150 151 1 0 .3 0 15 20 25 1.025
0 Transformer Data
200 201 1 200 1.1 .9 1.1 .9
150 151 1 151 1.1 .9 1.1 .9
0
0
0 No Other Data
0
0
0
0
Figure 13-2. Input Data for Small Example System
The steps in constructing the power flow case for dynamics use are shown in Figure 13-3.
The power flow case is first solved in the conventional manner to establish the initial con-
dition values of all bus voltages, generator reactive outputs, and phase angles. The initial
condition power flow, still in conventional form for power flow studies, is saved in file
SME1, which is not useful for dynamics work.
The conversion of the power flow case so that its models are valid in dynamics (t  t+)
rather than steady-state (t  t–) conditions is handled by activities CONL and CONG. These
are spotted in Figure 13-3. The dialog for CONL changes all loads from pure constant
(MW, Mvar) to a composite characteristic in which 60% of the real power load follows a
constant current characteristic and 40% behaves as a constant admittance. The split is des-
ignated as 50/50 in reactive power. This rearrangement of load characteristics may be
made on a bus-by-bus basis, with different splits at different buses, by using CONL with
its selection options, which are as outlined in Section 6.6.1.
13-3
Activity CONG should always be followed by activity ORDR because the change of gener-
ator modeling to dynamic form removes all swing (Type 3) buses from the system, hence
changing its topology with respect to optimal ordering.
The converted power flow case is saved in file SME2 for subsequent use as the source of
network data in dynamic simulations. All generators will, then, be represented in the
dynamic simulations by the dynamic impedance (normally the subtransient impedance) as
contained in the power flow case SME2, and all loads will be modeled with the voltage
dependence, as set up by CONL, unless this is specifically overridden in setting up the sim-
ulation.
Activities FACT and TYSL were used before the converted case was saved in file SME2.
These steps give a refinement of the power flow solution to obtain the smallest possible
mismatch at all buses. Such refinement is optional, but recommended, because:
1. The refinement’s initial condition estimate of all generator currents is as perfect as pos-
sible. Therefore, the generator internal flux linkages calculated in the dynamics
initialization will be, as nearly as possible, a perfect set of steady-state values.
2. Refining the power flow solution will reveal any significant imperfections. If
activity TYSL takes more than two or three iterations to reach its tolerance, it is likely
that the power flow case to which CONL and CONG were applied had not been brought
to a small enough mismatch to provide a good set of initial conditions for dynamics
work.
The converted saved case should always be reordered with ORDR after execution of
CONG and before it is saved; otherwise, the ordering will have to be done each time a
dynamic simulation is started.
Figure 13-4 shows a full data listing of the power flow case for reference purposes. This
listing describes the solved, but unconverted, initial condition power flow as saved in file
SME1.
13-4
$ psslf4
POWER TECHNOLOGIES INCORPORATED
12000 BUS POWER SYSTEM SIMULATOR--PSS®E-22.0
INITIATED AT LOAD FLOW ENTRY POINT ON MON FEB 14, 1994 15:11
ACTIVITY? read
ENTER INPUT FILE NAME (0 TO EXIT, 1 FOR TERMINAL): SMALL
ENTER IC, SBASE
ENTER TWO LINE HEADING
ENTER BUS DATA
ENTER GENERATOR DATA
ENTER BRANCH DATA
ENTER TRANSFORMER ADJUSTMENT DATA
ENTER AREA INTERCHANGE DATA
ENTER TWO-TERMINAL DC LINE DATA
ENTER SWITCHED SHUNT DATA
ENTER TRANSFORMER IMPEDANCE CORRECTION DATA
ENTER MULTI-TERMINAL DC LINE DATA
ENTER MULTI-SECTION LINE DATA
ENTER ZONE NAME DATA
ENTER AREA TRANSACTION DATA
BUILDING TABLES--WAIT...
ACTIVITY? SOLV
ITER DELTAV/TOL BUS REAL(DELTAV) IMAG(DELTAV)

1 984.908 201 0.2840E-01 0.9431E-01
2 594.432 201 0.4574E-02 -0.5927E-01
3 491.451 151 -0.4901E-01 0.3633E-02
4 253.457 201 0.1111E-02 -0.2532E-01
5 92.747 201 -0.3286E-02 0.8673E-02
6 52.375 151 0.3459E-02 -0.3933E-02
7 70.165 151 -0.5852E-02 0.3871E-02
8 41.591 151 0.3665E-02 -0.1965E-02
9 11.221 151 -0.8610E-03 0.7196E-03
10 10.485 201 -0.3681E-03 0.9818E-03
11 7.224 151 -0.6373E-03 0.3400E-03
12 5.429 151 0.5383E-03 -0.7123E-04
13 3.718 201 -0.2832E-03 -0.2409E-03
14 3.091 201 -0.2623E-03 0.1636E-03
15 2.446 201 -0.2378E-03 -0.5694E-04
16 2.384 201 -0.2375E-03 0.2116E-04
17 2.219 201 -0.2218E-03 -0.4107E-05
18 2.038 201 -0.2037E-03 0.7415E-05
19 1.954 201 -0.1953E-03 0.7045E-05
20 1.831 201 -0.1831E-03 -0.1460E-05
21 1.733 201 -0.1732E-03 0.6491E-05
22 1.619 201 -0.1619E-03 0.1621E-05
23 1.536 201 -0.1535E-03 0.4196E-05
24 1.442 201 -0.1442E-03 0.1949E-05
25 1.358 201 -0.1358E-03 0.2968E-05
26 1.278 201 -0.1278E-03 0.2533E-05
27 1.206 201 -0.1205E-03 0.2271E-05
28 1.135 201 -0.1135E-03 0.2319E-05
29 1.068 201 -0.1068E-03 0.1985E-05
30 1.009 201 -0.1009E-03 0.2015E-05
31 0.950 201 -0.9499E-04 0.1782E-05
REACHED TOLERANCE IN 31 ITERATIONS
Figure 13-3. Preparation of Power Flow Case for Use in Dynamic Simulation
(Sheet 1 of 2)
13-5
LARGEST MISMATCH: 0.00 MW 0.01 MVAR 0.01 MVA-BUS 150 [LOAD 33.0]
SYSTEM TOTAL ABSOLUTE MISMATCH: 0.02 MVA
ACTIVITY? ORDR
DIAGONALS = 4 OFF-DIAGONALS = 3 MAX SIZE = 6
ACTIVITY? FNSL
ENTER ITERATION NUMBER FOR VAR LIMITS

0 FOR IMMEDIATELY, -1 TO IGNORE COMPLETELY: 0
ITER DELTAP BUS DELTAQ BUS DELTA/V/ BUS DELTAANG BUS

0 0.0000( 200) 0.0002( 200)
0.00017( 200) 0.00003( 201)
1 0.0000( 200) 0.0018( 201)
0.00135( 201) 0.00008( 201)
2 0.0000( 200) 0.0001( 201)
LARGEST MISMATCH: 0.00 MW -0.01 MVAR 0.01 MVA-BUS 200 [STEAM 33.0]
ACTIVITY? SAVE SME1
CASE SAVED IN FILE SME1.SAV ON MON FEB 14, 1994 15:12

Save unconverted initial condition power flow
ACTIVITY? CONL ALL
ENTER % CONSTANT I, % CONSTANT G FOR REAL POWER: 100 0
ENTER % CONSTANT I, % CONSTANT B FOR REACTIVE POWER: 0 100
Convert load
LOAD TO BE REPRESENTED AS: characteristics to
REAL REACTIVE
0.00% 0.00% CONSTANT POWER form needed in
100.00% 0.00% CONSTANT CURRENT
0.00% 100.00% CONSTANT ADMITTANCE dynamic simulation
ENTER 1 IF O.K., 0 OTHERWISE:
LOADS CONVERTED AT 2 OF 2 LOAD BUSES
ACTIVITY? CONG
GENERATORS CONVERTED
ACTIVITY? ORDR Convert generators for dynamic simulation.

DIAGONALS = 5 OFF-DIAGONALS = 5 MAX SIZE = 10
ACTIVITY? FACT
5 DIAGONAL AND 5 OFF-DIAGONAL ELEMENTS

Reordering is essential
after converting
ACTIVITY? TYSL
generators
ITER DELTAV/TOL BUS REAL(DELTAV) IMAG(DELTAV)
1 2.161 200 0.2134E-04 -0.3395E-05
2 0.060 151 -0.5960E-06 0.6706E-07
LARGEST MISMATCH: 0.00 MW 0.00 MVAR 0.00 MVA-BUS 100 [HYDRO 33.0]
SYSTEM TOTAL ABSOLUTE MISMATCH: 0.00 MVA Refine solution with generator
ACTIVITY? SAVE SME2 flux linkage fixed
CASE SAVED IN FILE SME2.SAV ON MON FEB 14, 1994 17:05
Save converted case for use in dynamic simulation
Figure 0-0 (Cont). Preparation of Power Flow Case for Use in Dynamic Simulation (Sheet 2 of 2)
13-6
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E MON FEB 14, 1994 17:17
SMALL EXAMPLE SYSTEM SYSTEM SUMMARY
---------------------BUSES--------------------- GENERATION AREAS ZONES SYSTEM

TOTAL PQ<>0. PQ=0. PE/E PE/Q SWING OTHER PLANTS MACHS USED USED BASE
5 2 1 1 0 1 0 2 4 2 1 100.0
------------------AC BRANCHES------------------MULTI-SECTION DC LINES TRANS
TOTAL RXB RX RXT RX=0. IN OUT LINES SECTNS 2-TRM N-TRM ACTN
5 3 0 2 0 5 0 0 0 0 0 0
TOTAL GENERATION PQLOAD I LOAD Y LOAD SHUNTS CHARGING LOSSES SWING

MW 20.0 20.0 0.0 0.0 0.0 0.0 0.0 12.0
MVAR 3.8 9.5 0.0 0.0 0.0 7.3 1.6 1.5
TOTAL SYSTEM MISMATCH = 0.01 MVA X-----AT BUS-----X X------SWING-----X

LARGEST BUS MISMATCH = 0.01 MVA 200 STEAM 33.0 100 HYDRO 33.0
HIGH VOLTAGE = 1.0500 PU 100 HYDRO 33.0
LOW VOLTAGE = 0.9947 PU 151 LOAD 3.30 THRSHZ
0.000100
X-------SOLV AND MSLV-------X X---NEWTON---X X------TYSL------X BLOW PQ
ACCP ACCQ ACCM TOL ITER ACCN TOL ITER ACCTY TOL ITER UP BRAK
1.600 1.600 1.000 0.00010 100 1.00 0.100 20 1.000 0.000010 20 3.0 0.70
ENTER 0 TO EXIT, 1 FOR NEXT PAGE, 2 FOR NEXT DATA CATEGORY:
SMALL EXAMPLE SYSTEM WORST
BOONDOCKS POWER AND LIGHT MISMATCHES
BUS# NAME BSKV MW MVAR MVA
200 STEAM 33.0 0.00 0.01 0.01
201 STEAM 3.30 0.00 -0.01 0.01
ENTER 0 TO EXIT, 1 FOR NEXT PAGE:
SMALL EXAMPLE SYSTEM BUS DATA
BUS# NAME BSKV CODE VOLT ANGLE PLOAD QLOAD S H U N T AREA ZONE
100 HYDRO 33.0 3 1.0500 0.0 0.0 0.0 0.0 0.0 1 1
150 LOAD 33.0 1 1.0450 -0.6 5.0 1.6 0.0 0.0 2 1
151 LOAD 3.30 1 0.9947 -3.1 15.0 7.9 0.0 0.0 2 1
200 STEAM 33.0 1 1.0500 -0.1 0.0 0.0 0.0 0.0 1 1
201 STEAM 3.30 2 1.0168 3.5 0.0 0.0 0.0 0.0 1 1
SMALL EXAMPLE SYSTEM SWITCHED
BOONDOCKS POWER AND LIGHT SHUNT DATA
BUS# MOD VHI VLO SHUNT X-------X X-------X X-------X X-------X REMOTE
* NONE *
SMALL EXAMPLE SYSTEM GENERATING
BOONDOCKS POWER AND LIGHT PLANT DATA
BUS# NAME BSKV COD MCNS PGEN QGEN QMAX QMIN VSCHED VACT. REMOT PCT Q
100 HYDRO 33.0 3 3 12.0 1.5 6.0 0.0 1.0500 1.0500
201 STEAM 3.30 2 1 8.0 2.4 4.0 0.0 1.0500 1.0500 200 100.0
SMALL EXAMPLE SYSTEM GENERATOR
BOONDOCKS POWER AND LIGHT UNIT DATA
BUS# NAME BSKV COD ID ST PGEN QGEN QMAX QMIN PMAX PMIN MBASE Z S O R C E X T R A N GENTAP
100 HYDRO 33.0 3 1 1 4 0 2 0 9999-9999 5 0.0000 0.2500 0.0000 0.0900 1.0250
100 HYDRO 33.0 3 2 1 4 0 2 0 9999-9999 5 0.0000 0.2500 0.0000 0.0900 1.0250
100 HYDRO 33.0 3 3 1 4 0 2 0 9999-9999 5 0.0000 0.2500 0.0000 0.0900 1.0250
201 STEAM 3.30 2 1 1 8 2 4 0 9999-9999 10 0.0000 0.2000
SMALL EXAMPLE SYSTEM BRANCH DATA
FROM TO CKT NAME NAME LINE R LINE X CHRGING TP ST RATA RATB RATC
100* 150 1 HYDRO LOAD 0.0100 0.1000 0.0180 1 15 18 20
100* 200 1 HYDRO STEAM 0.0200 0.2000 0.0300 1 15 18 20
150* 151 1 LOAD LOAD 0.0000 0.3000 0.0000 F 1 15 20 25
150* 200 1 LOAD STEAM 0.0100 0.1000 0.0180 1 15 18 20
200* 201 1 STEAM STEAM 0.0000 0.8000 0.0000 F 1 10 12 15
SMALL EXAMPLE SYSTEM TRANSFORMER DATA
FROM TO CKT TP RATIO ANGLE RG CONT RMAX RMIN VMAX VMIN STEP TABLE CR CX
150 151 1 F 1.0250 0.00 1 151 1.1000 0.9000 1.1000 0.9000 0.00625
200 201 1 F 1.0500 0.00 1 -200 1.1000 0.9000 1.1000 0.9000 0.00625
Figure 13-4. Data for Small Example System

as Established in Power Flow Cases SME1 and SME2
13-7
13.2.4 Equipment Model Selection

A model must be selected from the library for each piece of equipment to be represented
in the simulation. One convenient way of recording the model selection is to fill out a data
form for each item of equipment, as shown in Figure 13-5, for the sample case. Each model
data sheet contains spaces for parameter and bus number values describing the equipment
and its location, and index values indicating the elements of the CON, VAR, STATE, and
ICON arrays assigned to the equipment model. Only the first category of entries is
required for a basic system model setup, and only these basic entries are completed in
Figure 13-5.
All data entered in dynamic simulation tables are expressed with respect to the base MVA
of the equipment described. The generator data sheets, in particular, require the user to
identify MVA. Figure 13-6, which pertains to a hydro plant with three units of 5 MVA each,
represents only one unit at bus 100. The constant data values, indicated by a star, are the
parameters of one 5-MVA generator expressed with respect to a 5-MVA base. The value
of ZSORCE is, similarly, the subtransient impedance of one 5-MVA generator expressed
with respect to a 5-MVA base. The fact that there are three identical units connected on the
bus is handled by setting the MBASE parameter to 3  5= 15 MVA. The behavior of three
identical 5-MVA units is the same, mathematically, as that of one 15-MVA unit where per
unit data on a 15-MVA base is the same as the per unit data of the 5-MVA units on 5-MVA
base. The value of Xd entered on GENSAL, GENROU, and GENDCO data sheets must
be identical to the imaginary part of ZSORCE, which is entered in the power flow.
13.2.5 PSS®E Dynamics Data File

The parameters chosen when the models were selected and entered on the data sheets,
must be entered in the Dynamics Data File, SMEDD, shown in Figure 13-7. (Refer to
Section 14.1, Reading Dynamics Model Data of the PSS®E Program Operation Manual for
details of the raw data file.) Each record contains equipment bus number, model name, and
the parameter list in the same order as it appears in the data sheet. The order in which
records appear in the Dynamics Data File is not important.
It may be desirable to include entries in the Dynamics Data File for items of equipment
that are not in service in the initial condition power flow case (SME2 in this example), so
that the file can be held as a complete record of system data for studies of a variety of sys-
tem conditions. Unneeded records will simply be ignored when the file is used by
activity DYRE.
13-8
GENROU
Round Rotor Generator Model (Quadratic Saturation)
This model is located at system bus # 201* IBUS,

machine # 1* I,
This model uses CONs starting with #____ J,
and STATEs starting with #____ K.
The machine MVA is 10* for each of 1* unit = 5*
MBASE.
ZSORCE for this machine is 0 + j 0.2 on the above
MBASE
*See power flow data, Figure 13-4.
CONs # Value Description STATEs # Description

J 6 T’do (>0) (sec) K E’q
J+1 0.05 Tdo (>0) (sec) K+1 E’d
J+2 1 T’qo (>0) (sec) K+2 kd
J+3 0.05 Tqo (>0) (sec) K+3 kq
J+4 3 Inertia, H K+4  speed (pu)
J+5 0 Speed damping, D K+5 Angle (radians)
J+6 1.4 Xd
J+7 1.35 Xq
J+8 0.3 X’d
J+9 0.6 X’q
J+10 0.2 Xd = Xq
J+11 0.1 Xl
J+12 0.03 S(1.0)
J+13 0.4 S(1.2)
Xd, Xq, X’d, X’q, Xd, Xq, Xl, H, and D are in

pu, machine MVA base.
Xq must be equal to Xd.
IBUS, ’GENROU’, I, T’do, T’do, Tqo, Tqo, H, D, Xd, Xq, X’d, X’q, Xd, Xl, S(1.0), S(1.2)/
Figure 13-5. Sample Dynamic Data Sheet

(Sheet 1 of 12)
13-9
GENSAL
Salient Pole Generator Model (Quadratic Saturation on d-Axis)

machine # 1* I.
Pm PMECH SPEED Speed
Efd EFD ISORCE
and STATEs starting with #____ K. Source Current
VOLT at GENSAL ETERM
The machine MVA is 5* for each of 1* units = 5* VT Terminal Voltage
MBASE. Terminal
Bus
ZSORCE for this machine is 0* + j 0.25* on the
above MBASE. ANGLE Angle

J+1 0.05 Tdo (>0) (sec) K+1 q
J+2 0.06 Tqo (>0) (sec) K+2 kd
J+3 5.084 Inertia, H K+3  speed (pu)

K+4 Angle (radians)
J+4 1 Speed damping, D
J+5 1.5 Xd
J+6 1.2 Xq
J+7 0.4 X’d
J+8 0.25 Xd = Xq
J+9 0.12 Xl
J+10 0.03 S(1.0)
J+11 0.25 S(1.2)
Xd, Xq, X’d, Xd, Xq, Xl, H, and D are in pu,

machine MVA base.
IBUS, ’GENSAL’, I, T’do, Tdo, Tqo, H, D, Xd, Xq, X’d, Xd, Xl, S(1.0), S(1.2)/
Figure 0-0 (Cont). Sample Dynamic Data Sheet

(Sheet 2 of 12)
13-10
GENSAL

machine # 2* I.
Efd EFD ISORCE
MBASE. Terminal
Bus

J+1 0.05 Tdo (>0) (sec) K+1 q
J+2 0.06 Tqo (>0) (sec) K+2 kd

K+4 Angle (radians)
J+4 1 Speed damping, D
J+5 1.5 Xd
J+6 1.2 Xq
J+7 0.4 X’d
J+8 0.25 Xd = Xq
J+9 0.12 Xl
J+10 0.03 S(1.0)
J+11 0.25 S(1.2)

machine MVA base.

(Sheet 3 of 12)
13-11
GENSAL

machine # 3* I.
Efd EFD ISORCE
MBASE. Terminal
Bus

J+1 0.05 Tdo (>0) (sec) K+1 q
J+2 0.06 Tqo (>0) (sec) K+2 kd
J+5 1.5 Xd
J+6 1.2 Xq
J+7 0.4 X’d
J+8 0.25 Xd = Xq
J+9 0.12 Xl
J+10 0.03 S(1.0)
J+11 0.25 S(1.2)

machine MVA base.

(Sheet 4 of 12)
13-12
SCRX
Bus Fed or Solid Fed Static Exciter
This model is located at system bus # 100 IBUS, ECOMP

machine # 1 I, VOTHSG
This model uses CONs starting with #____ J, VUEL
EFD
VOEL SCRX
XADIFD
ETERM

J 0.1 TA/TB K First integrator
J+1 10 TB (>0) (sec) K+1 Second integrator
J+2 200 K
J+3 0.05 TE (sec)
J+4 0 EMIN (pu EFD base)
J+5 5 EMAX (pu EFD base)
J+6 0 CSWITCH
J+7 0 rc/rfd
Set CSWITCH = 0 for bus fed.

Set CSWITCH = 1. for solid fed.
Set CON(J+7) = 0 for exciter with negative field
current capability.
Set CON(J+7) > 0 for exciter without negative
field current capability.
(Typical CON(J+7) = 10.)
IBUS, ’SCRX’, I, TA/TB, TB, K, TE, EMIN, EMAX, CSWITCH, rc/rfd/
CSWITCH = 0. CSWITCH = 1.
VREF
EMAX Et 1.0
+ Ebridge
EC – 1 + TAs K
 X Negative
(pu) 1 + TBs 1 + TEs EFD
Current Logic
LAFDIFD
+
EMIN
VS
VS = VOTHSG + VUEL + VOEL

(Sheet 5 of 12)
13-13
SCRX

EFD
VOEL SCRX
XADIFD
ETERM

J+2 200 K
J+3 0.05 TE (sec)
J+6 0 CSWITCH
J+7 0 rc/rfd

current capability.
VREF
EMAX Et 1.0
+ Ebridge
EC – 1 + TAs K
 X Negative
Current Logic
LAFDIFD
+
EMIN
VS
Figure 0-0 (Cont). Sample Dynamic Data Sheet (Sheet 6 of 12)
13-14
SCRX

EFD
VOEL SCRX
XADIFD
ETERM

J+2 200 K
J+3 0.05 TE (sec)
J+6 0 CSWITCH
J+7 0 rc/rfd

current capability.
VREF
EMAX Et 1.0
+ Ebridge
EC – 1 + TAs K
 X Negative
Current Logic
LAFDIFD
+
EMIN
VS

(Sheet 7 of 12)
13-15
SEXS
Simplified Excitation System
This model is located at system bus # 201 IBUS,

ECOMP
machine # 1 I,
This model uses CONs starting with #____ J, VOTHSG
EFD
VUEL SEXS
VOEL

J+2 100 K
J+3 0.1 TE (sec)
IBUS, ’SEXS’, I, TA/TB, TB, K, TE, EMIN, EMAX/
VREF EMAX
EC – 1 + TAs K
 EFD
(pu) 1 + TBs 1 + TEs
+
EMIN
VS

(Sheet 8 of 12)
13-16
HYGOV
Hydro Turbine-Governor

machine # 1 I,
This model uses CONs starting with #____ J, SPEED PMECH
HYGOV
and STATEs starting with #____ K, Speed
and VARs starting with #____ L.

J 0.05 R, permanent droop K e, filter output
J+1 0.75 r, temporary droop K+1 c, desired gate
J+2 8 Tr (>0) governor time constant K+2 g, gate opening
J+3 0.05 Tf (>0) filter time constant K+3 q, turbine flow
J+4 0.5 Tg (>0) servo time constant
J+5 0.2 + VELM, gate velocity limit VARs # Description
J+6 1 GMAX, maximum gate limit L Speed reference
J+7 0 GMIN, minimum gate limit L+1 h, turbine head
J+8 1.3 TW (>0) water time constant
J+9 1.1 At, turbine gain
J+10 0.5 Dturb, turbine damping
J+11 0.8 qNL, no power flow
IBUS, ’HYGOV’, I, R, r, Tr, Tf, Tg, VELM, GMAX, GMIN, TW, At, Dturb, qNL/
VAR(L) + 1 e 1 + T rs c 1 g
nref 
1 + Tfs rTrs 1 + Tgs
– Velocity and
Speed + Position Limits
SPEED  SPEED
R Dturb
X
–
g – 1 q +
 X   X At  PMECH
Tws +
q h
+ –
1. qNL
13-17
HYGOV

machine # 2 I,
HYGOV

VAR(L) + 1 e 1 + T rs c 1 g
nref 
– Velocity and
SPEED  SPEED
R Dturb
X
–
g – 1 q +
Tws +
q h
+ –
1. qNL
13-18
HYGOV

machine # 2 I,
HYGOV

VAR(L) + 1 e 1 + T rs c 1 g
nref 
– Velocity and
SPEED  SPEED
R Dturb
X
–
g – 1 q +
Tws +
q h
+ –
1. qNL
13-19
TGOV1
Steam Turbine-Governor

machine # 1 I,
SPEED PMECH
This model uses CONs starting with #____ J, TGOV1
and STATEs starting with #____ K,
and VAR #____ L.

J 0.05 R K Valve opening
J+1 0.5 T1 (>0) (sec) K+1 Turbine power
J+2 1.0 VMAX
J+3 0.3 VMIN VAR # Description
J+4 1.0 T2 (sec) L Reference
J+5 1.0 T3 (>0) (sec)
J+6 0. Dt
Note: VMAX, VMIN, Dt are in per unit on gener-

ator base.
T2/T3 = high-pressure fraction.
T3 = reheater time constant.
IBUS, ’TGOV1’, I, R, T1, VMAX, VMIN, T2, T3, Dt/
VMAX
+ 1 1 1 + T2s +
Reference  PMECH

VAR(L) R 1 + T1s 1 + T3s
– –
VMIN
 Dt
SPEED

(Sheet 12 of 12)
13-20
GENSAL

machine # 1* I,
Efd EFD ISORCE
and STATEs starting with #____ K, Source Current
MBASE. Terminal
Bus
*Power flow quantities.

J+1 0.05 Tdo (>0) (sec) K+1 q
J+2 0.06 Tqo (>0) (sec) K+2 kd
J+5 1.5 Xd
J+6 1.2 Xq
J+7 0.4 X’d
J+8 0.25 Xd = Xq
J+9 0.12 Xl
J+10 0.03 S(1.0)
J+11 0.25 S(1.2)

machine MVA base.
Figure 13-6. Dynamic Data Used for Representation of One Hydro Unit
13-21
3 × 5 MVA 1 × 10 MVA
Hydro Unit Steam Unit
kV
0.9 pf
kV
0.85 pf
SMALL SYSTEM EXAMPLE WED FEB 23, 1994 09:28
FILE SMEDD
100 'GENSAL' 1 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
100 'GENSAL' 2 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
100 'GENSAL' 3 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
201 'GENROU' 1 6 .05 1 .05 3 0 1.4 1.35 .3 .6 .2 .1 .03 .4/
100 'SCRX' 1 .1 10 200 .05 0 5 0 0/
100 'SCRX' 2 .1 10 200 .05 0 5 0 0/
100 'SCRX' 3 .1 10 200 .05 0 5 0 0/
201 'SEXS' 1 .1 .1 100 .1 0 3/
100 'HYGOV' 1 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
100 'HYGOV' 2 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
100 'HYGOV' 3 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
201 'TGOV1' 1 .05 .5 1. .3 1. 1. 0./
Figure 13-7. Basic Dynamics Data File for Setup of Simulation Model
of Small Example System
13-22
13.2.6 Dynamics Data Entry and CONEC, CONET Construction

Until valid subroutines CONEC and CONET have been loaded, the dynamic simulation
section of PSS®E is only a skeleton, used to bootstrap itself into the fully setup form. The
first step in the bootstrap process is the entry of dynamics data and construction of subrou-
tines CONEC and CONET.
Sheet 1 of Figure 13-8 shows the dialog of this process. First, the PSS®E link activity LOFL
is used to gain access to power flow activities. The converted saved case file, SME2, is
retrieved with activity CASE, and activity FACT is used to establish the required factors of
the network admittance matrix. Recovery and preparation of the network data is com-
pleted by activity RTRN, which regains access to dynamic simulation activities.
The reading of CON data values, assignment of storage in the dynamics data arrays (see
Table 1 and Section 12.4.3), and construction of subroutines CONEC and CONET, is han-
dled by activity DYRE on Sheet 1 of Figure 13-8. DYRE is instructed to place the CONEC
subroutine in file CC1 and the CONET subroutine in file CT1, respectively, as it builds
them.
The summary output produced by DYRE shows the number of CON, STATE, VAR, and
ICON array elements used by the models included in subroutines CONEC and CONET.
Figure 13-9 shows files CC1 and CT1, the CONEC and CONET subroutines constructed
by DYRE. In the file with CONEC is a routine USRXXX. Any user-written models would
be called here. The user should not modify this routine.
The last line printed by DYRE, at the bottom of Sheet 2 of Figure 13-8, shows that the sim-
ulation setup produced by DYRE uses CONS 1 through 123, STATES 1 through 43, VARs
1 through 7, and no ICONs. These numbers are stored in the Snapshot for future reference
because later steps of the simulation setup will require the assignment of additional space
in the CON, STATE, VAR, and ICON arrays.
Before exiting activity DYRE, the user is asked for a compiling command file for subrou-
tine CONEC and CONET, so the proper compile options are used.
While DYRE was executing, it ran certain data checks regarding the number of input val-
ues and whether the expected value was real or a character. The user should run DYCH
next to look for any model calling inconsistencies. This activity looks for items such as an
exciter on an SVS or a stabilizer on a machine with no exciter. This function was not run
in this example as there are no inconsistencies in the data. DYCH can also be used later to
remove or disconnect equipment for certain simulations.
13.2.7 Output Channel Selection and CONET Construction

Output channel selection is handled by activity CHAN and should be done immediately
after activity DYRE. Sheet 2 of Figure 13-8 shows the dialog of activity CHAN.
Activity CHAN first asks the user to specify the first channel, VAR and ICON to be
assigned in setting up output channels. These channels will normally be the next available
elements when assigning channels in an initial simulation setup. Because the program
keeps track of these values, the user will usually just hit the carriage return key.
The CHAN dialog assigns 19 variables from the simulation to the first 19 output channels.
The final summary output of activity CHAN, as spotted on Sheet 4 of Figure 13-8, shows
the next available entry in the channel arrays of Table 3 and in VAR and ICON to be 25,
13-23
17, and 13, respectively.

$ pssds4 Start PSS®E skeleton to initiate setup process
Starting skeleton PSSDS...
INITIATED AT DYNAMICS ENTRY POINT ON TUE FEB 15, 1994 10:08
ACTIVITY? LOFL
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? CASE SME2

CASE SME2.SAV WAS SAVED ON MON FEB 14, 1994 17:05 Pick up initial
condition power flow
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? FACT
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? RTRN
ACTIVITY? DYRE DYRE to read data and build subroutines

ENTER DYNAMICS DATA SOURCE FILENAME: SMEDD Dynamics Data File
ENTER FILENAME FOR SUBROUTINE CONEC: CC1 Place CONEC source in file CC1
ENTER FILENAME FOR SUBROUTINE CONET: CT1 Place CONET source in file CT1
ENTER FILENAME FOR PSSPLT RELAY CHARACTERISTIC DATA: CR1
NEXT AVAILABLE ADDRESSES ARE: File for relay

CON STATE VAR ICON characteristic data: CR1
1 1 1 1
ENTER STARTING CON, STATE, VAR, ICON OR CARRIAGE RETURN
OUT OF FILE DATA--SWITCH TO TERMINAL INPUT MODE
GENERATOR MODELS USE:

CONS 1- 50
STATES 1- 21
EXCITER MODELS USE:
CONS 51- 80
STATES 22- 29
GOVERNOR MODELS USE:

CONS 81- 123
STATES 30- 43
VARS 1- 7
SUMMARY OF MODELS READ:
GENS: GENROU GENSAL

1 3
EXSYS: SCRX SEXS

3 1
GOVS: TGOV1 HYGOV
1 3
NEXT AVAILABLE ADDRESSES ARE:

CON STATE VAR ICON Storage locations assigned by DYRE
124 44 8 1
Figure 13-8. Dialog of Basic PSS®E Dynamic Simulation Model Setup

(Sheet 1 of 4)
13-24
NO MODEL CALLS IN CONEC/CONET/USRXXX--DYNAMICS SKELETON MAY BE USED
ENTER FILENAME FOR COMPILING FILE (0 TO EXIT): COMPILE Creates command file
ACTIVITY? CHAN Use CHAN to set up outputs and for proper options

to add to subroutine CONET
CHANNEL VAR ICON
1 8 1 Program keeps locations used
ENTER STARTING CHANNEL, VAR, ICON INDICES OR CARRIAGE RETURN
ENTER OUTPUT CATEGORY:

0 = EXIT 1 = ANGLE 2 = PELEC
3 = QELEC 4 = ETERM 5 = EFD
6 = PMECH 7 = SPEED 8 = XADIFD
9 = ECOMP 10 = VOTHSG 11 = VREF
12 = BSFREQ 13 = VOLTAGE 14 = VOLT & ANG
15 = FLOW (P) 16 = FLOW (P&Q) 17 = FLOW (MVA)
18 = RELAY2 (R&X) 19 = VAR 20 = STATE
21 = MACH ITERM 22 = MACH APP IMP: 1
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 100,1
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 100,2 First three channels
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 201 will show rotor angle
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 0
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 100,2 Channels 4, 5, 6 will show
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 201 machine electrical power

ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 201 machine terminal voltage

Figure 0-0 (Cont). Dialog of Basic PSS®E Dynamic Simulation Model Setup (Sheet 2 of 4)
13-25

ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 201 machine field voltage

ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 0 turbine mechanical power
ENTER BUS NUMBER, 'IDENTIFIER': 150
ENTER BUS NUMBER, 'IDENTIFIER': 151 Channels 16, 17, 18 will
ENTER BUS NUMBER, 'IDENTIFIER': 0 contain bus voltages
ENTER FROM BUS, TO BUS, CIRCUIT ID, 'IDENTIFIER': 100 200 1
ENTER SECOND 'IDENTIFIER':
ENTER FROM BUS, TO BUS, CIRCUIT ID, 'IDENTIFIER': 100 150 1 Channels 19 through 24
ENTER FROM BUS, TO BUS, CIRCUIT ID, 'IDENTIFIER': 200 150 1 contain line P, Q flows
ENTER FROM BUS, TO BUS, CIRCUIT ID, 'IDENTIFIER': 0


CHANNEL VAR ICON
25 17 13
13-26
ACTIVITY? SNAP SSN1 Record setup progress to date

NUMBER OF ELEMENTS IN USE ARE:
CONS STATES VARS ICONS CHANNELS
123 43 16 12 24
ENTER NUMBER TO BE SAVED OR CARRIAGE RETURN FOR ABOVE VALUES
SNAPSHOT STORED IN FILE SSN1.SNP AT TIME = 0.000
ACTIVITY? STOP End of initial setup process with PSS®E skeleton
SUBROUTINE CONEC
C
$INSERT COMON4.INS
C
C
RETURN
END File CC1
SUBROUTINE USRXXX(MC,SLOT,IT)
INTEGER MC,SLOT,IT
SELECT (IT)
FIN
RETURN
END
SUBROUTINE CONET
C
$INSERT COMON4.INS
C
C
IF (.NOT. IFLAG) GO TO 9000
C
C NETWORK MONITORING MODELS File CT1
C
C
9000 CONTINUE
C
RETURN
END
Figure 13-9. Basic CONEC and CONET Subroutines as Built by Activity DYRE
13.2.8 Initial Condition Snapshot

Completion of activities DYRE and CHAN leaves completed dynamic simulation data
arrays in the PSS®E working files. These arrays must be preserved for use in the next
stages of the simulation work by making a Snapshot before PSS®E is stopped.
Activity SNAP is instructed to save the elements of arrays CON, STATE, VAR, and ICON
being used. By hitting the carriage return, the user saves only the arrays being used and
does not waste disk memory space.
13-27
13.2.9 End of Skeleton PSS®E Execution

Storage of the Snapshot completes the first part of the setup process. It is now necessary to
stop PSS®E, add any user code, compile CONEC and CONET, and load these into the
PSS®E skeleton to produce a completed PSS®E setup.
13.2.10 Refinement of CONEC and CONET

Subroutines CONEC and CONET, as built by activity DYRE, provide the basic level of
system modeling. The statements built into CONEC and CONET by this activity usually
represent the major bulk of the required code. In many applications, however, it is advan-
tageous to insert special Fortran logic into CONEC and/or CONET. The user may insert
any meaningful Fortran statements into these subroutines before compiling them and link-
ing them into PSS®E. It is suggested that the CONEC and CONET subroutines be left in
files with names related to their application (e.g., CC1 and CT1 in this case).
Advanced uses of this capability are demonstrated in Section 19.9, Advanced Uses of
CONEC and CONET of the PSS®E Program Operation Manual.
13.2.11 Compilation and Linking of CONEC/CONET

The CONEC and CONET subroutines must be compiled and linked into the PSS®E skele-
ton.
Successful execution of these steps leaves a completed copy of the PSS®E dynamic simu-
lation code in the user’s User File Directory. This copy of PSS®E will carry out dynamic
simulations of the power system described by the CONEC and CONET subroutines with
which it was loaded.
A user may keep any number of pairs of CONEC and CONET subroutines in differently
named files in a User File Directory. The loaded copy of PSS®E in the User File Directory
includes only the intelligence from the CONEC and CONET subroutines last loaded with
the CLOAD4 command, however. Because the CLOAD4 procedure always links the sub-
routines for which compiled codes are held in files, CONEC and CONET, filenames must
be reassigned as necessary. One convenient arrangement is to keep all CONEC and
CONET subroutines in source file form only, and to compile each with the commands
specified above each time it is needed.
13.2.12 Start PSS®E

After PSS®E has been loaded with the appropriate CONEC and CONET subroutines, it is
ready for use in the checkout and execution of dynamic simulations. The dialog needed to
start PSS®E and initialize a dynamic simulation is shown in the first page of Figure 13-10.
13-28
a. Initialization of Dynamics Data
Figure 13-10. Checkout of Dynamics Simulation Setup

(Sheet 1 of 5)
13-29
b. Review of Dynamic Model Data via Activity DOCU
Figure 0-0 (Cont). Checkout of Dynamics Simulation Setup (Sheet 2 of 5)
13-30
TA/TB TB K TE EMIN EMAX SWITCH RC/RFD

0.100 10.000 200.0 0.050 0.00 5.00 0.0 0.00
** HYGOV ** BUS NAME BSKV MACH C O N ' S STATE'S V A R ' S

100 HYDRO 33.0 3 105- 116 38- 41 5- 6
R-PERM R-TEMP TR TF TG VELM GMAX GMIN TW AT DTURB QNL

0.050 0.750 8.00 0.050 0.500 0.200 1.00 0.00 1.30 1.10 0.00 0.080
ENTER 0 TO END LIST, 1 FOR NEXT PAGE:
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E TUE FEB 15, 1994 16:39
REPORT FOR ALL MODELS AT ALL BUSES BUS 201 [STEAM 3.30] PLANT MODELS
** GENROU ** BUS NAME BSKV MACH C O N ' S STATE'S

201 STEAM 3.30 1 37- 50 16- 21
MBASE Z S O R C E X T R A N GENTAP
10.0 0.00000+J 0.20000 0.00000+J 0.00000 1.00000
T'D0 T''D0 T'Q0 T''Q0 H DAMP XD XQ X'D X'Q X''D XL

6.00 0.050 1.00 0.050 3.00 0.00 1.4000 1.3500 0.3000 0.6000 0.2000 0.1000
S(1.0) S(1.2)
0.0300 0.4000
** SEXS ** BUS NAME BSKV MACH C O N ' S STATE'S

201 STEAM 3.30 1 75- 80 28- 29
TA/TB TB K TE EMIN EMAX

0.100 0.100 100.0 0.100 0.00 3.00
** TGOV1 ** BUS NAME BSKV MACH C O N ' S STATE'S VAR

201 STEAM 3.30 1 117- 123 42- 43 7
R T1 VMAX VMIN T2 T3 DT
0.050 0.500 1.000 0.300 1.000 1.000 0.000
REPORT FOR ALL MODELS AT ALL BUSES CHAN MODELS
*** CALL VOLMAG( 1, 8, 0) ***
ICON VOLTAGE VAR ANGLE VAR BUS NAME BSKV

1 8 0 150 LOAD 33.0
*** CALL VOLMAG( 2, 9, 0) ***

2 9 0 151 LOAD 3.30
*** CALL VOLMAG( 3, 10, 0) ***

3 10 0 200 STEAM 33.0
*** CALL FLOW1( 4, 11, 12, 0) ***
c. Activity DOCU Output
Figure 0-0 (Cont). Checkout of Dynamics Simulation Setup (Sheet 3 of 5)
13-31
ACTIVITY? DLST
ENTER OUTPUT DEVICE CODE:

0 FOR NO OUTPUT 1 FOR CRT TERMINAL
2 FOR A FILE 3 FOR QMS PS2000
4 FOR QMS_PS800 5 FOR HARD COPY TERMINAL
6 FOR ALTERNATE SPOOL DEVICE: 1
ENTER CON RANGE: 1 123
ENTER VAR RANGE: 1 16
ENTER STATE RANGE: 1 43
ENTER ICON RANGE: 1 12
ENTER OUTPUT CHANNEL RANGE: 1 24
CONS:
1: 5.000 2: 0.5000E-01 3: 0.6000E-01 4: 5.084

5: 1.000 6: 1.500 7: 1.200 8: 0.4000
9: 0.2500 10: 0.1200 11: 0.3000E-01 12: 0.2500
13: 5.000 14: 0.5000E-01 15: 0.6000E-01 16: 5.084
17: 1.000 18: 1.500 19: 1.200 20: 0.4000
21: 0.2500 22: 0.1200 23: 0.3000E-01 24: 0.2500
25: 5.000 26: 0.5000E-01 27: 0.6000E-01 28: 5.084
29: 1.000 30: 1.500 31: 1.200 32: 0.4000
33: 0.2500 34: 0.1200 35: 0.3000E-01 36: 0.2500
37: 6.000 38: 0.5000E-01 39: 1.000 40: 0.5000E-01
41: 3.000 42: 0.0000 43: 1.400 44: 1.350
45: 0.3000 46: 0.6000 47: 0.2000 48: 0.1000
49: 0.3000E-01 50: 0.4000 51: 0.1000 52: 10.00
53: 200.0 54: 0.5000E-01 55: 0.0000 56: 5.000
57: 0.0000 58: 0.0000 59: 0.1000 60: 10.00
61: 200.0 62: 0.5000E-01 63: 0.0000 64: 5.000
65: 0.0000 66: 0.0000 67: 0.1000 68: 10.00
69: 200.0 70: 0.5000E-01 71: 0.0000 72: 5.000
73: 0.0000 74: 0.0000 75: 0.1000 76: 0.1000
77: 100.0 78: 0.1000 79: 0.0000 80: 3.000
81: 0.5000E-01 82: 0.7500 83: 8.000 84: 0.5000E-01
85: 0.5000 86: 0.2000 87: 1.000 88: 0.0000
89: 1.300 90: 1.100 91: 0.0000 92: 0.8000E-01
93: 0.5000E-01 94: 0.7500 95: 8.000 96: 0.5000E-01
97: 0.5000 98: 0.2000 99: 1.000 100: 0.0000
101: 1.300 102: 1.100 103: 0.0000 104: 0.8000E-01
105: 0.5000E-01 106: 0.7500 107: 8.000 108: 0.5000E-01
109: 0.5000 110: 0.2000 111: 1.000 112: 0.0000
113: 1.300 114: 1.100 115: 0.0000 116: 0.8000E-01
117: 0.5000E-01 118: 0.5000 119: 1.000 120: 0.3000
121: 1.000 122: 1.000 123: 0.0000
VARS:
Call FLOW1 (7, 13, 14, 0)
1: 0.4043E-01 2: 1.000 3: 0.4043E-01 4: 1.000
5: 0.4043E-01 6: 1.000 7: 0.4000E-01 8: 1.045
9: 0.9947 10: 1.050 11: 1.008 12: -1.754
13: 11.01 14: 3.207 15: 9.007 16: 3.391
STATES:
1: 1.057 2: 0.8924 3: 0.4985 4: 0.0000

5: 0.7219 6: 1.057 7: 0.8924 8: 0.4985
9: 0.0000 10: 0.7219 11: 1.057 12: 0.8924
13: 0.4985 14: 0.0000 15: 0.7219 16: 1.014
17: 0.3086 18: 0.8830 19: 0.5554 20: 0.0000
21: 0.6981 22: 0.7746E-02 23: 1.721 24: 0.7746E-02
25: 1.721 26: 0.7746E-02 27: 1.721 28: 0.1670E-01
29: 1.856 30: 0.0000 31: 0.8086 32: 0.8086
33: 0.8086 34: 0.0000 35: 0.8086 36: 0.8086
37: 0.8086 38: 0.0000 39: 0.8086 40: 0.8086
41: 0.8086 42: 0.8000 43: 0.0000
d. Review of Data Values and Channels via Activity DLST
Figure 0-0 (Cont). Checkout of Dynamic Simulation Setup (Sheet 4 of 5)
13-32
ICONS:
1: 150 2: 151 3: 200 4: 100

5: 200 6: '1 ' 7: 100 8: 150
9: '1 ' 10: 200 11: 150 12: '1 '
Call FLOW1 (7, 13, 14, 0)

OUTPUT CHANNELS:
#:ADDR 1: 1 2: 2 3: 4 4: 4001 5: 4002

ROTOR ANGLE ROTOR ANGLE ROTOR ANGLE P ELECTRICAL P ELECTRICAL
BUS 100 MC 1 BUS 100 MC 2 BUS 201 MC 1 BUS 100 MC 1 BUS 100 MC 2
IDENT ANG1HYDRO ANG2HYDRO ANG1STEAM P 1HYDRO P 2HYDRO
VALUE 41.360 41.360 39.996 0.40074E-01 0.40074E-01
#:ADDR 6: 4004 7: 12001 8: 12002 9: 12004 10: 16001

P ELECTRICAL E TERMINAL E TERMINAL E TERMINAL FIELD VLTAGE
IDENT P 1STEAM ET 1HYDRO ET 2HYDRO ET 1STEAM EFD1HYDRO
VALUE 0.80000E-01 1.0353 1.0353 1.0168 1.7821
#:ADDR 11: 16002 12: 16004 13: 20001 14: 20002 15: 20004
FIELD VLTAGE FIELD VLTAGE P MECHANICAL P MECHANICAL P MECHANICAL
IDENT EFD2HYDRO EFD1STEAM PM 1HYDRO PM 2HYDRO PM 1STEAM
VALUE 1.7821 1.8556 0.80148 0.80148 0.80000
#:ADDR 16: 44008 17: 44009 18: 44010 19: 44011 20: 44012
BUS VOLTAGE BUS VOLTAGE BUS VOLTAGE MW FLOW MVAR FLOW
150 151 200 100-200CKT1 100-200CKT1
IDENT V-LOAD V-LOAD V-STEAM P 100- 200 Q 100- 200
VALUE 1.0450 0.99470 1.0500 1.0076 -1.7542
#:ADDR 21: 44013 22: 44014 23: 44015 24: 44016

MW FLOW MVAR FLOW MW FLOW MVAR FLOW
100-150CKT1 100-150CKT1 200-150CKT1 200-150CKT1
IDENT P 100- 150 Q 100- 150 P 200- 150 Q 200- 150
VALUE 11.015 3.2071 9.0075 3.3913
Call FLOW1
(7, 13, 14, 0)
e. Output of Activity DLST
Figure 0-0 (Cont). Checkout of Dynamic Simulation Setup (Sheet 5 of 5)
The dynamics CON data is returned to the working memory of PSS®E by the command
RSTR, SSN1. The file SSN1 is the Snapshot of working memory taken just before shut-
ting down PSS®E to compile and load CONEC/CONET (see Figure 13-8, Sheet 4). With
Snapshot SSN1 restored, the working memory contains the values of all CONs specified
in the Dynamics Data File and of all ICONs established by activity CHAN. It contains no
other valid data.
13.2.13 Retrieval of Network Data and Initial Conditions

The network data and initial condition power flow are retrieved from file SME2 where
they were placed, in power flow converted saved case form, prior to the start of work on
dynamics, (see Section 13.2.3). The network data is recovered and prepared for use by the
activity sequence:
LOFL
CASE, SME2
FACT
RTRN
13-33
Following RTRN, the PSS®E working memory contains proper values of dynamics con-
stant data, network parameters, and initial condition network and generator terminal volt-
ages and loadings. All VAR and STATE array elements and all output channels (elements
of array CHAN) are still invalid.
13.2.14 Entry of Additional Data

Additional data not provided by the prior power flow setup work or by activities DYRE
and CHAN should be entered after the snapshot/saved case pair have been returned to the
PSS®E working arrays. Additional data would be entered in an interactive dialog with
activity ALTR. No such additional data entry is necessary in this example.
13.2.15 Updated Snapshot

The dynamics Snapshot should be updated any time that CON, ICON, VAR, or output
channel data are changed. No such update is necessary in this example.
13.2.16 Initialization
The previous steps established all required constant (CON, ICON) data and output chan-
nel assignments in PSS®E working memory, together with generator terminal conditions
as provided by the converted power flow. It is now possible to determine the initial condi-
tion (t = t–) values of all algebraic variables (VAR, PMECH, EFD, etc.) and state variables
representing conditions inside the generators, excitation systems, and other equipment
modeled by the library subroutines called from CONEC and CONET. This process is han-
dled by activity STRT; its dialog is shown in the first page of Figure 13-10. Activity STRT
runs two functions simultaneously:
1. It sets state and algebraic variables to their initial condition values.
2. It makes a number of data consistency and operating limit checks.
STRT works backward through all dynamic simulation models to determine the initial
condition values of all quantities in correspondence to the generator terminal loadings. For
example:
• Generator current follows from terminal voltage, real power, and reactive power.
• Generator field voltage, electrical torque, and flux linkages follow from terminal volt-
ages and current.
• Excitation system conditions follow from field voltage, etc.
The chain ends with the initialization of the excitation system voltage reference setpoints
in array VREF and of the turbine-governor load reference setpoints in designated VARs.
Consistency checking is handled in conjunction with this backward initialization process.
Details of activity STRT’s error and diagnostic messages are covered in Section 13.3.2.
The STRT dialog in Figure 13-10 shows a trouble-free initialization. The main output
report, shown by the star in Figure 13-10, gives generator loading, field voltage, rotor
angle, and power factor. The rotor angles shown here are angular spatial positions of the
generator rotors. Their spread reflects both the angular spread between phase angles of the
13-34
generator terminal voltages and the internal angle of each machine. The last two columns
of the tabulation are the generator’s direct and quadrature axis components of terminal
current, expressed relative to the rotor axis and illustrated by the machine phasor diagram
in Figure 13-11.
Figure 13-11. Phasor Diagram Used in Initializing Synchronous Machines
13.2.17 Initial Condition Error Checking

The message INITIAL CONDITIONS CHECK OK indicates that the initial values of the
time derivatives of all state variables are zero, or adequately close to zero, and that the ini-
tial condition, therefore, represents steady-state operation. If any state variable has an ini-
tial time derivative of significant magnitude the INITIAL CONDITIONS CHECK OK
message will be replaced by an INITIAL CONDITION ERROR listing of the values and
time derivatives of all such state variables.
These error conditions should all be checked out before the system model setup is
accepted as a valid starting point for dynamic simulation runs. Perseverance in tracking
down the causes of these errors is essential if good quality simulation results are to be
obtained, because they are very often symptoms of major errors or incompatibilities in the
power flow and dynamic simulation data on which the setup is based. The question of ini-
tial condition errors is covered in detail in Section 13.3.2.
13-35
13.2.18 Initial Condition Snapshot

On exit from STRT, both the working memory and the Snapshot, if one was made or
updated, contain a complete and initialized dynamic simulation setup. In the example, the
complete record of the simulation model setup and its initial condition is now contained in
the four system files:
CC1 CONEC, equipment structure and connection
CT1 CONET, system metering connections
SME2 Transmission system data
SSN1 Dynamic simulation data
The output file, GOP1, is selected to receive the output channels during the simulation
run. The channel headings, as established by activity CHAN, are written into the output
file.
Because error-free completion of STRT is the key point of the job setup process, STRT
allows the user to update a Snapshot immediately, and before leaving the STRT activity
itself. In this example, the Snapshot is placed in file SSN1 and updates (by overwriting)
the partial setup Snapshot that was in the file prior to execution of STRT. All setup work
done prior to this stage can be recovered quickly for a new simulation run by recompiling
and loading CONEC and CONET from CC1 and CT1 if they are not the pair currently
loaded with PSS®E and re-running the dialog shown in Figure 13-10.
13.2.19 Dynamic Simulation Documentation
13.2.19.1 Activity DOCU

The last four pages of Figure 13-10 show the reports produced by activities DOCU and
DLST and their documentation of the dynamic simulation setup. Pages 2 and 3 of Figure 13-
10 show the report of activity DOCU. These reports restate the information entered on the
model data sheets in Figure 13-5, and add the index values allocated by activity DYRE, in
the places left open on the data sheets.
The DOCU reports list data and setup indices for each equipment model called from either
the PSS®E tables or subroutines CONEC and CONET. The markings on the DOCU report
show the correspondence of indices and data. The boxed line shows the beginning indices
for the second unit at bus 100. The second line expands the indices of the call statement to
show that the 12 constants required by model GENSAL (see Figure 13-5) lie in CON array
elements 13 through 24, and that the four state variables of this model are present in
STATE array elements 6 through 10. The third line shows the data that has been taken
from the power flow Saved Case and associated with this model. The fourth and subse-
quent lines for this model show the significance and value of the 12 CONs used by this
model. Counting from the left shows that CON(21) is the value of Xd, which must be
identical to the imaginary part of ZSORCE.
Activity DOCU reports on all models called from the PSS®E model library.
The DOCU report is the most convenient way of documenting the majority of the dynamic
simulation data used in a setup.
13-36
Program Application Guide: Volume II Error Checking
13.2.19.2 Activity DLST

The DLST report, shown on pages 4 and 5 of Figure 13-10, may list any CON, VAR,
STATE, ICON, or channel assignment used in a simulation setup. In this example, DLST
has been used to display all data used in the setup. All CON data shown previously are
listed by DLST; so are all VARs referred to by models.
The markings on the DLST output in Figure 13-10 show the ICON values assigned and
VAR/channel values that result. This call of subroutine FLOW1 was placed by CHAN
when the user requested that the flow in branch 100-150, circuit 1, be placed in channels.
Activity CHAN assigned ICONs 7 through 9 to FLOW1 and gave them values of 100, 150,
and 1, respectively. Activity CHAN also assigned VARs 13 and 14 to this FLOW1 call so
that it could set them to the P and Q flows in the branch. CHAN also sets the 21st and
22nd elements of the channel arrays (see Table 3) to pick up VARs 13 and 14 and display
them in the two corresponding channels.
13.3 Error Checking
13.3.1 Gross Parameter Errors

The example shown in Figures 13-1 through 13-10 culminates in an apparently correct ini-
tialization. The data used in the example, and summarized in the last four pages of
Figure 13-10, contain at least one major error, nevertheless. At least one erroneous data
value is listed in both the DOCU and DLST reports. It is unlikely, though, that an engineer
reading these reports would find it, and it is certain that reading DOCU and DLST reports
for a large system is not a practical way of locating errors.
A preliminary check on the reasonableness of the CON values entered into the PSS®E
database via the Dynamics Data File (Figure 13-7) or by alteration of CON values may be
obtained by using activity DOCU in the checking mode. The checking selection instructs
DOCU to make a reasonableness check of the values of every CON used by a dynamic
simulation model. For example it checks CONs representing the generator parameter, Xd,
to ensure that their values lie in the reasonable range for this parameter, 0.8 Xd 2.0.
The checking mode of DOCU displays warning messages pertaining to each dynamic sim-
ulation model just ahead of the data report on that model. Figure 13-12 shows the recovery
of the setup made in the previous example. This report shows warning messages for all the
excitation system models used, and for the TGOV1 governor model at bus 201 used in the
simulation. The warnings given for the governor model and the SCRX exciter models sug-
gest that the flagged parameter values be checked for correctness. All are correct —
though at one extreme or another of typical ranges. The warning message preceding the
SEXS call does reveal a serious error, however. The value of Tb in the SEXS model cer-
tainly should not be 0.1 as detected by the checking mode of DOCU. A check of the SEXS
data sheet in Figure 13-5 shows that the value should be 10, not 0.1; the value was
mistyped in the Dynamics Data File (Figure 13-7, line 8).
13-37
Error Checking Program Application Guide: Volume II
ACTIVITY? RSTR SSN1 Recover Snapshot of dynamics data

SNAPSHOT SSN1.SNP WAS SAVED ON TUE FEB 15, 1994 16:37
NUMBER OF ELEMENTS RESTORED:

123 43 16 12 24
ACTIVITY? LOFL

CASE SME2.SAV WAS SAVED ON MON FEB 14, 1994 17:05 Retrieve initial condition
power flow
ACTIVITY? DOCU ALL

0 FOR NO OUTPUT 1 FOR CRT TERMINAL Document to check
4 FOR QMS_PS800 5 FOR HARD COPY TERMINAL parameter values
ENTER 0 FOR REPORTING MODE, 1 FOR DATA CHECKING MODE: 1
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E WED FEB 16, 1994 11:30
DATA CHECK FOR ALL MODELS AT ALL BUSES BUS 100 [HYDRO 33.0] PLANT MODELS
BUS 100 MACHINE 1:

TA/TB= 0.1000 K= 200.0000 TRANSIENT GAIN= 20.0000
** SCRX ** BUS NAME BSKV MACH C O N ' S STATE'S BUS

100 HYDRO 33.0 1 51- 58 22- 23 FED

0.100 10.000 200.0 0.050 0.00 5.00 0.0 0.00
BUS 100 MACHINE 2:

TA/TB= 0.1000 K= 200.0000 TRANSIENT GAIN= 20.0000 Suspect values
100 HYDRO 33.0 2 59- 66 24- 25 FED

0.100 10.000 200.0 0.050 0.00 5.00 0.0 0.00
BUS 100 MACHINE 3:

TA/TB= 0.1000 K= 200.0000 TRANSIENT GAIN= 20.0000

100 HYDRO 33.0 3 67- 74 26- 27 FED

0.100 10.000 200.0 0.050 0.00 5.00 0.0 0.00
Figure 13-12. Use of Activity DOCU to Review CON Values for Reasonableness
(Sheet 1 of 2)
13-38
DATA CHECK FOR ALL MODELS AT ALL BUSES BUS 201 [STEAM 3.30] PLANT MODELS
BUS 201 MACHINE 1:

TB= 0.1000

201 STEAM 3.30 1 75- 80 28- 29

0.100 0.100 100.0 0.100 0.00 3.00
Suspect
BUS 201 MACHINE 1: values
T1= 0.5000
T2= 1.0000 T3= 1.0000

201 STEAM 3.30 1 117- 123 42- 43 7
0.050 0.500 1.000 0.300 1.000 1.000 0.000
ACTIVITY? STOP
BUS 201 MACHINE 1:

TB= 0.1000

201 STEAM 3.30 1 75- 80 28- 29

0.100 0.100 100.0 0.100 0.00 3.00
BUS 201 MACHINE 1:

T1= 0.5000
T2= 1.0000 T3= 1.0000

201 STEAM 3.30 1 117- 123 42- 43 7
0.050 0.500 1.000 0.300 1.000 1.000 0.000
ACTIVITY? STOP
Figure 0-0 (Cont). Use of Activity DOCU to Review CON Values for Reasonableness (Sheet 2 of 2)
The checking mode of activity DOCU is useful primarily for detecting gross errors like the
typographical one pointed out above. It is not a complete data check, however, and an
execution of it with no warnings does not imply that all data is correct. The checking
mode of DOCU is useful both in initial model setup when data is being established for the
first time, and when an extensive series of changes has been made and errors may have
crept in.
13.3.2 Initialization Errors

A dynamic simulation model must be correct with respect to both equipment parameters
and initial operating point. The previous check using DOCU gave some help in reviewing
equipment parameter data, but contributed nothing to the checking of the initial condition
operating point. The most important checking of the initial operating point must begin
during the setup of the initial condition power flows. Generators must be checked for
operation at an acceptable voltage and power factor. Their output must be above minimum
13-39
load and below maximum turbine output. The mismatches of the initial condition power
flow must be acceptably small. The power flow can only check operating point variables
outward from the generator terminals. The checking of the initial condition with respect to
variables within the equipment dynamic models is assisted by activity STRT.
Each equipment model checks all variables to which limits are applied to ensure that the
initial variable value, as calculated, falls between the limits. An error message is printed if
any initial variable value calculated during initialization falls outside of its operating lim-
its. Calculated values are overridden by the appropriate limits.
Activity STRT attempts to calculate initial values of all variables to correspond to steady-
state operation, and then makes a single test calculation of the time derivatives of all state
variables. STRT executes the same three major steps that are used to determine the time
derivatives during the dynamic simulation itself:
• Call CONEC with mode = 3.
• Solve the transmission network.
• Call CONEC with mode = 2.
The network solution should converge in one or two iterations and all time derivatives
should be close to zero. Some small deviation of the time derivatives from zero is inevita-
ble as a result of the small but nonzero mismatches of the initial condition converted
power flow. STRT prints error messages stating the value of both time derivative and state
variable for each state variable for which the calculated derivative exceeds a small toler-
ance from zero.
Figure 13-13 shows the report from an execution of STRT made after considerable altera-
tions had been made to the constant data in Snapshot SSN1, both to correct the error iden-
tified in Figure 13-12 and to refine other parameter values. Figure 13-13 shows both OUT-
OF-LIMIT messages and INITIAL CONDITIONS SUSPECT messages. The OUT-OF-
LIMIT messages draw attention to the plant at bus 201. The INITIAL CONDITION SUS-
PECT messages draw attention to state variables 16 and 43 at bus 201. As the printout
shows, the state variable 16 is used by model GENROU connected at bus 201, and that
state variable 43 is used by model TGOV1, also connected at bus 21. The STRT report in
Figure 13-13 is followed immediately by a selective DOCU report for bus 201.
The first step in error checking is to check the limit parameters in the model data against
the calculated initial conditions. Checking the Emax value for SEXS shows that the genera-
tor field voltage ceiling is 1.75 pu, while the initial value calculated from power flow con-
ditions is 1.86. The conflict results in the 1.86 value being overridden by the limit of 1.75.
The 1.75 value is wrong, it should be 2.75. A similar check on the Vmax parameter in
TGOV1 shows that the turbine output is limited to 0.75 pu of the rated value. Because the
base MVA of the unit at bus 201 is 10.0, this corresponds to a maximum power of
7.5 MW, while the power flow value is 8.0 MW. The 0.75 value is wrong, it should be
0.95.
The next step is to check out the nonzero time derivative errors. The STRT message shows
that STATE(16) is the first state variable used by GENROU. Reference to the GENROU
data sheet in Figure 13-5 shows that this first state variable represents E’q of the generator.
E’q is held in equilibrium by the applied field voltage. Its nonzero derivative is, therefore,
13-40
due to the limitation of the excitation system output to 1.75 pu instead of 1.86 as needed
for equilibrium. A similar check shows that the nonzero derivative of STATE(43) is due to
the limitation of the value position in TGOV1. Both of the nonzero time derivatives, there-
fore, will be rectified when the errors in the values of CON(80) (Emax) and CON(119)
(Vmax) are corrected.
13-41
ACTIVITY? STRT
SEXS AT BUS 201 MACHINE 1 INITIALIZED OUT OF LIMITS
TGOV1 AT BUS 201 MACHINE 1 INITIALIZED OUT OF LIMITS Variables in SEXS and TGOV1 not valid
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E THU, AUG 06 1987 12:00
INITIAL CONDITION LOAD FLOW USED 1 ITERATIONS
--------------------------MACHINE INITIAL CONDITIONS---------------------------

X------ BUS -----X ID ETERM EFD POWER VARS P.F. ANGLE ID IQ
100 HYDRO 33.0 1 1.0353 1.7822 4.01 0.48 0.9928 41.36 0.5880 0.5247
100 HYDRO 33.0 2 1.0353 1.7822 4.01 0.48 0.9928 41.36 0.5880 0.5247
100 HYDRO 33.0 3 1.0353 1.7822 4.01 0.48 0.9928 41.36 0.5880 0.5247
201 STEAM 3.30 1 1.0167 1.8555 8.00 2.38 0.9585 40.00 0.6560 0.4934
INITIAL CONDITIONS SUSPECT:

I DSTATE(I) STATE(I) MODEL STATE BUS X-- NAME --X MACH State variables cannot
16 -0.17591E-01 1.0142 GENROU K 201 STEAM 3.30 1 be initialized for zero
43 -0.50007E-01 0.80001 TGOV1 K+1 201 STEAM 3.30 1
time derivative
ENTER CHANNEL OUTPUT FILENAME:
ENTER SNAPSHOT FILENAME:
ACTIVITY? DOCU
EFD upper limit is 1.75,
ENTER OUTPUT DEVICE CODE: power flow requests 1.855
2 FOR A FILE 3 FOR VERSATEC
4 FOR PRINTER 5 FOR HARD COPY TERMINAL
ENTER UP TO 20 BUS NUMBERS
201
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E THU, AUG 06 1987 12:00
*** CALL GENROU( 201,’1’, 4, 37, 16) ***
BUS NAME BASKV MACH KOUNT C O N ‘ S STATE’S SLOT

201 STEAM 3.30 1 4 37- 50 16- 21 4
10.0 0.00000+J 0.20000 0.00000+J 0.00000 1.00000
T’D0 T’’D0 T’Q0 T’’Q0 H DAMP XD XQ X’D X’Q X’’D XL

6.00 0.050 1.00 0.050 3.00 0.00 1.4000 1.3500 0.3000 0.6000 0.2000 0.1000
S(1.0) S(1.2)
0.0300 0.4000
*** CALL SEXS( 201,’1’, 4, 75, 28) ***
BUS NAME BSVLT MACH KOUNT C O N ‘ S STATE’S SLOT

201 STEAM 3.30 1 4 75- 80 28- 29 4

0.100 0.100 100.0 0.100 0.00 1.75
*** CALL TGOV1( 201,’1’, 4, 117, 42, 7) ***
BUS NAME BSVLT MACH KOUNT C O N ‘ S STATE’S VAR SLOT

201 STEAM 3.30 1 4 117- 123 42- 43 7 4
0.050 0.500 0.750 0.300 1.000 1.000 0.000
0
Figure 13-13. Error Conditions Detected by STRT and Use of DOCU to Track Down Errors
13-42
Program Application Guide: Volume II Running Simulations
The error-free completion of activity STRT does not imply a correct system model setup or
valid initial condition. It remains entirely possible for the setup to contain errors with
regard to both parameter values and operating point. Equipment parameters can fall
entirely within reasonable ranges, satisfying both DOCU and STRT, and yet be incorrect.
Additional checking of equipment parameters is covered in Chapter 24. Operating condi-
tions can be completely within equipment limits and in perfect steady state, satisfying
STRT, and yet still contain errors that will invalidate a study.
13.4 Running Simulations
13.4.1 Basic Steps

A dynamic simulation setup of the type covered in Section 13.2 provides the basis for a
wide range of simulation runs, both from the original initial condition and from new initial
conditions in which loadings and equipment statuses may have changed. Figure 13-14
shows a simple simulation run based on the setup as built up in Section 13.2. The principal
steps are:
• Recovery of the setup data by picking up converted power flow saved case SME2 and
Snapshot SSN1.
• Reinitialization with activity STRT.
• Numerical integration of the differential equations with activity RUN to advance the time
trajectory of the state variables.
• Application and removal of perturbations, such as faults and line switchings, with
activity ALTR.
Note that data items were changed from those originally entered, based on data checking.
The run shown in Figure 13-14, in summary, employs the following activities:
RSTR Recovers the Snapshot
LOFL, CASE,
FACT, RTRN Recovers the initial condition converted power flow
STRT Initializes the dynamic simulation run at t = –2.DELT (standard procedure of

STRT)
RUN Advances the system, unperturbed to t = 0–
ALTR Perturbs the system by application of a three-phase short circuit at bus 200.
(The fault is represented by a large shunt admittance.)
RUN Advances the system to t = 0.1–
ALTR Removes the fault, by removing the large shunt, and open branch 100-200
RUN Advances time to t = 3.0–
SNAP Records the conditions at t = 3.0– in file SSN2 in case the simulation needs to
be continued later
13-43
Running Simulations Program Application Guide: Volume II
LOFL
SAVE Records the network data at t = 3.0– in file SLF2 in case the run needs to be
continued later
Annotations on Figure 13-14 point out details of the dialog with each activity used in the
run. The results of the simulation run are deposited in file GOP, and may be plotted or
listed by utility program PSSPLT. Figure 13-15 shows a typical dialog with PSSPLT to pro-
duce plots of output channels versus time, and Figure 13-16 shows a series of representa-
tive plots of the simulation results.
13-44
$ pssds4
Starting your PSSDS...
INITIATED AT DYNAMICS ENTRY POINT ON WED FEB 16, 1994 17:17
ACTIVITY? RSTR SSN1 Recover initial condition Snapshot

SNAPSHOT SSN1.SNP WAS SAVED ON WED FEB 16, 1994 12:15

123 43 16 12 24
ACTIVITY? LOFL

CASE SME2.SAV WAS SAVED ON MON FEB 14, 1994 17:05

Recover initial condition
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? FACT converted power flow case
ACTIVITY? STRT Initialize


100 HYDRO 33.0 1 1.0353 1.7821 4.01 0.48 0.9928 41.36 0.5880 0.5247
100 HYDRO 33.0 2 1.0353 1.7821 4.01 0.48 0.9928 41.36 0.5880 0.5247
100 HYDRO 33.0 3 1.0353 1.7821 4.01 0.48 0.9928 41.36 0.5880 0.5247
201 STEAM 3.30 1 1.0168 1.8556 8.00 2.38 0.9585 40.00 0.6560 0.4934
INITIAL CONDITIONS CHECK O.K. Put results in file GOP for

subsequent plotting Run to t = 0-; print
ENTER CHANNEL OUTPUT FILENAME: GOP
ENTER SNAPSHOT FILENAME: channels every 2 steps;
No need for updating Snapshot plot channels (to GOP)
ACTIVITY? RUN
every 2 steps
AT TIME = -0.017 ENTER TPAUSE, NPRT, NPLT, CRTPLT: 0 2 2
CHANNEL OUTPUT FILE IS GOP
TIME ANG1HYDRO ANG2HYDRO ANG1STEAM P 1HYDRO P 2HYDRO
-0.017 41.360 41.360 39.996 0.40074E-01 0.40074E-01
P 1STEAM ET 1HYDRO ET 2HYDRO ET 1STEAM EFD1HYDRO
6 0.80000E-01 1.0353 1.0353 1.0168 1.7821
EFD2HYDRO EFD1STEAM PM 1HYDRO PM 2HYDRO PM 1STEAM
11 1.7821 1.8556 0.80148 0.80148 0.80000
V-LOAD V-LOAD V-STEAM P 100- 200 Q 100- 200
16 1.0450 0.99470 1.0500 1.0076 -1.7542
P 100- 150 Q 100- 150 P 200- 150 Q 200- 150
21 11.015 3.2071 9.0075 3.3913
0.000 41.360 41.360 39.996 0.40074E-01 0.40074E-01

6 0.80000E-01 1.0353 1.0353 1.0168 1.7821
11 1.7821 1.8556 0.80148 0.80148 0.80000
16 1.0450 0.99470 1.0500 1.0076 -1.7542
21 11.015 3.2071 9.0075 3.3913
a. Start-Up (Prefault)
Figure 13-14. Dialog for Dynamic Simulation Run

(Sheet 1 of 4)
13-45
ACTIVITY? ALTR Use ALTR to apply 3-phase fault at bus 200

TIME = 0.0000
ENTER CHANGE CODE:

0 = NO MORE CHANGES 1 = OUTPUT CHANNEL DATA
2 = CONS 3 = VARS
4 = CRT PLOT CHANNELS 5 = ICONS
6 = SOLUTION PARAMETERS 7 = STATES
8 = CASE HEADING: 0 No change to CONs, etc.
NETWORK DATA CHANGES (1=YES, 0=NO)? 1 Yes, fault is a network change
PICK UP NEW SAVED CASE (1=YES, 0=NO)? 0 No need
ENTER CHANGE CODE:
0 = EXIT ACTIVITY 1 = BUS DATA
2 = GENERATOR DATA 3 = BRANCH DATA
4 = TRANSFORMER DATA 5 = AREA INTERCHANGE DATA
6 = TWO-TERMINAL DC LINE DATA 7 = SOLUTION PARAMETERS
8 = CASE HEADING 9 = SWITCHED SHUNT DATA
10 = IMPEDANCE CORRECTION TABLES 11 = MULTI-TERMINAL DC DATA
12 = ZONE DATA 13 = AREA TRANSACTIONS DATA: 1 Fault is bus data change
ENTER BUS NUMBER (0 FOR NEW CHANGE CODE, -1 TO EXIT): 200 Bus 200
BUS DATA FOR BUS 200 [STEAM 33.0]:
CODE PLOAD QLOAD S H U N T I L O A D Y L O A D
OLD 1 0.00 0.00 0.00 0.00 0.0 0.0 0.0 0.0 CHANGE IT? 1
ENTER CODE, PLOAD, QLOAD, G, B, IP, IQ, YP, YQ
,,,,-2E11
NEW 1 0.00 0.00 0.00-0.2E+12 0.0 0.0 0.0 0.0 Represent fault by
AREA VOLT ANGLE NAME BASVLT LOSZON
large admittance
OLD 1 1.0500 -0.11 STEAM 33.000 1 CHANGE IT? 0 to ground
ENTER BUS NUMBER (0 FOR NEW CHANGE CODE, -1 TO EXIT): -1
ACTIVITY? RUN Run to t = 0.1-; Print

AT TIME = 0.000 ENTER TPAUSE, NPRT, NPLT, CRTPLT: .1 1 1
and plot every step
0.000 41.360 41.360 39.996 0.82596E-03 0.82596E-03
6 0.00000 0.32043 0.32043 0.30718 1.7821
11 1.7821
V-LOAD
1.8556
V-LOAD
0.80148
V-STEAM
0.80148
P 100- 200
0.80000
Q 100- 200
Conditions
16 0.22053E-01 0.20413E-01 0.40423E-09 0.99867E-01 0.99564 at t = 0+
P 100- 150 Q 100- 150 P 200- 150 Q 200- 150
21 0.14792 1.0113 0.22254E-09 -0.88674E-08
0.100 49.627 49.627 54.119 0.37375E-03 0.37375E-03

6 0.18626E-08 0.21516 0.21516 0.21663 1.0845 Conditions
11
16
1.0845
0.14797E-01
2.7500
0.13680E-01
0.80207
0.27751E-09
0.80207
0.45006E-01
0.77546
0.44870
at t = 0.1–
21 0.67119E-01 0.45596 0.59865E-10 -0.40855E-08
b. Fault Application (During Fault)
Figure 0-0 (Cont). Dialog for Dynamic Simulation Run (Sheet 2 of 4)
13-46
ACTIVITY? ALTR

TIME = 0.1000
ENTER CHANGE CODE:

2 = CONS 3 = VARS
8 = CASE HEADING: 0
NETWORK DATA CHANGES (1=YES, 0=NO)? 1

PICK UP NEW SAVED CASE (1=YES, 0=NO)? 0
ENTER CHANGE CODE:

12 = ZONE DATA 13 = AREA TRANSACTIONS DATA: 1
ENTER BUS NUMBER (0 FOR NEW CHANGE CODE, -1 TO EXIT): 200

OLD 1 0.00 0.00 0.00-0.2E+12 0.0 0.0 0.0 0.0 CHANGE IT? 1
,,,,0
NEW 1 0.00 0.00 0.00 0.00 0.0 0.0 0.0 0.0

OLD 1 0.0000 46.30 STEAM 33.000 1 CHANGE IT? 0
ENTER CHANGE CODE:

ENTER FROM BUS, TO BUS, CIRCUIT IDENTIFIER

(FROM BUS = 0 FOR NEW CHANGE CODE, -1 TO EXIT): 100 200 1
BRANCH DATA FOR CIRCUIT 1 FROM 100 [HYDRO 33.0] TO 200 [STEAM 33.0]:
STATUS LINE R LINE X CHARGING RATE-A RATE-B RATE-C
OLD 1 0.02000 0.20000 0.03000 15.0 17.5 20.0 CHANGE IT? 1
ENTER STATUS, R, X, CHARGING, RATE-A, RATE-B, RATE-C, # OF CIRCUITS
0
NEW 0 0.02000 0.20000 0.03000 15.0 17.5 20.0
LINE SHUNTS: BUS 100 [HYDRO 33.0] BUS 200 [STEAM 33.0]
OLD 0.00000 0.00000 0.00000 0.00000 CHANGE IT? 0
METERED END IS BUS 100 [HYDRO 33.0]. ENTER 1 TO REVERSE: 0
ENTER FROM BUS, TO BUS, CIRCUIT IDENTIFIER

(FROM BUS = 0 FOR NEW CHANGE CODE, -1 TO EXIT): -1
c. Fault Clearance
13-47
ACTIVITY? RUN
AT TIME = 0.100 ENTER TPAUSE, NPRT, NPLT, CRTPLT: 3 35 3

0.100 49.627 49.627 54.119 0.24792E-01 0.24792E-01
6 0.51703E-01 0.65921 0.65921 0.66068 1.0758
11 1.0758 2.7500 0.80207 0.80207 0.77546
16 0.65843 0.62578 0.66287 0.00000 0.00000
P 100- 150 Q 100- 150 P 200- 150 Q 200- 150
21 7.4376 1.0787 5.1703 2.0607
0.392 119.35 119.35 117.78 0.42846E-01 0.42846E-01

6 0.67378E-01 1.0311 1.0311 0.99718 4.2097
11 4.2097 2.7500 0.80883 0.80883 0.70099
16 1.0225 0.97320 1.0262 0.00000 0.00000
21 12.854 4.1344 6.7378 2.2363
2.725 212.93 212.93 211.68 0.38933E-01 0.38933E-01

6 0.82793E-01 1.0428 1.0428 1.0253 1.9074
11 1.9074 2.0502 0.79535 0.79535 0.82726
16 1.0417 0.99156 1.0466 0.00000 0.00000
21 11.680 3.2766 8.2793 3.2893

3.000 203.25 203.25 201.25 0.39259E-01 0.39259E-01
6 0.81273E-01 1.0399 1.0399 1.0227 1.9243
11 1.9243 2.0501 0.79783 0.79783 0.83028
16 1.0389 0.98885 1.0437 0.00000 0.00000
P 100- 150 Q 100- 150 P 200- 150 Q 200- 150
21 11.778 3.2330 8.1273 3.3026
ACTIVITY? SNAP SSN2

123 43 16 12 24
ACTIVITY? LOLF
INVALID ACTIVITY--PLEASE TRY AGAIN
ACTIVITY? LOFL
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? SAVE SLF2
CASE SAVED IN FILE SLF2.SAV ON WED FEB 16, 1994 17:26
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? STOP
d. Saving Progress Conditions
13-48
$ PSSPLT Start plot program

CHANNEL OUTPUT FILE PLOTTING PROGRAM -- PSSPLT-15.0
INITIATED ON WED, JAN 07 1987 11:06
ACTIVITY? CHNF GOP Results were put in GOP by PSS®E

ACTIVITY? SUBT
SUBTITLE LINE 1 IS
ENTER NEW SUBTITLE LINE OR <CR> TO LEAVE UNCHANGED
SMALL EXAMPLE STABILITY RUN
SUBTITLE LINE 2 IS Title for plots

ENTER NEW SUBTITLE LINE OR <CR> TO LEAVE UNCHANGED
3 PH 6 CY FLT AT 200 OPEN 100 TO 200
ACTIVITY? TINT
URRENT VALUE FOR TSTART = -999.0000, TSTOP = 999.0000

ENTER TSTART, TSTOP: 0 3 Horizontal scale 0 to 3 sec
ACTIVITY? SLCT
ENTER CHANNEL NUMBER (0 FOR NO MORE): 4 Pick channel 4 and set scale
CHANNEL 4 [ P 1HYDRO ] MIN = 0.0, MAX = 0.0
ENTER CMIN, CMAX (OR ‘R’ TO RE-SELECT CHANNEL): 0 .05
ENTER CHANNEL NUMBER (0 FOR NO MORE): 6 Pick channel 6 and set scale
CHANNEL 6 [ P 1STEAM ] MIN = 0.0, MAX = 0.0
ENTER CMIN, CMAX (OR ‘R’ TO RE-SELECT CHANNEL): 0 .1
ENTER CHANNEL NUMBER (0 FOR NO MORE): 0 No more curves this plot

ACTIVITY? PLOT
ENTER 25 CHARACTER LABEL :

*************************
: ELECTRICAL POWERS Specify label
SUPPORTED PLOTTING DEVICES ARE:
0 = NONE 1 = VERSATEC
2 = HP 7221A 3 = TEKTRONIX 4010
4 = TEKTRONIX 4014 5 = TEKTRONIX 4014 W/EGM
6 = TEKTRONIX 4662 7 = TEKTRONIX 4663
10 = CALCOMP 12 = KMW V.P.
17 = TEKTRONIX 4105/04/06 18 = TEKTRONIX 4107/09
22 = TEKTRONIX 4114 23 = TEKTRONIX 4115/4125
27 = HP 7470A 28 = HP 7475A
31 = QMS LASERGRAFIX 33 = TEKTRONIX 4111
36 = IMAGEN 37 = TEKTRONIX FILE
38 = HP-GL FILE 99 = INDE. PLOT FILE
ENTER DESIRED PLOTTING DEVICE: 31 Choose QMS
ACTIVITY? STOP
ENTER NUMBER OF COPIES (0 TO 5), DEVICE NAME FOR QMS LASERGRAFIX: 1 Request one copy
Job 01011 entered on queue QMSPLOT.
of plot
Figure 13-15. Dialog for Plotting Program PSSPLT

to Plot Results From GOP as Made During Run Shown in Figure 13-14
13-49
11:08
POWER
TECHNOLOGIES,
SMALL STABILITY EXAMPLE
INC.R 3 PH 6 CY FLT AT 200 OPEN 100 TO 200
THU FEB 17, 1994

FILE: GOP
ELECTRICAL PLOT
CHNL# 6: [P 1STEAM]
0.10000 0.0
CHNL# 4: [P 1HYDRO]
3.0000
0.06000 0.0
2.7000
2.4000
2.1000
1.8000
TIME (SECONDS)
1.5000
1.2000
0.90000
0.60000
0.30000
0.0
a. Electrical Powers
Figure 13-16. Representative Plot of the Simulation Results

(Sheet 1 of 5)
13-50
11:35
POWERS
POWER
TECHNOLOGIES,
THU FEB 17, 1994

FILE: GOP
CHNL# 15: [PM 1STEAM]

1.0000 0.0
CHNL# 13: [PM 1HYDRO]
1.0000 0.0
CHNL# 6: [P 1STEAM]
0.20000 0.0
CHNL# 4: [P 1HYDRO]
0.20000 0.0
3.0000
2.7000
2.4000
2.1000
1.8000
TIME (SECONDS)
1.5000
1.2000
0.90000
0.60000
0.30000
0.0
b. Powers
Figure 0-0 (Cont). Representative Plot of the Simulation Results (Sheet 2 of 5)
13-51
GENS
POWER
TECHNOLOGIES,
FILE: GOP
11:53
CHNL# 12: [EFD1STEAM]
5.0000 0.0
CHNL# 10: [EFD1HYDRO]
THU FEB 17, 1994

5.0000 0.0
CHNL# 9: [ET 1STEAM]
1.2000 0.0
CHNL# 7: [ET 1HYDRO]
1.2000 0.0
3.0000
2.7000
2.4000
2.1000
1.8000
1.5000
TIME (SECONDS)
1.2000
0.90000
0.60000
0.30000
0.0
c. Generators
13-52

13:16
POWER
TECHNOLOGIES,
VOLTAGES
FILE: GOP
THU FEB 17, 1994

CHNL# 18: [V-STEAM]
1.2000 0.0
CHNL# 17: [V-LOAD]
1.2000 0.0
CHNL# 16: [V-LOAD]
1.2000 0.0
3.0000
2.7000
2.4000
2.1000
1.8000
TIME (SECONDS)
1.5000
1.2000
0.90000
0.60000
0.30000
0.0
d. Voltages
13-53

13:22
POWER
TECHNOLOGIES,
FILE: GOP
LINE FLOWS
THU FEB 17, 1994
CHNL# 24: [Q 200- 150]
20.000 0.0
CHNL# 23: [P 200- 150]
20.000 0.0
CHNL# 22: [Q 100- 150]
20.000 0.0
CHNL# 21: [P 100- 150]
20.000 0.0
3.0000
2.7000
2.4000
2.1000
1.8000
TIME (SECONDS)
1.5000
1.2000
0.90000
0.60000
0.30000
0.0
e. Line Flows
13-54
13.4.2 Simulation Procedure
13.4.2.1 Plotting Interval

The variable NPLT specified to activity RUN informs PSS®E that the output channels
should be transferred to the output file every NPLT time steps. Except for the initial brief
unperturbed run from initialization to the time at which the first perturbation is to be
applied (usually t = 0), NPLT should be an odd number.
Figure 13-17a shows the effect of changing the plotting interval on the display of simula-
tion results. Plotting results every time step is the best that can be achieved; it reveals the
high-frequency oscillation superimposed on the slower variation. Plotting every time step
is expensive in terms of disk file space, however; it makes the PSS®E dynamics output
files so large that it is not practical to retain more than one or two at a time on the disk.
Economical use of disk storage requires that output channels be recorded in the output file
as infrequently as is consistent with a reasonable quality of display.
Figure 13-17b shows that plotting every second time step would completely mask the pres-
ence of the high-frequency component. It also shows that, while it would not reveal the
details of the high-frequency component, plotting every third time step would at least indi-
cate that some such component is present in the result. Of course, a high-frequency ripple
with a true period close to three time steps, instead of two time steps, as in Figure 13-17,
would be masked by plotting every third step and indicated by plotting every second step.
A plotting interval of an odd number of steps is recommended because the most trouble-
some high-frequency ripples found in digital dynamic simulation results have a period of
two steps. The most common ripple causing problems is not a true high-frequency compo-
nent of the system’s response, but an alternating positive and negative deviation of calcu-
lated values from true values due to marginal numerical stability, or numerical instability
of the digital integration process. Such an alternating deviation has an effective period of
two time steps and hence is most readily revealed if output is plotted at intervals of an odd
number of steps.
While plotting every three, five, or seven time steps is usually a reasonable compromise
between plot quality and economy of disk space, it is not uncommon for numerical insta-
bilities to go unnoticed even when the above precautions are observed. In such trouble-
some situations, the only reliable alternative is to plot the output channels every time step
(NPLT = 1). Even plotting every step can sometimes leave doubt as to whether a jagged
plot is due to a true high-frequency component or to numerical instability. In these cases,
the only way to obtain an authoritative answer is to reduce the time step and rerun the sim-
ulation, again plotting every time step.
13.4.2.2 Unperturbed Simulation

Discrimination between a true oscillatory or unstable simulation result and a numerical
instability is often assisted by making a simulation run with no perturbation. Such a run
should, in a perfect computer, remain completely steady with no deviation from the initial
condition, regardless of the number of integration steps taken. In practice, the finite word
length and associated rounding errors in the computer will give rise to small deviations
from the initial condition. If both the system being simulated and the numerical integration
13-55
process are stable, these errors will not build up beyond a small level corresponding to the
machine’s precision limitation, regardless of the number of time steps taken. If either the
system being simulated or the numerical integration is unstable, the unperturbed simula-
tion will eventually exhibit a growing deviation from the initial condition. Hence, if a sim-
ulation that is behaving suspiciously in a perturbed simulation run but does not deviate in
a long unperturbed simulation, it is at least reasonable to suppose that the numerical pro-
cess is stable.
13-56
a. True Result
Plotted Every 2 Time Steps
Plotted Every
3 Time Steps
b. Calculated and Plotted Results
Figure 13-17. Distortion of Simulation Display by Adjustment of Plotting
13-57
If, on the other hand, the unperturbed simulation fails to remain steady, the question of
numerical instability versus true instability remains. In this case, the diagnosis should be
based on:
1. Numerical instabilities are almost always overcome by reduction of the integration step
width.
2. The majority of numerical instabilities experienced with the algorithms used in PSS®E
and in power system applications develop in unperturbed runs as a deviation of one or
more state variables with growing amplitude and alternating sign.
13.4.2.3 Applying and Removing Disturbances

The majority of disturbances are applied and removed by using the NETWORK DATA
CHANGES option of activity ALTR. This option causes ALTR to initiate an automatic
linkage to activities CHNG and FACT to modify the network data. The specific actions to
be used for some more common disturbances are summarized in Figure 13-18 and Table 2.
Network switching changes such as fault application, fault clearance, and branch tripping
may be applied either to the network data that is currently in working memory, or to a new
set of network data that is picked up from a converted power flow saved case. New net-
work data is picked up by the PICK UP NEW SAVE CASE option of activity ALTR. A
reply of yes to the question PICK UP NEW SAVE CASE? causes PSS®E to pick up new
network data, but not a new voltage solution, from a designated power flow saved case.
This option often avoids complications in undoing the application of a fault or other dis-
turbance by returning directly to the predisturbance network data.
Dynamics Changes (New Values for CONs, VARs, ICONs, etc.)
Network Changes?
No Yes
New Network Case?
Yes
Get New Load No
Flow Case Name
CHNG
Network Data
CASE Name
and Loading Special Fixed
Changes Sequence Linkage
of Power Flow
New Ordering Needed? Activities
Yes
ORDR No
FACT
Dynamics Activity Selection
Figure 13-18. Implied Linkage via Activities CHNG and FACT from Activity ALTR
13-58
Table 2: Application of Some Common Perturbations
Operation or Disturbance Implementation in PSS®E

Add large shunt admittance B = –2E10 at bus i
Apply 3-phase short circuit at bus i.
with ALTR.
Either pick up prefault power flow case with
PICK UP NEW SAVE CASE option of ALTR, or
Clear fault at bus i.
return shunt admittance at bus i to prefault
value with ALTR.
Open circuit. Set circuit status to 0* with ALTR.
Reclose circuit. Set circuit status to 1** with ALTR.
Clear a bus by tripping all circuits connected Branch to power flow activities with LOFL and
to it. use activity DSCN.
Trip all generators at a bus. Change bus type code to 1 with activity ALTR.
Change MBASE for the bus from m × UBASE to
Trip one or more of a group of identical units
n × UBASE with ALTR; where UBASE = unit
at a bus that are represented by a single
® base MVA and number of units is changed from
generator in PSS E.
m to n.
Trip a single generator, leaving other gen- Change status, ISG, of generator to 0, leaving
erators connected. status of other generators at the bus as 1.
*If line reactors are represented by bus data, instead of as direct line connected reactors,
adjust shunt susceptance on bus at each end.
** If outage isolates any bus, change its type code to 4.
For example, short circuit faults involve the addition of a shunt admittance to the shunt
that is already present at a bus. This initial shunt represents the effect of a shunt connected
device. Clearance of such faults requires restoration of the shunt admittance at the bus to
its prefault value. This can be done without the need to take note of the prefault shunt
value by using the PICK UP NEW SAVE CASE option of activity ALTR to pick up the
same saved case that was used in initialization. Consider a simulation involving a three-
phase short circuit at the bus shown in Figure 13-19.
The sequence of activities are shown in Table 3 .
The alternative procedure for fault clearance would be to use ALTR, without the PICK UP
NEW SAVE CASE option, to change the reactive shunt at the faulted bus from –2E10
back to –7.109 before opening the branch. This has the same effect as recovering the pre-
fault Saved Case but is more susceptible to error, particularly when several prefault data
values must be re-entered to accomplish removal of a disturbance.
Take care in using the PICK UP NEW SAVE CASE option when the simulation includes
relay models or dynamic load models. For details refer to the PSS®E Program Operation
Manual Section 19.7.2, Applying Disturbances.
13-59
v = 1.05
B = +j20
P + jQ = 50 + j30
a. From Case F1 Before CONL
v = 1.05
+j20
–j27.21
I p = 47.62
b. After CONL (P = 100% Constant I; Q = 100% Constant B) Case F2
v0
–j2E10
–j27.21
Ip  0
c. After Application of Large Fault Admittance
v = 0.791
+j20
-j27.21
I p = 47.62
d. After Recovery of Converted Saved Case F2 NEW SAVED CASE

Option of Activity ALTR and Opening Branch
Figure 13-19. Use of NEW SAVE CASE Option

to Recover Prefault System Data and Status
13-60
Table 3: Sequence of Activities to Recover Prefault System Data and Status
RSTR Pick up Snapshot

LOFL, CASE,F2, FACT, Pick up initial condition power
Bus as in Figure 13-19b
RTRN flow
STRT Initialize
RUN Run to t = 0–
ALTR (no new Saved Case);
Apply fault; add to present data Bus as in Figure 13-19c
change shunt to (0,–2E10)
RUN Run through fault
Clear fault; recover data from
ALTR (new Saved Case, F2); before fault; make changes from Bus as in Figure 13-19b;
open branch prefault conditions rather than bus as in Figure 13-19d
faulted
RUN Postfault simulation
13.4.2.4 Saved Case - Snapshot Pairs

The complete condition of a dynamic simulation at any instant may be recorded by mak-
ing a comparison Snapshot and power flow Saved Case. This comparison was done, for
example, at the end of the run shown in Figure 13-14 using file SSN2 for the snapshot and
file SLF2 for the power flow case.
The creation of the power flow case, while not essential in this example, is still advisable.
The run could be restarted by picking up the Snapshot, SSN2; picking up the prefault
power flow, SME2; and then using ALTR to open branch 100-200 to reconstitute the con-
ditions at t = 3.0–. This practice is susceptible to error, though, and the use of companion
Snapshot/Saved Case pairs is preferable.
A Snapshot and/or Saved Case may be made at any point in a simulation. It is recom-
mended that these files be made immediately before any major switching or other change
is applied in a simulation run.
1. An error in setting up the switched system condition can be corrected by exiting from
activity ALTR, re-entering ALTR, picking up the Saved Case just made with the PICK
UP NEW SAVE CASE option, and redoing the sequence of changes.
2. A disturbed network condition can be undone by recovering the predisturbance net-

work condition with the PICK UP NEW SAVE CASE option of activity ALTR, rather than
by a reverse sequence of changes.
3. A new run can take place with an alternative switching or disturbance if it is discovered
after a run has been advanced beyond the switching before an error is discovered. A
new run is made by recovering both the power flow case, with LOFL-CASE-FACT, and
the Snapshot as made before the erroneous switching, and then re-executing
activities ALTR, RUN, and so on.
13-61
13.4.2.5 Restarting a Simulation Run

The most basic restarting procedure follows this order:
RSTR (Snapshot at time, t)
LOFL
CASE (power flow Saved Case for time, t)
FACT
RTRN
RUN—to continue simulation
For example, the small example run of Figures 13-14 and 13-15 can be restarted and
extended to a longer time span by picking up files SSN2 and SLF2. Figure 13-20 shows the
dialog for this restart and extension of the run from t = 3 to t = 4 sec with no new perturba-
tion. At t = 4, the conditions are preserved in a new Snapshot File, SSN3, and a new power
flow Saved Case File, SLF3. These files will allow the run to be continued still longer, if
desired.
A simulation need not be carried on as a simple extension after a restart. Figure 13-21
shows the dialog and resulting plot for a restart from the end point, t = 3.0, of the run
shown in Figure 13-14 and continuation with the steam turbine generator at bus 201 being
tripped at t = 3.0+ sec. The basic restart procedure in this case may omit the FACT step,
because the recovery of the Snapshot and Saved Case is to be followed immediately by
activity ALTR, which will refactorize the network after the imposition of network
changes. The ALTR dialog in Figure 13-21a shows:
1. Change of bus 201 type code from 2 to 1 to remove all of its generation from service.
2. A yes reply to the question DO YOU NEED TO RE-ORDER? because a bus type code
has been changed.
Activity ALTR is followed by activity RUN and from this point onward, the dialog is no dif-
ferent in this restart case than it would be in an uninterrupted run.
13.4.2.6 Multiple - Alternative Runs

A PSS®E dynamics simulation Snapshot contains both the name of the dynamics output
file, if any, currently in use, and the position of the writing pointer within that file. Restart-
ing a dynamic simulation run from a Snapshot automatically causes the output channel
results for the continuation of the simulation to be placed immediately behind the results
recorded up to the point where the Snapshot was made. If the restart point is before the end
of the run in which the snapshot was made, the continuation results will overwrite the cor-
responding results from the original run. If the restart point is the end of the original run,
the file will be extended to accommodate the continuation results.
Figures 13-22 and 13-23 illustrate the management of output files when making three simu-
lation runs to evaluate alternative control actions:
Run 1: 5-sec simulation, no load shedding.
Run 2: 5-sec simulation, load shedding option A applied at t = 3.
Run 3: 5-sec simulation, load shedding option B applied at t = 4.
13-62
Figure 13-22 shows the form of results expected. Run 1 will be done first with Snapshots
being made at t = 3 and t = 4 so that run 2 and run 3 can be restarts; this will allow the 15
sec of simulation results to be produced with 5 + 2 + 1 = 8 sec of simulation calculations.
The results of the three runs must be stored in separate files so that comparison plots can
be made later. The dynamics output files to be used are OREF (Run 1), OSWA (Run 2),
and OSWB (Run 3).
13-63
$ pssds4
INITIATED AT DYNAMICS ENTRY POINT ON THU FEB 17, 1994 14:28
ACTIVITY? RSTR SSN2 Recover dynamics data from SSN2


123 43 16 12 24
ACTIVITY? LOFL
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? CASE SLF2

Recover network data
CASE SLF2.SAV WAS SAVED ON WED FEB 16, 1994 17:26 from Saved Case SLF2
5 DIAGONAL AND 4 OFF-DIAGONAL ELEMENTS Run to t = 4.0; print

ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? RTRN every 15 steps; plot
every 3 steps
ACTIVITY? RUN Continue simulation run
New output is
CHANNEL OUTPUT FILE IS GOP appended to existing
TIME ANG1HYDRO
3.000 203.25
ANG2HYDRO
203.25
ANG1STEAM
201.25
P 1HYDRO
0.39259E-01
P 2HYDRO
0.39259E-01
output in file GOP
6 0.81273E-01 1.0399 1.0399 1.0227 1.9241
11 1.9241 2.0501 0.79783 0.79783 0.83028
16 1.0389 0.98885 1.0437 0.00000 0.00000
P 100- 150 Q 100- 150 P 200- 150 Q 200- 150
21 11.778 3.2330 8.1273 3.3026
3.125 199.64 199.64 198.10 0.38909E-01 0.38909E-01

6 0.82105E-01 1.0388 1.0388 1.0216 1.9337
11 1.9337 2.0514 0.79877 0.79877 0.82872
16 1.0378 0.98778 1.0426 0.00000 0.00000
21 11.673 3.2323 8.2106 3.2904
3.875 192.69 192.69 190.74 0.39218E-01 0.39218E-01

6 0.80482E-01 1.0353 1.0353 1.0180 1.9673
11 1.9673 2.0533 0.80214 0.80214 0.81080
16 1.0341 0.98430 1.0389 0.00000 0.00000
21 11.765 3.2178 8.0482 3.2662
4.000 193.61 193.61 191.44 0.39385E-01 0.39385E-01

6 0.79922E-01 1.0350 1.0350 1.0176 1.9709
11 1.9709 2.0530 0.80231 0.80231 0.80726
16 1.0338 0.98401 1.0386 0.00000 0.00000
21 11.816 3.2162 7.9922 3.2648
ACTIVITY? SNAP SSN3

CONS STATES VARS ICONS CHANNELS Make new Snapshot and
123 43 16 12 24
ENTER NUMBER TO BE SAVED OR CARRIAGE RETURN FOR ABOVE VALUES Saved Case of conditions
at t = 4.0 to prepare for
further continuation
ACTIVITY? LOFL
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? SAVE SLF3
CASE SAVED IN FILE SLF3.SAV ON THU FEB 17, 1994 14:31
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? STOP
Figure 13-20. Simple Continuation of Dynamic Simulation Run
13-64
a. ALTR Dialog
Figure 13-21. Restarting Simulation Run with Immediate Application of New Perturbation
Trip All Generation at Bus 201 (Sheet 1 of 2)
13-65
15:33
POWER
TECHNOLOGIES,
CONTINUATION OF BASIC EXAMPLE RUN
TRIP STEAM UNIT AT 3 SEC
UNIT TRIP RUN

INC.R
THU FEB 17, 1994

FILE: GOP
CHNL# 17: [V-LOAD]

1.0500 0.80000
1.0500 0.80000
CHNL# 4: [P 1HYDRO]
0.20000 0.0
6.0000
5.4000
4.8000
4.2000
3.6000
TIME (SECONDS)
3.0000
2.4000
1.8000
1.2000
0.60000
0.0
b. Result of Continuation Run
Figure 0-0 (Cont). Restarting Simulation Run with Immediate Application of New Perturbation: Trip All
Generation at Bus 201 (Sheet 2 of 2)
13-66
System Average
Frequency
Plot File OSWA
2
Plot File OSWB
3
1
Plot File
OREF
0 1 2 3 4 5 Time
Saved Case L3
Saved Case L4
Snapshot S3
Snapshot S4
Figure 13-22. Use of Snapshot, Saved Case, and Dynamics Output Files
for Three Runs Examining Switching Alternatives
13-67
Run 1, Initialize and Run to 5 sec

No Switching
Output File = GOP
File GOP
Unswitched Run 0 to 5 sec
File OREF
Unswitched Run COPY GOP OREF
0 to 5 sec
Run 2, Restart at t = 3 sec from S3, L4

Switch Option A
Output File = GOP
Run to 5 sec
File GOP
Switch Option A
0 to 5 sec
File OSWA
COPY GOP OSWA
Switch Option A
COPY OREF GOP
0 to 5 sec
Run 3, Restart at t = 4 sec from S4, L4

Switch Option B
Output File = GOP
Run to 5 sec
File GOP
Switch Option B
0 to 5 sec
File OSWB
Switch Option B COPY GOP OSWB
0 to 5 sec
Figure 13-23. Management of Dynamics Output Files in Switching Operation Runs
All simulation runs should be made with a single working plot file, which can be copied
into three storage files as the runs are completed. Figure 13-23 shows the sequence of
PSS®E runs and file manipulations needed to accomplish the three runs.
13-68
Program Application Guide: Volume II Use of CONET in System Modeling
The first run places the entire 5-sec unswitched result into file GOP, which is immediately
duplicated for reference in file OREF. The second run reuses file GOP, leaving the results
of the first 3 sec untouched and overwriting the channel values of the last 2 sec with the
results for switch option A. File GOP is then duplicated again into file OSWA. Because
file GOP no longer contains the required result for Run 3 in the interval from 3 to 4 sec, its
original unswitched form is recovered by duplicating file OREF in place of GOP. The
restart procedure is then used again to make Run 3.
13.5 Use of CONET in System Modeling

The following example uses obsolete load type models that are called from subroutine
CONET. At present, load type models are called from the internal PSS®E tables, similar to
plant-related models, and are not called from subroutine CONET. The following is
included only as an example of using CONET. There are still some categories of models
that exist in CONET and, in future PSS®E releases, will also be executed from the PSS®E
internal tables. As this happens, the following example will no longer be relevant.
13.5.1 Calling Models via CONET

The basic example covered in Section 13.2 involved only equipment associated with gener-
ating plants. None of the equipment modeled involved a direct algebraic voltage-current
relationship. The CONET subroutine used in Section 13.2 was produced automatically by
activity DYRE and required no refinement prior to its use in the simulation.
Models not part of the PSS®E tables involving algebraic relationship between current and
voltage at a load (Type 1) bus must be called from subroutine CONET so that they are
executed at each iteration of the network solution process (Figure 12-10). Because
activity DYRE automatically includes equipment models and the program keeps track of
storage locations, it is recommended that new equipment models be added automatically.
Manual addition by allocating storage is prone to mistakes.
The following subsections illustrate the use of equipment modeling by showing an exam-
ple case in which the simulation setup from Section 13.2 is augmented by additional detail
in its representation of loads.
13.5.2 Example Using CONET Models
13.5.2.1 Added Modeling Requirements

Consider the following augmentations of the basic simulation setup from Section 13.2:
1. Recognize the sensitivity of all network impedances and admittances to local
frequency.
2. Model the real power constant current component of load at buses 150 and 151 as
varying according to:
13-69
Use of CONET in System Modeling Program Application Guide: Volume II
3. Model three-stage load shedding at bus 151 by a solid-state type underfrequency relay.
The augmented setup is to be constructed on the basis of the same converted power flow
case, SME2, used in Section 13.2, and is not to disturb the original setup. Accordingly, the
new setup will be recorded in three new files, SNU1, CC2, and CT2, for a new CONEC
subroutine and a new CONET subroutine, respectively.
When building upon a previous setup automatically, it is necessary to set up temporary
CONEC and CONET files, which have been named TEMCOC and TEMCOT. Table 4
shows the original file allocation table (Table 1) as augmented.
Table 4: Files to be Used in First Dynamic Simulation Setup Example
Power Flow Raw Data SMALL

Solved Initial Condition Power Flow SME1
Solved Converted IC Power Flow SME2
Dynamics Raw Data SMEDD
Initial Condition Snapshot SSN1
First Plotting Output File GOP1
Progress Snapshot SSN2
Progress Power Flow Case SLF2
Additional Power Flow Raw Data NEWMOD
Initial Condition Snapshot SNU1
Temporary CONEC Source TEMCOC
Temporary CONET Source TEMCOT
13.5.2.2 Model Data Sheets

The first step in adding equipment to an existing setup of the modeling subroutines is to
complete the data sheets for each model to be used, with respect to both parameter values
(Figures 13-24 through 13-26). The data is transferred to a new raw data dynamics file,
NEWMOD, Figure 13-27 shows a listing of the new raw data file.
13.5.2.3 Construction of Base CONEC/CONET Subroutines

The base CONEC subroutine is already available from the previous work (Section 13.2); it
is recorded in file CC1 and will have no changes to it because the new equipment needed
is all CONET models.
Figure 13-28 shows the dialog that creates the new model calls starting from the original
Snapshot, this is recorded in file TEMCOT. Figure 13-29 shows the new temporary subrou-
13-70
tine CONET that DYRE created and placed in file TEMCOT.
13-71
NETFRQ
Network Frequency Dependence Model
CALL NETFRQ from CONET

This model makes network parameters dependent
upon local bus frequency
0, ’NETFRQ’/
Figure 13-24. Network Frequency Dependence Model (NETFRQ) Model Data Sheet
13-72
LOADF
Frequency Sensitive Load Model
CALL LOADF (I,J) from CONET

The bus number where this model is
called is in ICON #______ I,
This model uses CONs starting with #______ J.
ICON # Value Description

I 151 Bus number
CONs # Description
J 1.6 Current load exponent, k
J+1 0 Real power load exponent, m
J+2 0 Reactive power load exponent, n
The constant current and constant power load components are made sensitive to bus frequency
according to:
IBUS, ’LOADF’, k, m, n/
Figure 13-25. Frequency Sensitive Load Model (LOADF) Model Data Sheet
for Bus 151
13-73
LODSHD
Underfrequency Load Shedding Relay Model
CALL LODSHD (I,J,L,M) from CONET

The bus number where this model is
called is in ICON #______ I,
It uses CONs starting with #______ J,
and VARs starting with #______ L,
and ICONs starting with #______ M.
ICON # Value Description

I 151 Bus number
CONs # Value Description

J 59.7 First load shedding point (Hz)
J+1 0.05 First point pickup time (sec)
J+2 0.5 First fraction of load to be shed
J+3 59.5 Second load shedding point
(Hz)
J+4 0.05 Second fraction pickup time
(sec)
J+5 0.3 Second fraction of load to be
shed
J+6 59.0 Third load shedding point (Hz)
J+7 0.05 Third point pickup time (sec)
J+8 0.2 Third fraction of load to be
shed
J+9 0.085 Breaker time (sec)
J+10 0 Nominal shunt Mvar not sub-
ject to load shedding
ICONs # Description
M First point delay flag
M+1 First point time-out flag
M+2 First timer status
M+3 Second point delay flag
M+4 Second point time-out flag
M+5 Second timer status
M+6 Third point delay flag
13-74
ICONs # Description
M+7 Third point time-out flag
M+8 Third timer status
VARs # Description
L First timer memory
L+1 Second timer memory
L+2 Third timer memory
IBUS, ’LODSHD’, CON list/
Figure 13-26. Underfrequency Load Shedding Relay Model (LDSHD) Model Data Sheet
0 'NETFRQ' / FILE NEWMOD

150 'LOADF' 1.6 0 0/
151 'LOADF' 1.6 0 0/
151 'LOADSHD' 59.7 .05 .5 59.5 .05 .3 59.0 .05 .2 .085 0/
Figure 13-27. Additional Dynamics Data File for Adding to Basic Model
13-75
$ pssds4
INITIATED AT DYNAMICS ENTRY POINT ON FRI FEB 18, 1994 09:38
ACTIVITY? RSTR SSN1 Recover previous initial condition snapshot


123 43 16 12 24
ACTIVITY? LOFL

Recover initial condition
converted power flow case

DYRE to read new data and
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? RTRN build new subroutines
ACTIVITY? DYRE,ADD CONEC and CONET
ENTER DYNAMICS DATA SOURCE FILENAME: NEWMOD
ENTER FILENAME FOR CONEC ADDITIONS: TEMCOC
ENTER FILENAME FOR CONET ADDITIONS: TEMCOT Place sources in files TEMCOC and TECCOT
ENTER FILENAME FOR PSSPLT RELAY CHARACTERISTIC DATA: TEMCOR

CON STATE VAR ICON Start allocating arrays after
124 44 17 13
ENTER STARTING CON, STATE, VAR, ICON OR CARRIAGE RETURN those already in use
OTHER MODELS USE:

CONS 124- 140
VARS 17- 19
ICONS 13- 24
MISC: LOADF LODSHD NETFRQ

2 1 1

CON STATE VAR ICON
141 44 20 25 Summary of locations in use
EDIT NEW CALLS INTO CONEC AND CONET--THEN COMPILE AND CLOAD4
Figure 13-28. Dialog for Adding More Raw Data

(Sheet 1 of 2)
13-76
ACTIVITY? CHAN

CHANNEL VAR ICON
25 20 25 Use activity CHAN to add
ENTER STARTING CHANNEL, VAR, ICON INDICES OR CARRIAGE RETURN
more output channels
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 100 1
ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 100 2 Channels 26-28 will be machine rotor speeds



CHANNEL VAR ICON
31 20 25
ACTIVITY? SNAP SNU1 Save new data

140 43 19 24 30
SNAPSHOT STORED IN FILE SNU1.SNP AT TIME = -0.017
ACTIVITY? STOP Must stop PSS®E to

recompile and relink
Figure 0-0 (Cont). Dialog for Adding More Raw Data (Sheet 2 of 2)
C--REPLACE IF (.NOT. IFLAG) STATEMENT IN ORIGINAL CONET FILE WITH THE FOLLOWING
C
CALL NETFRQ
CALL LOADF ( 13, 124)
C
C
C NETWORK MONITORING MODELS
C
CALL LODSHD( 15, 130, 17, 16)
Figure 13-29. Files Created by DYRE and CHAN in Making New Setup
13-77
13.5.2.4 Combining Automatic and Manual CONEC/CONET Sections

The combining of the old and new generated sections of CONET requires the use of the
editor program. PSS®E must be stopped, and a Snapshot taken to preserve work done to
date.
Because no user models or models being handled by subroutine CONEC are being used in
this example, file TEMCOC can be ignored.
Figure 13-30 shows the completed CONET subroutine after TEMCOT has been inserted
with the editor. The new CONEC and CONET subroutines must be incorporated into the
executable code of PSS®E, in place of the previous versions, before any additional work is
done on the new setup by relinking and reloading (see Section 13.2.11).
SUBROUTINE CONET
C
$INSERT COMON4.INS
C
CALL NETFRQ
C
C PLACE ABOVE CALLS BEFORE IFLAG TEST IN CONET
C
C
C ADD FOLLOWING CALLS AFTER IT
C
CALL LODSHD( 15, 130, 17, 16)
C
C
C
9000 CONTINUE
C
RETURN
END
Figure 13-30. Editing CONETs to Combine Old and New Setups
13.5.2.5 Recording of Simulation Setup Details

Both the executable code and the dynamics data arrays of PSS®E are now completely set
up in accordance with the new requirements outlined in Section 13.5.2.1 and simulations
may now be run with the new system representation. Before doing simulation runs, how-
ever, it is advisable to record full details of the simulation setup for future reference. A
complete record of the setup is given by the following data:
1. The name of the initial condition power flow Saved Case File, SME1.
2. The name of the converted power flow Saved Case File, SME2.
3. The name of the initial condition or preinitial condition Snapshot File, SNU1.
4. The names of the CONEC and CONET subroutine source files, CC2 and CT2.
5. The number of the last CON, VAR, ICON, STATE, and channel used in the setup; 140,
19, 25, 43 and 30, respectively, in this example.
6. Complete printed reports from activities DOCU and DLST, as shown in Figures 13-31
and 13-32, together with printed listings of the CONEC and CONET source files.
13-78
ACTIVITY? RSTR SNU1

SNAPSHOT SNU1.SNP WAS SAVED ON FRI FEB 18, 1994 11:09

140 43 19 24 30
ACTIVITY? LOFL

ACTIVITY? DOCU CN Document CONEC models

ACTIVITY? DOCU CT Document CONET models

PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E FRI FEB 18, 1994 11:10
REPORT FOR CONET MODELS AT ALL BUSES
*** CALL NETFRQ ***
*** CALL LOADF( 13, 124) ***
BUS NAME BSKV ICON C O N ' S

150 LOAD 33.0 13 124- 126
I EXP P EXP Q EXP

1.6000 0.0000 0.0000
*** CALL LOADF( 14, 127) ***
BUS NAME BSKV ICON C O N ' S

151 LOAD 3.30 14 127- 129
I EXP P EXP Q EXP

1.6000 0.0000 0.0000
*** CALL LODSHD( 15, 130, 17, 16) ***
BUS NAME BSKV ICON C O N ' S V A R ' S ICON'S

151 LOAD 3.30 15 130- 140 17- 19 16- 24
HZ-1 T1 FRAC-1 HZ-2 T2 FRAC-2

59.700 0.050 0.500 59.500 0.050 0.300
HZ-3 T3 FRAC-3 TB EXCL MVAR

59.000 0.050 0.200 0.085 0.0
Figure 13-31. DOCU of New Models Only

Rest Like Before (Figure 0-0)
13-79
ACTIVITY? DLST

ENTER STATE RANGE: 0
CONS:
124: 1.600 125: 0.0000 126: 0.0000 127: 1.600

128: 0.0000 129: 0.0000 130: 59.70 131: 0.5000E-01
132: 0.5000 133: 59.50 134: 0.5000E-01 135: 0.3000
136: 59.00 137: 0.5000E-01 138: 0.2000 139: 0.8500E-01
140: 0.0000
VARS:
17: 0.0000 18: 0.0000 19: 0.0000
ICONS:
13: 150 14: 151 15: 151 16: 0

17: 0 18: 0 19: 0 20: 0
21: 0 22: 0 23: 0 24: 0
25: 0
OUTPUT CHANNELS:
#:ADDR 25: 24001 26: 24002 27: 24004 28:116002 29:116003

SPD DEVIAT'N SPD DEVIAT'N SPD DEVIAT'N BUS FREQ DEV BUS FREQ DEV
BUS 100 MC 1 BUS 100 MC 2 BUS 201 MC 1 150 151
IDENT SPD1HYDRO SPD2HYDRO SPD1STEAM FRQ-LOAD FRQ-LOAD
VALUE 0.00000 0.00000 0.00000 0.00000 0.00000
#:ADDR 30:116004
BUS FREQ DEV
200
IDENT FRQ-STEAM
VALUE 0.00000
Figure 13-32. Listing of New Data by Activity DLST
Items 1 through 4 are recorded for this example in Table 4. Item 5 is stored in the Snapshot.
It is strongly recommended that similar tabulations, together with the printed reports of
item 6 be stored together in a notebook for every dynamic simulation setup involved in an
engineering study. All data in items 1 through 6 are needed for a complete reconstruction
of a simulation setup, and all are needed for troubleshooting work.
13.5.2.6 Simulation Run With New Setup

The new setup may now be used to make new simulation runs. Figure 13-33 shows the dia-
log for a simulation run, based on the new setup, in which the 10-MVA generator (carrying
8 MW) is tripped off the line at bus 201, leaving the small Boondocks Power and Light
system with 20 MW of load (see Figure 13-4), and 12.02 MW of generation from the three
remaining hydro units at bus 100. The purpose of the simulation run is to check that the
load shedding relay settings specified in Figure 13-26 will enable the system to recover
13-80
with the essential load still energized and without catastrophic voltage and frequency
excursions. Figure 13-33 shows by messages that the first and second stages of load shed-
ding pick up, and that the corresponding feeder circuit breakers open at t = 0.333 and t =
0.442, respectively. This removes (0.5 + 0.3) per unit of the 15-MW load (that is 12 MW)
at bus 151, leaving the system with a nominal 8 MW of load (3 MW remaining at bus 151
and 5 MW at bus 150); the shaft speed and bus frequency channels then indicate rapid
recovery of frequency.
The results of this simulation are plotted in Figures 13-34a through 0-0d. Figure 13-34a
shows the shaft speed of one of the hydro units and frequency at bus 150. The slight differ-
ence is due to the five time step filter time constant of the bus frequency detector model,
BUSFRQ. The load shedding certainly reverses the decline of frequency and, because it
leaves a 12-MW hydro turbine output available for 8 MW (nominal) of load, it produces a
major overfrequency as the hydro governors run their power back from their initial power
output of 0.8 per unit of rating to 0.6 per unit. The hydro turbine power response is shown
in Figure 0-0c. Figure 0-0b shows hydro unit terminal voltage jumps from below unity to
over 1.25 per unit when load is shed. This rapid rise of voltage, in conjunction with the
load real power characteristic (60% constant current and 40% constant admittance, as
shown in Figure 13-3) produces a rapid increase in load on remaining feeders when the first
stage of load is shed. Figure 0-0d shows how the output of the hydro plant is split between
lines 100-150 and 200-150 after tripping of the steam unit.
13-81
ACTIVITY? STRT Pick up Snapshot SNU1 and Saved Case SME2 before this

100 HYDRO 33.0 1 1.0353 1.7821 4.01 0.48 0.9928 41.36 0.5880 0.5247
100 HYDRO 33.0 2 1.0353 1.7821 4.01 0.48 0.9928 41.36 0.5880 0.5247
100 HYDRO 33.0 3 1.0353 1.7821 4.01 0.48 0.9928 41.36 0.5880 0.5247
201 STEAM 3.30 1 1.0168 1.8556 8.00 2.38 0.9585 40.00 0.6560 0.4934
INITIAL CONDITIONS CHECK O.K.
ENTER CHANNEL OUTPUT FILENAME: GOP

ENTER SNAPSHOT FILENAME: SNU1 Save initialized setup in SNU1,
NUMBER OF ELEMENTS IN USE ARE: replacing previous uninitialized setup
140 43 19 24 30
SNAPSHOT STORED IN FILE SNU1.SNP AT TIME = -0.017
ACTIVITY? RUN Predisturbance run

AT TIME = -0.017 ENTER TPAUSE, NPRT, NPLT, CRTPLT: 0 2 2
ACTIVITY? ALTR

TIME = 0.0000 Activity ALTR to trip all

ENTER CHANGE CODE: generation at bus 201
2 = CONS 3 = VARS
8 = CASE HEADING: 0 No dynamics array changes
NETWORK DATA CHANGES (1=YES, 0=NO)? 1
PICK UP NEW SAVED CASE (1=YES, 0=NO)? 0
ENTER CHANGE CODE:

Select bus 201

OLD 2 0.00 0.00 0.00 0.00 0.0 0.0 0.0 0.0 CHANGE IT? 1
1
NEW 1 0.00 0.00 0.00 0.00 0.0 0.0 0.0 0.0
Status to 1 to ignore
OLD 1 1.0168 3.50 STEAM 3.300 1 CHANGE IT? 0 all generation
Figure 13-33. Simulation Run with New Setup (Section 13.5)

to Observe Action of Load Shedding Relay at Bus 151 (Sheet 1 of 2)
13-82
ENTER BUS NUMBER (0 FOR NEW CHANGE CODE, -1 TO EXIT): -1 Reorder because a bus type
5 DIAGONAL AND 5 OFF-DIAGONAL ELEMENTS code was changed
ACTIVITY? RUN Run for 5 sec
0.000 41.360 41.360 0.00000 0.60132E-01 0.60132E-01
6 0.00000 0.95017 0.95017 0.00000 1.7821
11 1.7821 0.00000 0.80148 0.80148 0.00000
16 0.94114 0.89562 0.94460 4.5152 -0.61377
P 100- 150 Q 100- 150 P 200- 150 Q 200- 150 SPD1HYDRO
21 13.524 3.5345 4.5106 2.0234 0.45186E-09
SPD2HYDRO SPD1STEAM FRQ-LOAD FRQ-LOAD FRQ-STEAM
26 0.45186E-09 0.00000 0.50281E-08 0.43926E-08 0.49347E-08
LODSHD AT BUS 151 STAGE 1 PICKUP TIMER STARTED AT TIME = 0.200 FREQ = 59.696
Messages from
LODSHD AT BUS 151 STAGE 1 BREAKER TIMER STARTED AT TIME = 0.250 LODSHD model
0.250 18.270 18.270 0.00000 0.58129E-01 0.58129E-01
6 0.00000 0.92897 0.92897 0.00000 4.1877 indicating
11 4.1877 0.00000 0.79909 0.79909 0.00000 disconnection
16 0.91945 0.87496 0.92283 4.3647 -0.55138
21 13.074 3.4432 4.3602 1.9479 -0.84376E-02 of load
26 -0.84376E-02 0.00000 -0.66198E-02 -0.65662E-02 -0.66253E-02
LODSHD AT BUS 151 STAGE 2 PICKUP TIMER STARTED AT TIME = 0.308 FREQ = 59.486
LODSHD AT BUS 151 STAGE 1 BREAKER TIMER TIMED OUT AT TIME = 0.333
50.00 PERCENT OF INITIAL LOAD SHED
7.5 MW AND 4.0 MVAR (NOMINAL) SHED. VOLT = 0.9020 FREQUENCY = 59.430
LODSHD AT BUS 151 STAGE 2 BREAKER TIMER STARTED AT TIME = 0.358
LODSHD AT BUS 151 STAGE 2 BREAKER TIMER TIMED OUT AT TIME = 0.442
30.00 PERCENT OF INITIAL LOAD SHED
4.5 MW AND 2.4 MVAR (NOMINAL) SHED. VOLT = 1.1855 FREQUENCY = 59.272

0.483 -39.425 -39.425 0.00000 0.34178E-01 0.34178E-01
6 0.00000 1.2923 1.2923 0.00000 0.00000
11 0.00000 0.00000 0.79444 0.79444 0.00000
16 1.3532 1.3139 1.3554 2.5666 -4.3283
21 7.6868 -2.1124 2.5655 1.1030 -0.12384E-01
26 -0.12384E-01 0.00000 -0.11382E-01 -0.11369E-01 -0.11384E-01
5.000 2789.7 2789.7 0.00000 0.30796E-01 0.30796E-01

6 0.00000 1.0343 1.0343 0.00000 1.0074
11 1.0074 0.00000 0.74026 0.74026 0.00000
16 1.0890 1.0573 1.0909 2.3130 -3.1204
21 6.9259 -1.8077 2.3119 0.64822 0.61197E-01
26 0.61197E-01 0.00000 0.60943E-01 0.60942E-01 0.60943E-01
ACTIVITY? RUN Go on another 4 sec to see hydro governor action

5.000 2789.7 2789.7 0.00000 0.30796E-01 0.30796E-01
6 0.00000 1.0343 1.0343 0.00000 1.0074
11 1.0074 0.00000 0.74026 0.74026 0.00000
16 1.0890 1.0573 1.0909 2.3130 -3.1205
P 100- 150 Q 100- 150 P 200- 150 Q 200- 150 SPD1HYDRO
21 6.9259 -1.8079 2.3118 0.64824 0.61197E-01
SPD2HYDRO SPD1STEAM FRQ-LOAD FRQ-LOAD FRQ-STEAM
26 0.61197E-01 0.00000 0.60943E-01 0.60942E-01 0.60943E-01
ACTIVITY? STOP
Figure 0-0 (Cont). Simulation Run with New Setup (Section 13.5) to Observe Action of Load Shedding
Relay at Bus 151 (Sheet 2 of 2)
13-83
11:52
FREQ
POWER
TECHNOLOGIES,
INC.R TRIP STEAM UNIT
FRI FEB 18, 1994

FILE: GOP
CHNL# 29: [FRQ-LOAD]

0.15000 -0.0500
CHNL# 25: [SPD1HYDRO]
0.15000 -0.0500
9.0000
8.1000
7.2000
6.3000
5.4000
TIME (SECONDS)
4.5000
3.6000
2.7000
1.8000
0.90000
0.0
a. Frequency
Figure 13-34. Simulation Run with New Setup

(Sheet 1 of 4)
13-84
15:58
POWER
TECHNOLOGIES,
HYDRO PLANT 1
FRI FEB 18, 1994
FILE: GOP
CHNL# 10: [EFD1HYDRO]

5.0000 -1.000
1.4000 0.50000
CHNL# 4: [P 1HYDRO]
0.08000 0.0
7.0000
6.2000
5.4000
4.6000
3.8000
TIME (SECONDS)
3.0000
2.2000
1.4000
0.60000
-0.2000
-1.000
b. Hydro Plant 1
Figure 0-0 (Cont). Simulation Run with New Setup (Sheet 2 of 4)
13-85
15:10
POWER
TECHNOLOGIES,
FILE: GOP
FRI FEB 18, 1994

HYDRO TURBINE POWER
CHNL# 13: [PM 1HYDRO]
1.0000 0.0
9.0000
8.0000
7.0000
6.0000
5.0000
TIME (SECONDS)
4.0000
3.0000
2.0000
1.0000
0.0
-1.000
c. Hydro Turbine Power
13-86
16:07
POWER
TECHNOLOGIES,
FLOWS TO LOAD
FILE: GOP
FRI FEB 18, 1994

CHNL# 23: [P 200- 150]
10.000 0.0
CHNL# 21: [P 100- 150]
20.000 0.0
9.0000
8.0000
7.0000
6.0000
5.0000
TIME (SECONDS)
4.0000
3.0000
2.0000
1.0000
0.0
-1.000
d. Flows to Load
13-87
13-88
Chapter 14
Generator Modeling
Chapter 14 - Generator Modeling
14.1 Introduction
The dynamic simulation model library of PSS®E represents a wide, and constantly grow-
ing, variety of electromechanical equipment. New models are added to the library con-
stantly as new types of equipment are developed by the manufacturers, as research work
provides improved information on existing equipment, and as PSS®E user needs evolve.
In addition to calling on the PSS®E library models, users may create their own models to
handle the simulation of special items of equipment. The coding of model subroutines is
discussed in the PSS®E Program Operation Manual.
14.2 Standards on Equipment Modeling and Data Exchange

While there are no nationally or internationally sanctioned standards on the modeling of
power system equipment or on the exchange of data between simulation programs, there
are several data file formats that have become quite widely used and recognized as conve-
nient vehicles for mechanizing the burdensome task of building up databases for simula-
tions of interconnected systems. The PSS®E Dynamics Data File format may reasonably
be regarded as one of these widely recognized formats.
The PSS®E package includes its own data file formats, and provides interface programs
for several other data formats (see PSS®E Program Operation Manual). PTI endorses the
recommendations of the IEEE regarding the exchange of power flow and dynamic simula-
tion data. The IEEE recommendations reflect the views of many of the electric utility
companies, equipment manufacturers, and software suppliers presently active in North
America. These recommendations are published in the following sources:
1. "Computer Representation of Excitation Systems," IEEE Transactions on Power Appa-
ratus and Systems, Vol. PAS-87, No. 6, pp. 1460-1468, 1968 (Committee Report).
2. "Dynamic Models for Steam and Hydro Turbines in Power System Studies," IEEE
Transactions on Power Apparatus and Systems, Vol. PAS-92, pp. 1904-1915, 1973
(Committee Report).
3. "Procedures for the Exchange of Power Plant and Load Data for Synchronous Stability
Studies," IEEE Transactions on Power Apparatus and Systems, Vol. PAS-100, No. 7,
pp. 3229-3245, 1981 (Committee Report).
14-1
Generator Modeling PSS®E 33.4
Rules for Using PSS®E Equipment Models Program Application Guide: Volume II
4. "Excitation System Models for Power System Stability Studies," IEEE Transactions on
Power Apparatus and Systems, Vol. PAS-100, No. 2, pp. 494-509, 1981 (Committee
Report).
The recommendations given in these reports represent stages in the constant evolution of
power system equipment modeling. They are not, and cannot be, a final statement of prac-
tice because the designs of new equipment will continue to change, and the modeling
needs will change accordingly, as the equipment technology advances. These references
provide a base for the establishment of system dynamic simulation setups, but leave a con-
stant requirement for the introduction of new models to handle new equipment and study
situations where the common viewpoint taken by the IEEE Committees is not applicable.
The PSS®E library includes models corresponding to most of those included in references
1, 2, and 3, but also includes many models that are not covered in these references.
The selection of the best model for a given study must be made by the user of PSS®E. The
models corresponding to IEEE recommendations have the advantage, in some situations,
in that data is already available for them in existing computer data files. However, in other
situations the models may be insufficiently detailed or unable to represent an effect of spe-
cific interest, regardless of the availability of data.
14.3 Rules for Using PSS®E Equipment Models
14.3.1 Memory Reference

Each model is a Fortran subroutine for which the function is to pick up equipment param-
eter values from the CON array, terminal conditions from the system condition arrays,
state variables from the STATE array, and to place time derivative values into the
DSTATE array. The arguments of each model’s subroutine call direct the model to spe-
cific areas of the CON, system condition, and STATE and DSTATE arrays.
The model CALL for each individual unit of equipment must refer to a unique area of the
STATE, VAR and system condition arrays, but CALLs for identical units may refer to the
same areas in the CON and ICON arrays.
14.3.2 Model Calling Sequence

In order to properly coordinate the outputs and inputs of the library models, they should be
called from CONEC in the following order:
1. Generator models.
2. Current compensating models.
3. Excitation stabilizer models.
4. Excitation system models.
5. Turbine-governor models.
6. Miscellaneous models.
14-2
PSS®E 33.4 Generator Modeling
Program Application Guide: Volume II Generator Models
14.4 Generator Models
14.4.1 Source Impedance

The PSS®E generator models range from the simplest to highly elaborate representations
of the synchronous machine. All models share certain common features, however. All
generator models ultimately present the electric transmission network with a positive
sequence source voltage where instantaneous amplitude and phase are known and where
current is to be determined. The equivalent circuit used to represent the generator as a
boundary condition in the network solution is described in Section 5.6 and Figure 5-7.
Although, physically, a generator is best thought of as a voltage source behind the step-up
transformer and a dynamic impedance, it is represented in PSS®E by a Norton equivalent
in which the voltage source is replaced by an equivalent current source, ISORCE. The
determination of the Norton source current in the principal generator models is illustrated
in Figure 14-1.
Rotor flux linkage transients and magnetic saturation are the principal factors affecting the
dynamic behavior of synchronous machines in the perturbation frequency bandwidth, 0 to
about 10 Hz, covered by PSS®E dynamic simulation. The magnitude and phase of the
source current are determined at any instant as a function of the instantaneous values of
generator state variables (i.e., rotor circuit flux linkages, shaft speed, and rotor angle). The
value of the generator’s effective dynamic impedance, ZSORCE, may be either its tran-
sient or subtransient impedance, depending on which dynamic model is chosen to repre-
sent the behavior of rotor circuit flux linkages.
ISORCE
To System
GTAP:1 ZTRAN
1
y =
ZSOURCE
ISOURCE = (iq - jid) source × (cos + j sin)
For GENROU, GENSAL, GENDCO, GENROE, GENSAE:


( d" + jq" ) 
o
(iq - jid) source =
ZSOURCE
For GENCLS:
0 + jEFD
(iq - jid) source =
ZSOURCE
Figure 14-1. Generator Model Equivalent Current Source and Norton Equivalent Circuit
14-3
Generator Models Program Application Guide: Volume II
The theoretical background on the modeling of synchronous machines in conjunction with

the interconnecting electrical network is outlined in these references:
1. "Structure in the Computation of Power System Nonlinear Dynamical Response,"
IEEE Transactions on Power Apparatus and Systems, Vol. PAS-88, pp. 1-6, 1969,
(J.M. Undrill).
2. "FACE Multimachine Power System Simulator Program, IEEE PICA Conference

Record, 1969, (D.N. Ewart, R.P. Schulz).
3. "Equipment and Load Modeling in Power System Dynamic Simulation," U.S. Energy
Research & Development Agency Publication CONF 750867, pp. 394-418,
(J.M. Undrill).
The generator parameters ZSORCE, GENTAP, and ZTRAN must be included in the cor-
responding arrays of the power flow case used as the base for the dynamic simulation. If
the generator step-up transformer is represented as part of the transmission network, the
values of GENTAP and ZTRAN should be left with default values of unity and zero,
respectively.
14.4.2 Generator Rotor Modeling

The PSS®E library includes a family of generator models, as summarized in Table 1, that
allows generator rotor effects to be modeled in different levels of detail depending upon
study requirements, availability of data, and users’ preferences. Table 1 summarizes the
standard generator models in terms of effects modeled and data requirements.
Table 1: Summary of Generator Models in Terms of Data Used
Reactance and Model

Time Constants GENSAL and GENROU and
Used GENDCO GENTRA GENCLS
GENSAE GENROE
Xd     
Xq    
Xd     
Xq  
Xd   
Xq   
Xl    
Tdo    
Tqo  
Tdo    
Tqo    
Saturation
   
Factors
14-4
Table 1: Summary of Generator Models in Terms of Data Used (Cont.)
Reactance and Model

Time Constants GENSAL and GENROU and
Used GENDCO GENTRA GENCLS
GENSAE GENROE
Ta 
*Xq is assumed to be equal to Xd.
Table 2 summarizes the operational impedances used to represent the rotor flux dynamics,
and Figures 14-2 through 14-4 show the corresponding transfer function block diagrams.
All synchronous machine models use the generator open-circuit magnetization curve as
their source of data for representing saturation.
Figure 14-5 shows typical magnetization curves in per-unit form and Figure 14-6 defines
two S parameters used to codify the magnetization curves for computer modeling.
The generator rotor electromagnetic models, except GENCLS, inherently represent both
steady-state and dynamic machine properties. The steady-state properties of greatest
importance are the initial condition (t–) values of flux linkages and of the per-unit field
voltage, Efd, needed to maintain a given real and reactive power loading. These are deter-
mined by the machine reactances and the modeling of saturation. The dynamic effects of
greatest importance are the development of synchronizing and damping torques where
instantaneous magnitudes during nonsteady conditions are different from those existing in
the steady state. These effects are determined by the relative values of the synchronous,
transient, and subtransient reactances, and by the rotor circuit time constants. If they are
used with valid values for all of their data models, GENSAL, GENROU, GENSAE, GEN-
ROE, and GENDCO are an inherently correct representation of electromagnetic synchro-
nizing and damping effects over the entire 0 to 10 Hz bandwidth to which PSS®E is
applicable. No limit exists on the elapsed time period that these three models can simulate.
Models GENTRA and GENCLS represent synchronous machines at a more empirical and
less accurate level than the three more complete models. GENTRA is included in PSS®E
solely for compatibility with older stability analysis programs that used its level of repre-
sentation; its use is not recommended by PTI. GENTRA will usually represent transient
synchronizing torque effects reasonably accurately during the first swing of rotor angle
after a simple short circuit fault, but it does not give adequate representation of damping
effects.
The GENCLS model is intended to be used primarily as an effective short circuit current
source in setting up approximate equivalents of segments of large interconnected power
systems that are far removed from the area of specific interest.
14-5
Generator Models
Generator Modeling
Table 2: Generator Rotor Transfer Functions
Model GENSAL* GENROU**, GENDCO GENTRA GENCLS

Machine type Salient pole Solid rotor Laminated rotor Classical
d-axis rotor circuits Field plus one amortisseur Field plus one iron circuit Field winding only None
q-axis rotor circuits One amortisseur Two iron circuits None None
L’d
L’d L"d
Ld 1 + T’dos
Ld 1 + T’do s 1 + T"do s Ld
Ld(s) As for GENSAL N/A
Ld L’d
(1 + T’dos)
(1 + T’dos) (1 + T"dos)
Lq
L’ L"
14-6
L" Lq 1 +
q
T’qos 1+ T"qos
Lq(s) Lq 1 + T"qos Lq L’q N/A
Lq
(1 + T"qos) (1 + T’qos) (1 + T"qos)
Program Application Guide: Volume II

d-axis mutual inductance as a d- and q-axis mutual inductances
Saturation As for GENSAL None
function of Eq as a function of 
*GENSAE – same as GENSAL except saturation of d- and q-axis mutual inductances as an exponential function of 
** GENROE – same as GENROU except saturation of d- and q-axis mutual inductances as an exponential function of 
PSS®E 33.4
L d – L l
--------------------
L d – L l
– +
+ 1 1 L d – L d
E fd  ------------- +
 -------------- ----------------------
- +
  d
T do s E' T do s L d – L l
q
– –
L d – L d
--------------------------
- L d – L l d-axis
2
 L d – L l 
+ + +
+ L d – L d + id
 
–
1
 -------------- –  q
T qo s
+ q-axis
L q – L q iq
a. Model
Figure 14-2. Electromagnetic Model of Salient Pole Generator
14-7
L d – L l
-------------------
-
L d – L l
– +
+ 1 + 1 L d – L d +
E fd  -------------  -------------- -----------------------   d
T do s E q T do s L d – L l
– –
L d – L d
--------------------------
- L d – L l d-axis
L ad i fd 2
 L d – L l 
+ +
+ L d – L d + id
 

----------d-
//
L ad i fd
//
 q  L q – L l  
------------------------------------
-
// (L d – L l 
–
1
-------------- –  q
 T qo s
+ q-axis
L q – L q iq
b. Model
Figure 0-0 (Cont). Electromagnetic Model of Salient Pole Generator
14-8
L d – L l
--------------------
L d – L l
– +
+ 1 + 1 L d – L d +
E fd  -------------  -------------- ----------------------
-   d
T do s T do s L d – L l
– –
L d – L d
--------------------------
- Ld  – Ll d-axis
L ad i fd 2
 L d – L l 
+
+
+ +
 L d – L d  id

----------d-
//
L ad i fd
//
 q  L q – L l 
------------------------------------
- 
// (L d – L l 
+
+ L q – L q + iq
 
–
–
L q – L q
--------------------------
- L q – L l
2 q-axis
 L q – L l 
+
1 1 L q – L q
------------- + -------------- ----------------------
- + – q
T qo s  T qo s L q – L l 
– +
L q – L l
--------------------
L q – L l
Figure 14-3. Electromagnetic Model of Round Rotor Generator

(Models GENROU, GENROE, and GENDCO)
14-9
+ 1
E fd  -------------- +   d
T do S
–
– +
L d
L ad i fd
E q
+
– L d – L d id

–  d
Lq iq
Figure 14-4. Electromagnetic Model of Salient Pole Generator

Considering Field Winding Transients Only (Model GENTRA)
1.4
Machine A
1.2 Machine B
 e ac = E' q = 
1.0
0.8
Terminal Voltage, pu
0.6
0.4
0.2
0.0
0.0 0.4 0.8 1.2 1.6 2.0 2.4
Field Voltage, pu  E fd = L ad i fd 
Figure 14-5. Typical Generator Open-Circuit Saturation Curves
14-10
Figure 14-6. Definition of Saturation Factor, S, for Entry as Generator Data
14.4.3 Magnetic Saturation

All PSS®E generator models except GENCLS recognize that magnetic saturation affects
the various mutual and leakage inductances within the machine. The specific assumptions
regarding which inductances are affected by saturation, and the relative effect of that satu-
ration, are different in the salient pole and solid rotor models.
• GENSAL and GENTRA assume that saturation affects only the direct axis and that the
mutual inductances vary as a function of the flux linkage, E q , behind transient
reactance.
• GENROU, GENROE, GENSAE, and GENDCO assume that saturation affects both
direct and quadrature axis reactances and that the mutual inductances vary as a func-
tion of the flux linkage behind subtransient reactance.
14-11
• GENROU, GENDCO, GENSAL, and GENTRA assume the saturation curve to be qua-
dratic, while GENROE and GENSAE assume the saturation curve to be exponential.
When the quadratic saturation is used, saturation is calculated using the equation:
B(E - A)2
S =
E (1.1)
where E is the input and A and B are such that the points (1.0, S1.0) and (1.2, S1.2) lie
on the curve as shown for the exciter in Section 16.3.
When the exponential saturation is used, saturation is calculated by the equation:
S = S1.0 × EX
where:
.and E is the input.
All six of these generator models recognize that the extent of the saturation effect depends
upon both rotor (field) and stator currents. Accordingly, they derive the effective saturated
parameters of the machine at each instant by internal calculation from the specified (con-
stant) unsaturated values of machine reactances and the instantaneous internal flux level.
The user should enter unsaturated values of all machine reactances in setting up PSS®E
data, and must enter an appropriate set of open-circuit magnetization curve data for each
machine.
14.4.4 Generator Rotor Speed Damping

All generator models calculate the rotor acceleration by:
dn Pmech - Den
2H = - Telec
dt 1+n (1.2)
where:
Pmech = (Pmech) nominal speed – Dtn.
Telec = The generator rotor electrical torque.
n= The per unit speed deviation of the generator.
Pe
De = = Represents the variation of electrical load with frequency,
n as seen from the generator.
Pt
Dt = = Represents the variation of turbine power with shaft speed.
n
De = Specified in generator data.

Dt = Entered as part of the turbine-governor model data.
14-12
The parameter, De, gives an approximate representation of the damping effect contributed
by the speed sensitivity of system loads. The parameter should not be used to represent the
effect of damping torques developed within the synchronous machines; these torques are
modeled explicitly by GENSAL, GENROU, GENDCO, GENSAE, and GENROE and are
included in the Telec term of (14.2).
The value of De could range from near zero for systems with predominantly resistive load
to approximately two for systems with a large percentage of pumping, fan, and other
industrial load.
The limitations of the use of De to represent load damping effects are discussed in Refer-
ence 3 of Section 14.4.1. If load damping is considered to be critical, it is advisable to
model load speed variation effects on the basis of local bus frequency by calling the load
model subroutine such as LOADF from subroutine CONET.
14.4.5 Generator Model Details
14.4.5.1 GENSAL and GENSAE (Salient Pole Machines)

GENSAL and GENSAE represent salient pole machines at the subtransient level. GEN-
SAL and GENSAE require that the subtransient reactance specified in the machine data be
exactly equal to the reactive part of the corresponding ZSORCE as used in the power flow
case. GENSAL and GENSAE treat the unit as follows:
• Online when the terminal bus type is 2 and machine status is 1.
• Off-line when the terminal bus type is 1 or 4 or machine status is 0.
When GENSAL or GENSAE represents several identical units on a plant bus (in a hydro
plant, for example), all data should be entered with respect to the base MVA of a single
unit, and the value of MBASE entered in the power flow case should be the sum of the rat-
ings of all connected units.
Reducing the value of MBASE has the effect in GENSAL or GENSAE of tripping the cor-
responding fraction of the units connected to the bus. The value of MBASE should not be
increased during a simulation run.
The tripping and subsequent reconnection of all units represented by a call to GENSAL or
GENSAE may be implemented by changing the bus type code to 1 (for tripping) and then
returning it to 2 (for reclosing) without changing the value of MBASE, or by changing the
unit status to 0 (for tripping) and returning it to 1 (for reclosing).
The values of Tdo and Tqo entered for GENSAL or GENSAE should not be less than
four times the integration time step, DELT. Typical values of DELT for simulation runs
using GENSAL or GENSAE are:
0.01 sec (1/2 cycle at 50 Hz)
0.00833 sec (1/2 cycle at 60 Hz)
If the value of DELT is too big (with regard to GENSAL or GENSAE) in any given study,
most likely it will be revealed by SUSPECT INITIAL CONDITIONS messages from
activity STRT referring to STATEs K+1 and K+2 of GENSAL or GENSAE models. Three
available remedies exist:
14-13
1. If only STATE (K+2) indicates suspect initial conditions, and if valid data are available,
use GENROU or GENROE to obtain a better q-axis representation.
2. Decrease DELT.
3. Increase Tdo or Tqo if these data values themselves are suspect or assumed values.
Tdo and Tqo seldom exceed 0.1 sec for typical generators.
14.4.5.2 GENROU and GENROE (Solid Rotor Generators)

GENROU and GENROE represent solid rotor generators at the subtransient level. Apart
from the additional q-axis data values, Tqo and xq, GENROU and GENROE are used in
exactly the same way as GENSAL.
14.4.5.3 GENCLS (Classical Generator Model)

GENCLS is the classical constant voltage behind transient reactance generator model. Set-
ting the type code of the terminal bus to 1, or the generator status flag to zero, removes the
unit from service.
GENCLS initializes EFD(I) to the initial condition value of Eq for the generator. After
being initialized in STRT, EFD(I) for GENCLS models should not be changed during a
run. It is not valid to use an excitation system to vary EFD in conjunction with the GEN-
CLS model.
14.4.5.4 GENTRA (Transient Saliency Level Generator Model)

GENTRA is a two-rotor model that neglects the amortisseur windings. Use of this model
will not reduce time in transient stability studies. Because there is no state in the q-axis,
the model must have its own acceleration procedure and may require more power flow
iterations. The model should not be used beyond 1 sec. Because the damping contributed
by the amortisseur windings is neglected, a larger time step to extend dynamic stability
cases would not be justified. A typical value for the acceleration factor is 0.3 but a smaller
value may be necessary. Care should be taken in using this model. A higher-order model
with assumed data may be preferable.*
14.4.5.5 GENDCO (Solid Rotor Generator Model Including Stator dc Offset Effects)
GENDCO is a detailed round rotor model that includes dc offset effects. The model
includes shaft torsional effects when it is associated with a call to the model SHAF25. The
model is intended for use only in shaft torsional studies. The time constant, Ta, for this
model should normally be zero except after a balanced simultaneous three-phase switch.
The dc offset approximation is not valid for any nonsymmetrical unbalances such as line-
ground faults; i.e., GENDCO should always have a Ta equal to zero in these situations.
Simulation Technique in PSS®E for Shaft Studies

The key requirement in shaft impact torque studies is correct dynamic characterization of
sudden changes in generator electromagnetic torque. The transient electromagnetic torque
* "Model Selection and Data Assembly for Power System Simulations," J.M. Undrill and T.F.
Laskowski, IEEE Transactions on Power Apparatus and Systems, Sept. 82, pp. 3333–3341.
14-14
profile of a generator following a disturbance on the transmission network contains many

components, some of which decay rapidly to zero, and some of which vary quite slowly
and persist as long as rotor angles continue to move. The transient torque produced by
each switching or sudden change (such as fault) on the transmission may be visualized as
containing the following components that are pertinent to shaft impact studies:
1. A square step change of torque.
2. A slow variation of torque in accordance with the relative angular positions of generator
rotors throughout the power system. This component falls in the frequency spectrum
between about 0.1 and 20 rad/sec.
3. A series of exponentially decaying components that round-off the corner associated

with component i. These components are associated with changing flux linkages in cir-
cuits (discrete or within the iron mass) on the generator rotor.
4. A power frequency (60 Hz) oscillating component associated with the decaying dc
offset component of current produced in the generator stator by a network disturbance.
5. A double power frequency (120 Hz) component associated with the negative-sequence
currents induced by an unbalanced fault.
The PSS®E dynamic simulation program represents torque components 1, 2, and 3,

including the subtransient part of component 3, in all of its generator models (except the
classical model), but does not reproduce components 4 and 5. Exact reproduction of com-
ponent 4, the dc offset effect, would require simulation of the entire transmission system
on a differential equation basis as used in switching surge studies. This simulation would
be prohibitively expensive in studies involving extensive networks and requiring exami-
nation of transients for the period of about 0.5 sec over which shaft impacts are of interest.
The PSS®E generator models GENDCO and SHAF25 handle the dc offset torque compo-
nent by taking advantage of the fact that the initial amplitude of this torque component can
be determined exactly from a pair of network solutions made for the instants immediately
before and immediately after the network switching. Because the corresponding rate of
decay of the dc offset torque component cannot be similarly determined, models SHAF25
and GENDCO require the user to enter an estimate of the time constant governing their
decay. This time constant is loosely approximated by the network R/X ratio as seen by the
generator. A representative value of this time constant is 0.15 sec.
The last component of generator torque associated with negative sequence currents is of
significance in shaft impact studies only if a shaft has a natural frequency very close to
120 Hz. It is neglected by the dynamic models of PSS®E.
Figure 14-7 illustrates the way in which a generator’s response to a balanced network
switching is made up by the addition of components 1 through 4.
14-15
Torque Torque Component
1 plus 2
1 plus 2 plus 3
1 plus 2, 3, 4
Time
Figure 14-7. Composition of Generator Transient Torque in Response to Switching
Components 1, 2, and 3 are considered to be significant in conventional transient stability

studies. Component 3 includes torques decaying with transient and subtransient compo-
nents. Component 1 is of concern in connection with steady-state stability because it
determines the final load on the generator. Component 2 is of concern principally from the
viewpoint of transient stability. Component 3 is of concern in dynamic stability studies
because, in addition to its rounding off of the first component, it contributes to the damp-
ing of rotor angle swings.
Component 4, while it can be an oscillation with a large initial peak-to-peak amplitude, is
usually neglected in transient stability studies because it decays rapidly to zero, has an
average close to zero, and because its initial amplitude is strongly limited by the presence
of impedance between the generator and the point of switching. Component 5 is normally
neglected in transient stability studies.
Shaft impact torques are influenced principally by components 1 and 4 of the generator
electromagnetic torque transient. Accordingly, shaft impact studies require generators to
be modeled at a level of detail where components 1 through 4 of the torque transient can
be represented. The subtransient part of component 3 should be represented accurately
because it may not have decayed to zero by the time of the next switching, and hence can
influence the magnitude of the step component of torque induced by this event.
14-16
14.4.5.6 FRECHG (Frequency Changer)

The frequency changer, FRECHG, models a motor-generator set with a salient pole syn-
chronous machine at each end of a common shaft. Each of the synchronous machines has
its own electromechanical characteristic but the mechanical speeds of the two must be the
same. A 60/50-Hz frequency changer could, therefore, be built using a 12-pole machine
for the 60-Hz side and a 10-pole machine for the 50-Hz side.
This subtransient level frequency changer model recognizes the components of torque
produced by the rotor body and amortisseur winding currents in both of the synchronous
machines.
Data for each machine should be entered on its own base. The inertia value entered for
each machine should be that of the entire shaft. Because the model allows different size
machines, the only data requirement for the set of salient pole machines is that data for the
inertias be entered so that the kinetic energy equation K.E. = 1/2 I2 is satisfied on both
ends (H is entered for the entire shaft).
Losses in the set are negligible so that in the power flow the real electrical power trans-
ferred through the air gap of the motor and generator should have the same magnitude
(i.e., after specifying an armature resistance in ZSORCE, the user must calculate and spec-
ify the electrical power at the terminals such that the sets have the same internal power).
If during initialization either machine is off-line, an appropriate message is printed and the
set is assumed off-line. Either machine can be tripped during a simulation by setting the
machine status to zero.
14.4.5.7 CGEN1 and TGEN1 (Third-Order Equivalent Circuit Model)

The third-order equivalent circuit models, CGEN1 and TGEN1, represent the most
detailed rotor circuit structure suggested for usage in power system stability studies. Pro-
ponents of models derived from standstill frequency response tests have at times used
third-order models to better fit the test data. The CGEN1 and TGEN1 models were devel-
oped in response to individual user requests. Inclusion of those models in the PSS®E
model library should not be interpreted as an endorsement of its use in preference to the
widely used GENROU and GENSAL models.
The format for this model departs from the traditional set of time constants and reactances
as used by GENROU and GENSAL. Instead, the rotor is represented by the equivalent cir-
cuits as shown in Figure 14-8 for the two axes with the data being entered as per-unit values
for the resistances and reactances. The data sheet for this model is shown in Figure 14-9.
Some of the resistances and inductances in the equivalent circuit can be set to zero. while
other quantities requiring nonzero values are indicated by >0 on the data sheet. Allowing
some quantities to be set to zero permits this model to represent a second-order circuit
such as being modeled in GENROU.
14-17
Rmd (2) Rfd

Rmd (1)
Lld Lmd (1) Lmd (2) Lfd
Lkd (1) Lkd (2)
Id Lad Ifd EFD
Rkd (1) Rkd (2)
d-Axis
Rmq (2) Rkq (3)

Rmq (1)
Llq Lmq (1) Lmq (2) Lkq (3)
Lkq (1) Lkq (2)
Iq Laq
Rkq (1) Rkq (2)
q-Axis
Figure 14-8. CGEN1 Equivalent Circuits
14-18
CGEN1
Third-Order Complex Generator Model
This model is located at system bus #____ IBUS,

machine #____ I, Pm PMECH SPEED
Speed
Efd EFD ISORCE Source
and STATEs starting with #____ K, Current
CGEN1
and VARs starting with L. VOLT TGEN1 ETERM Terminal
VT
The machine MVA base is __________ for each of at Terminal Voltage
Bus
___________ units = __________ MBASE. ANGLE Angle
ZSORCE for this machine is_________+ j
_________on the above MBASE.

J+24 0 Lmq (2)
J 4 Inertia, H
J+25 0 Rkq (2)
J+1 0.1 S(1.0)
J+26 0 Lkq (2)
J+2 0.4 S(1.2)
J+3 0.17 Lld > 0 STATEs # Description
J+4 1.63 Lad > 0 K  speed (pu)
J+5 0.00081 Rfd > 0 K+1 Angle (radians)
J+6 0.20234 Lfd > 0 K+2 rd (1)
J+7 0 Rmd (1) K+3 rd (2)
J+8 0 Lmd (1) K+4 rd (3)
J+9 0.01194 Rkd (1) K+5 rq (1)
J+10 0.09 Lkd (1) K+6 rq (2)
J+11 0 Rmd (2) K+7 rq (3)
J+12 0 Lmd (2)
J+13 0 Rkd (2) VARs # Description
J+14 0 Lkd (2) L Internal memory
J+15 0.17 Llq > 0 L+1 Internal memory
J+16 1.58 Laq > 0
J+17 0.011036 Rkq (3) > 0
All constants except S(1.0) and S(1.2) are in
J+18 0.500333 Lkq (3) > 0 pu machine MVA base.
J+19 0 Rmq (1) Set Rmd (2), Lmd (2), Rkd (2) and Lkd (2) to 0 for
J+20 0 Lmq (1) 2nd order d-axis model.
Set Rmq (2), Lmq (2), Rkq (2) and Lkq (2) to 0 for
J+21 0.00997 Rkq (1)
2nd order q-axis model.
J+22 0.07125 Lkq (1) See diagram below for definition of various
J+23 0 Rmq (2) resistances and inductances.
14-19
IBUS, ’CGEN1’, I, H, S(1.0), S(1.2), Lld, Lad, Rfd, Lfd, Rmd (1), Lmd (1), Rkd (1), Lkd (1), Rmd (2), Lmd
(2), Rkd (2), Lkd (2), Llq, Laq, Rkq (3), Lkq (3), Rmq (1), Lmq (1), Rkq (1), Lkq (1), Rmq (2), Lmq (2), Rkq
(2), Lkq (2)/
Rmd (2) Rfd

Rmd (1)
Lld Lmd (1) Lmd (2) Lfd
Lkd (1) Lkd (2)
Id Lad Ifd EFD
Rkd (1) Rkd (2)
d-Axis
Rmq (2) Rkq (3)

Rmq (1)
Llq Lmq (1) Lmq (2) Lkq (3)
Lkq (1) Lkq (2)
Iq Laq
Rkq (1) Rkq (2)
q-Axis
CGEN1 Equivalent Circuit
Figure 14-9. CGEN1 Data Sheet
14-20
Using the feature to represent a second-order response, this model was compared with
GENROU by simulation runs. Starting with derived parameters, the equivalent circuit
parameters were calculated to obtain the same rotor response. The data for the derived
parameters are the following:
Tdo = 6 sec
Tdo = 0.06 sec
Tqo = 0.5 sec
Tqo = 0.12 sec
T = 4 MW-sec/MVA
Xd = 1.8 pu
Xd = 0.35 pu
Xd = 0.23 pu
Xq = 1.75 pu
Xq = 0.55 pu
Xq = 0.26 pu
Xl = 0.17 pu
S(1.0) = 0.1
S(1.2) = 0.4
To calculate the equivalent circuit parameters the following equations were used:
Lad = Ld - Ll
1
Lfd =
1 1
-
L’d - Ll Lad
1
Lsd (1) =
1 1 1
- Ll - +
L"d Lad Ldf
Lad + Lfd
rfd =
T’do × o
Lad × Lfd
Lsd(1) +
Lad + Lfd
rsd (1) =
T"do × o
Laq = Lq - Ll
14-21
1
Lsq (3) =
1 1
-
L’q - Ll Laq
1
Lsq (1) =
1 1 1
- -
L"q - Ll Laq Lsq(3)
Laq + Lsq(3)
rsq (3) =
T’qo × o
Laq × Lsq(3)
Lsq(1) +
Laq + Lsq(3)
rsq (1) =
T"qo × o
The values for equivalent circuits that will produce the same response as GENROU with
the above data are shown in Figure 14-10. The disturbance for the simulation runs was a
three-phase fault (on the infinite bus) for three cycles. The field voltage was kept constant
in the simulation runs. The plots shown in Figure 14-11 reveal that the response of the two
models are the same. Because the representation of saturation is different for the two mod-
els, the initial values for the rotor angle and the field current were slightly different. In the
CGEN1 and TGEN1 model the saturation is accounted by changing the inductances Lad
and Laq by determining the K factors from the open circuit saturation curves while the
GENROU model accounts for saturation by adding a term at the summing junction as
shown in Figure 14-11. Also, the air gap flux is used to account for saturation in CGEN1
while in GENROU it is the subtransient flux.
Another feature of the CGEN1 and TGEN1 model is that it can represent subtransient
saliency, i.e., Ld Lq. A simulation run was made comparing the response for a machine
with and without subtransient saliency using an equivalent circuit for which the value for
Lq would be 0.26 pu and Ld = 0.23 pu. The plots shown in Figure 14-12 reveal hardly any
differences.
14-22
0.17 pu 0.2023 pu 0.00081 pu

•
0.09 pu
1.63 pu vfd
0.01194 pu
•
Direct Axis
0.17 pu
0.07125 pu 0.5003 pu
1.58 pu
0.00997 pu 0.01104 pu
Quadrature Axis
Figure 14-10. Equivalent Circuit for Machine Model Used in the Simulation
14-23
COMPARISON BETWEEN GENROU AND CGEN1

INITIAL CONDITION LOAD FLOW
19:54
POWER
TECHNOLOGIES,
INC.R
CHNL# 3: [Q CGEN1]
WED MAY 04, 1994

0.20000 FILE: cgen1 -0.2000
GENROU VS. CGEN1

CHNL# 2: [P CGEN1]
0.30000 FILE: cgen1 -0.3000
CHNL# 1: [ANGLE CGEN1]
60.000 FILE: cgen1 30.000
CHNL# 3: [Q GENROU]
0.20000 FILE: genrou -0.2000
CHNL# 2: [P GENROU]
0.30000 FILE: genrou -0.3000
CHNL# 1: [ANGLE GENROU]
60.000 FILE: genrou 30.000
4.0000
3.6000
3.2000
2.8000
2.4000
TIME (SECONDS)
2.0000
1.6000
1.2000
0.80000
0.40000
0.0
Figure 14-11. Comparison Between GENROU and CGEN1

(Sheet 1 of 2)
14-24
COMPARISON BETWEEN GENROU AND CGEN1

INITIAL CONDITION LOAD FLOW
19:56
POWER
TECHNOLOGIES,
INC.R
CHNL# 6: [EFD CGEN1]
WED MAY 04, 1994

4.0000 FILE: cgen1 1.0000
GENROU VS. CGEN1

CHNL# 5: [SPEED CGEN1]
0.00500 FILE: cgen1 -0.0050
CHNL# 4: [ET CGEN1]
1.2000 FILE: cgen1 0.30000
CHNL# 6: [EFD GENROU]
4.0000 FILE: genrou 1.0000
CHNL# 5: [SPEED GENROU]
0.00500 FILE: genrou -0.0050
CHNL# 4: [ET GENROU]
1.2000 FILE: genrou 0.30000
4.0000
3.6000
3.2000
2.8000
2.4000
TIME (SECONDS)
2.0000
1.6000
1.2000
0.80000
0.40000
0.0
Figure 0-0 (Cont). Comparison Between GENROU and CGEN1 (Sheet 2 of 2)
14-25
CGEN1 WITH AND WITHOUT

SUBTRANSIENT SALIENCY
08:26
POWER
TECHNOLOGIES,
INC.R
CHNL# 3: [Q CGEN1]
THU MAY 05, 1994

CGEN1 BEHAVIOR
0.20000 FILE: cgen1 -0.2000
CHNL# 2: [P CGEN1]
0.30000 FILE: cgen1 -0.3000
60.000 FILE: cgen1 30.000
CHNL# 3: [Q CGEN1]
0.20000 FILE: cgen1sal -0.2000
CHNL# 2: [P CGEN1]
0.30000 FILE: cgen1sal -0.3000
60.000 FILE: cgen1sal 30.000
4.0000
3.6000
3.2000
2.8000
2.4000
TIME (SECONDS)
2.0000
1.6000
1.2000
0.80000
0.40000
0.0
Figure 14-12. Comparison Between CGEN1 With and Without Subtransient Saliency
(Sheet 1 of 2)
14-26
CGEN1 WITH AND WITHOUT

SUBTRANSIENT SALIENCY
08:46
POWER
TECHNOLOGIES,
INC.R
THU MAY 05, 1994

CGEN1 BEHAVIOR
4.0000 FILE: cgen1 1.0000
0.00500 FILE: cgen1 -0.0050
CHNL# 4: [ET CGEN1]
1.2000 FILE: cgen1 0.30000
4.0000 FILE: cgen1sal 1.0000
0.00500 FILE: cgen1sal -0.0050
CHNL# 4: [ET CGEN1]
1.2000 FILE: cgen1sal 0.30000
4.0000
3.6000
3.2000
2.8000
2.4000
TIME (SECONDS)
2.0000
1.6000
1.2000
0.80000
0.40000
0.0
Figure 0-0 (Cont). Comparison Between CGEN1 With and Without Subtransient Saliency (Sheet 2 of 2)
The run times for the various models calibrate the performance of the two models. The
fastest simulation run is obtained using GENROU. When the CGEN1 and TGEN1 model
is used without subtransient saliency, the ratio of CPU time was 1.98:1. With subtransient
saliency the ratio was 2.07:1.
14-27
Models CIMTR1 and CIMTR3 (Induction Generators) Program Application Guide: Volume II
The time step used in the simulation runs was equal to a half cycle for a 60-Hz system. As
mentioned in Section 14.4.5.1, the values for  should be at 1/4 of the smallest short circuit
time constant. To determine the smallest short circuit time constant, one approach is to
calculate the equivalent L/R for each rotor resistance with the stator shorted and the other
rotor resistances set to zero. For example with the model used in the simulation the value
for the time constant would be:
As can be seen from the above, the smallest value is greater than four times the time step
and there should be no numerical problems.
14.5 Models CIMTR1 and CIMTR3 (Induction Generators)

CIMTR1 and CIMTR3 model either a single-cage or double-cage induction generator
including rotor flux dynamics. In the power flow, a generator with positive electrical
power should be used. To model a single-cage machine, either T or X should be set to
zero and ZSORCE in the power flow should be set to X. To model a double-cage
machine, a value must be provided for both T and X, and ZSORCE of the power flow
should be set to X. For this model, an extra CON (labeled SWITCH) is used only for
minimization of computer code and does not affect the model. None of the standard
PSS®E governor models can be used with this model because they assume zero speed
deviation during initialization.
A double-cage machine will have time constants much smaller than T and, therefore, a
time step less than the normal value of 1/2 cycle may be necessary.
14-28
Program Application Guide: Volume II Models CIMTR1 and CIMTR3 (Induction Generators)
CIMTR3 uses one extra state variable but is more accurate than CIMTR1. It is recom-
mended for large frequency deviations and is required for extended term. CIMTR3 also
has an initial CON where the mechanical power is entered at synchronous speed. This
CON is ignored if the machine is in-service in the power flow (i.e., JCODE > 2 and
MCSTAT = 1). The machine can be started by simply placing it in service. Starting the
machine in state space will generally require a very small time step, sometimes in the
order of 1/16 of a cycle. When using the extended algorithms to start the motor, 1/2 cycle
usually works. This model cannot be the only machine in an island when using the con-
stant island frequency mode.
14-29
This page intentionally left blank.
14-30
Chapter 15
Static var Devices
Chapter 15 - Static var Devices
15.1 Principles
A static var generator (SVG or SVS or SVC) is a shunt reactor operating subject to some
form of control to regulate the voltage at a bus. While there is a wide variety of ways of
building an SVG, shunt reactors or special transformers are used in combination with con-
trolled thyristors. Some representative arrangements are shown in Figure 15-1. Figure 15-1a
shows arrangements in which the thyristors are in series with a reactor and control the cur-
rent in it directly. Figure 15-1b shows an alternative arrangement in which the thyristor ele-
ments exert control by altering the degree of saturation in the core of the reactor. In all
arrangements the effect of controlled firing of the thyristors is to control the effective fun-
15-1
Static var Devices PSS®E 33.4
Principles Program Application Guide: Volume II
damental frequency admittance of the thyristor-reactor unit as seen from its high-voltage
terminals.
a. Reactors With Current Controlled Directly by
ac
Supply
Control Winding
b. Reactors With Current Controlled by Saturation Induced by dc Control
Figure 15-1. Control of Reactor Current by Thyristors
The controllable shunt reactor may be used as an alternative to a synchronous condenser

to control voltage on a transmission system. One basic application and its overall charac-
teristic are shown in Figure 15-2. With the reactor turned off, the installation is a shunt
capacitor that will supply reactive power to the system. The reactor may be turned on to
absorb reactive power, giving the flat section of the characteristic until it is fully on. After
the reactor is fully on, the characteristic is that of the reactor, which may saturate exten-
sively at high voltages, in parallel with the capacitor. While the characteristic shown in
Figure 15-2 is achieved by turning the reactor on in proportion to bus voltage, virtually any
static and dynamic characteristic can be obtained within the range offered by the rating of
the reactor and capacitor.
15-2
PSS®E 33.4 Static var Devices
Program Application Guide: Volume II Dynamic Characteristics
a. Controlled Reactor and Capacitor to Provide Controlled var
V
Composite
Characteristic
Controlled
Reactor Characteristic
Characteristic Capacitor
(Without Saturation) Characteristic
Reactor
Characteristic
(With Saturation)
Effective Characteristic
b. Composite Characteristics of Controlled Reactor and Capacitor
Figure 15-2. Static var Source Application
15.2 Dynamic Characteristics
15.2.1 Representation
As with dc transmission terminals, the dynamic behavior of controlled reactors should be
viewed at two distinct levels. The firing and commutation of the thyristors is a broad band-
width process for which the response is very rapid in relation to the bandwidth of simula-
tion in PSS®E. The operating point of this process is set by a control unit where bandwidth
is generally within that of PSS®E. The overall behavior of the controlled reactor is as
shown in Figure 15-3.
The inherent dynamic characteristics, H(s) and Z(s), of the thyristor-reactor unit and the
system interact with one another in a complex and nonlinear way over a broad bandwidth.
The reference to this combination is provided by the control unit C(s), where behavior is
15-3
Dynamic Characteristics Program Application Guide: Volume II
designed over the bandwidth recognized by PSS®E. Because the details of H(s) and Z(s) at
frequencies above 30 r/s are unseen by PSS®E, all controlled reactor models in PSS®E
take the general form shown in Figure 15-4. The overall characteristic of the controlled
reactor is approximated by a single time constant and limits, as appropriate, relating fun-
damental frequency admittance to the output signal of the control unit. The control unit
may be modeled in detail.
Controller Thyristor-Reactor Unit Power System
s
+
v ref + Cs Hs  Z s v

c yr ir
– Control Unit Thyristor- Power System
Controlled Impedance
Reactor
Characteristic
v = Bus voltage
ir = Controlled reactor current
y r = Effective controlled reactor admittance
 c = Reference signal for thyristor firing control
 s = Supplementary input signal
v ref = Reference voltage
Figure 15-3. Controlled Reactor Dynamics
15-4
Program Application Guide: Volume II Dynamic Characteristics
Power Frequency
Controller Thyristor-Reactor Unit Network Solution
s
v ref +  1 z thev vr
Cs Limits --------------- Limits 
c 1 + Ts yr
–
Figure 15-4. Form of Static var System Representation Used in PSS®E
15.2.2 Key Time Constant

Consider a static var device controlled by a pure gain transfer function, as shown in
Figure 15-5.
+ 1 -
v ref  K -------------- y
1 + Ts
–
v
Figure 15-5. Simple Static var Device Control Arrangement
The var flow into this device is:

q = v2y
and, for small perturbations about an operating point, vo, yo:
q = 2voyo v + vo2y
Assume that vo is approximately unity, giving:
q = 2yo v + y
where yo is the admittance of the static var device at the unperturbed operating point. The
controller and time constant give:
- K (ref - v)
y =
1 + Ts
15-5
Static var Device Models Program Application Guide: Volume II
and the power system apparent impedance (or admittance) gives:

- q = Yov2
Taking vo = 1 gives:
- q = 2Yov
where Yo is the effective admittance of the system as seen from the terminals of the static
var device.
Combining the perturbation equations gives:
K vref
v =
K + 2(Yo + yo) + 2(Yo + yo)Ts
The overall time constant of the static device and this simple controller is, then:
2(Yo + yo)T
 =
K + 2(Yo + yo)
This result indicates that caution is needed in simulating static var devices. Representative
application data for a static var device on a moderately weak system might give:
Yo = 5
yo = 1
The approximating time constant, T, is generally quite short, say 0.05 sec. Close control of
voltage requires a high value of K. With K set to 50, the effective time constant would be:
2 × (5 + 1) × 0.05
 =
50 + 2 × (5 + 1)
12
= 0.05 × = 0.01 sec
62
This time constant is too short to be represented in simulations using the usual PSS®E
integration step of 0.00833 sec. Hence, a static var device with a high gain controller may
require that the PSS®E integration step width be reduced, in order to maintain numerical
stability of the simulation.
15.3 Static var Device Models
15.3.1 System Interface

The PSS®E dynamic models require that each SVG be represented by either a switched
shunt or a generator in the power flow case. The recommended approach is to use the
switched shunt in the power flow. The following rules apply to using the switched shunt
model:
1. The switched shunt control mode should be continuous (MODSW = 2).
2. All steps and blocks at that switched shunt are assumed to be controlled by the PSS®E
dynamic model.
15-6
Program Application Guide: Volume II Static var Device Models
Any fixed reactors or capacitors should be modeled independently of the switched shunt.
Voltage at the SVG terminals or remotely controlled bus is obtainable via the standard
CHAN selection of voltage at a bus. Admittance of the device can be obtained by assign-
ment of the appropriate var to a channel.
The following rules apply to the assignment of the generator attributes:
1. The bus at which the SVG is connected must be of Type 2 or 3 and must have a gen-
erator assigned and online to represent the SVG.
2. The generator’s MBASE value must equal the MVA rating of the controlled inductor ele-
ment of the SVG installation.
3. The step-up transformer element of the generator must not be used; its parameters
(ZTRAN and GENTAP) must be zero and unity, respectively.
4. The ZSORCE value of the generator must be set to a large number, such as (0 + j999)
per unit, to assure that the SVG is not a current source in switching solutions.
5. The var limits assigned to the generator must reflect:
a. The effective admittance of the controlled reactor in its full on/off conditions.
b. The admittance of any shunt capacitors that will be represented as a part of the
SVG. (Shunt devices connected independently at the bus do not affect the var
limits associated with the SVG.)
c. The nature of any current and/or MVA limits inherent in the SVG gating controls
and the actual voltage at the bus.
The SVG model CSVGN1 allows shunt capacitance to be represented as a part of the
SVG. Hence, an installation of the form shown in Figure 15-2 could be represented by any
of the following:
either a generator representing the net effect of the inductor and capacitor,
with its var limits set in accordance with their joint admittance.
or a fixed bus-connected shunt capacitor and a generator where var limits
are set in accordance with the operating range of the reactor alone.
or a switched shunt capacitor on the bus and a generator where var limits
are set in accordance with the operating range of the reactor alone.
15-7
Static var Device Models Program Application Guide: Volume II
For SVG models represented by a generator in the power flow, the voltage at the SVG bus,
complex power flow into it, its compensated terminal voltage, and so on, are handled, and
are available for output channels, exactly as if the device were a synchronous machine.
The models, however, represent internal dynamic behavior that is very different from that
of a synchronous machine.
15.3.2 Handling SVGs in Power Flow Solutions

Switched shunts representing SVGs in the power flow do not require any special handling
of boundary conditions, therefore, this representation is recommended.
Generators representing SVGs should be treated with special care in power flow solutions
because the constant-Q reactive power limits used as the standard generator boundary
condition are unlikely to be a true representation of the limits affecting an SVG. If the
SVG is modeled as a reactor-capacitor compendium, for example, its reactor fully off con-
dition would normally produce a reactive power versus voltage boundary condition corre-
sponding to a capacitive admittance. If, on the other hand, all capacitance is represented
by independent bus-connected devices, the reactor fully off limit would match the stan-
dard var limit condition with QMAX = 0. In the maximum reactor load condition, the true
form of the power flow boundary condition depends on the nature of the gating controls of
the SCR unit. If gated wide open, the reactor would present an inductive susceptance
boundary condition. Protective logic in the gating controls may prevent this condition,
however, and produce a constant current or other boundary characteristic.
The SVG model, CSVGN1, does not represent any protective or current-limiting control
logic. It allows the reactor to be gated fully on regardless of the current and voltage on it.
To represent this condition, the generator var limits should ideally be set in accordance
with:
Qmax = (Bc - Boff) voff2 × SBASE
Qmin = (Bc - Bon) von2 × SBASE

where:
Bc = Admittance of capacitor included in SVG installation, system base.
Boff = Effective admittance of controlled inductor when gated fully off, system base.
Bon = Effective admittance of controlled inductor when gated fully on, system base.
voff, von = Voltage at SVG bus when controlled reactor is at off and on limit, respectively.
Because the power flow logic does not handle variable var limits, a pair of reasonable con-
stant var limits must be chosen. These limits should be chosen so that the SVG always
appears in the power flow solution to be operating within its continuous regulation range.
If limited behavior is called for, this practice should have the unit operating within its reg-
ulating range, but just short of the limit. The var limits should then be:
Qmax = (Bc - Boff) vsched2 × SBASE
Qmin = (Bc - Bon) vsched2 × SBASE
15-8
Program Application Guide: Volume II Static var Device Models
If an initial power flow solution results in an SVG loading on the high var limits with volt-
age below schedule, a corrected power flow with SVG output just within limit can usually
be produced by the following process:
1. Using CHNG to reduce the scheduled voltage to a value slightly below the actual value
from the var-limited solution.
2. Adjusting the high var limit to correspond to the new scheduled voltage.
3. Resolving and repeating the procedure as necessary.
15.3.3 Model CSVGN1

The CSVGN1 model represents an SCR-controlled shunt reactor and a parallel connected
capacitor, if present. The SCR gate is controlled by voltage error with an auxiliary signal.
The voltage reference is contained in the VREF array, and the auxiliary signal is obtained
from the VOTHSG array, exactly as if the device was a synchronous machine.
The physical size of the reactor is specified by the base MVA, MBASE, assigned to the
SVG in the power flow data. The size of the capacitor is given by the parameter CBASE
(CON(J+9)) in the model data. This constant is zero if no capacitor is represented by the
model.
The limits VMAX and VMIN specify the active range of the voltage control loop. Normal val-
ues of VMAX and VMIN are unity and zero, respectively, though they could have more
restrictive values if the controls are adjusted for an active range corresponding to only a
part of the reactor’s rating. The minimum admittance limit, RMIN, specifies the effective
admittance of the reactor when the SCR control element is turned completely off. The
value of RMIN would be zero, or very nearly zero, in units where the SCR element controls
main current directly and can turn the reactor completely off and nonzero in elements
where the SCR controls work through a secondary winding and allow the reactor to draw
magnetizing current even in the off condition.
The time constant, T5, allows a rough approximation of delays in the reactor’s response to
control signals. Its value should, in the light of the fast response observed in most tests of
SVG units, be 0.05 sec. or less. The time constants T1 through T4 provide for transient
gain reduction in the control loop and hence allow for a high value, K, of steady-state volt-
age control gain.
15.3.4 Model CSVGN3

This static var model is identical to CSVGN1 except for an extra controller. For this
model, if the voltage magnitude deviates from nominal by Vov per unit, the reactor will be
either gated all the way on or off.
15.3.5 Model CSVGN4

This static var model is identical to CSVGN3 except that it allows the static var source to
try to regulate the voltage on a remote bus. If the remote bus specified is out-of-service
(this includes being disconnected during a simulation) there will be no regulation, i.e., the
reactor will remain at its admittance at the time of bus disconnection. When this occurs, an
15-9
Example Case Program Application Guide: Volume II
appropriate message is printed. If the remote bus cannot be found, the model will try to
hold its own voltage. If a remote bus is specified, its voltage is placed in array ECOMP
and the terminal bus of the static var source is placed in ETERM. Both ECOMP and
ETERM values may be assigned to channels.
15.3.6 Models CSVGN5 and CSVGN6

CSVGN5 and CSVGN6 are static var system models that were written for a correspond-
ing model in the WSCC stability program. The features of these models include a fast
override and remote bus voltage control. For a specified remote bus that is out-of-service
or not in the network, these models use the same rules as CSVGN4. The fast override is
activated when the voltage error exceeds a threshold values during major disturbances
such as faults near the SVC or switching.
The corresponding WSCC model does not separate the equipment to identify capacitor
banks and reactors. To maintain the WSCC model structure and to include frequency
dependence, CSVGN5 assumes that the output is equal to BMAX and the thyristor-con-
trolled reactor is shut off. If BMAX is positive then the capacitor banks are equal to
BMAX times the MVA rating in the power flow from the generator slot. If BMAX is neg-
ative, then the equipment is assumed to consist of just reactors.
CSVGN5 requires nonzero values for the parameters VEMAX, TS3, KSVS, BMAX,
BMAX, BMIN, BMIN, and TS6. The time constants, TS3 and TS6, must be greater than
twice DELT to avoid numerical instability. The other time constants, KSD and DV, may be
set to zero. If DV is set to zero then the threshold limits for the fast override are calculated
as follows:
1. Upper limit = BMAX/KSVS.
2. Lower limit = BMIN/KSVS.
By setting a nonzero value for DV, the limits are then equal plus or minus the value of DV.
CSVGN6 is different from CSVGN5 in that controller limits are nonwindup. Additional
data is specified to have additional shunt equipment switched in-service and a bias is also
provided.
15.3.7 Model CSSCS1

CSSCS1 incorporates all the features of CSVGN1, CSVGN3, and CSVGN4. This model
is to be used when the static var system is modeled as a switched shunt in the continuous
mode in the power flow. The CSSCS1 model assumes that the full range of the switched
shunt is continuously controllable.
15.4 Example Case

Figure 15-6 shows a small power system model for use in illustrating the application of
model CSVGN1. This system model corresponds to a long distance transmission scheme
of about 300 miles at 500 kV. The receiving system has only a very small amount of local
generation and cannot meet reactive power demands at high load without additional reac-
15-10
Program Application Guide: Volume II Example Case
tive sources. These sources must be controlled in order to maintain proper management of
voltage during light load operation and hence must be synchronous condensers or SVGs.
Figure 15-7 shows the base case power flow data and solution for dynamic simulations.
Type 2 buses 4 and 5 represent the tertiary terminals of the receiving transformers, each
has a generator connected to represent an SVG. Each SVG consists of a 500-Mvar SCR-
controlled reactor with a large capacitor connected in parallel with it. The capacitors do
not appear in the power flow case because they are treated as an internal part of the SVG
and will be so modeled by CSVGN1. The scheduled voltage at buses 4 and 5 is 1.0 pu.
The high var limit is set at 500 Mvar, which is the maximum output to be allowed from the
SVG in steady-state conditions, the size of the capacitor bank is not defined. If the capaci-
tor should have a rating of 500 Mvar, the var limit setting would be interpreted as a fixed
approximation to a variable limit. In this case, however, expect to use a capacitor rating
greater than 500 Mvar in heavy load studies; the limit is simply a convenience to ensure
that the SVG is initialized well within its dynamic range. The low var limits are set at zero
because they are not expected to be active in heavy load cases and the shunt capacitor rat-
ing has not yet been selected.
SVG 500 MW
Steam
HYDRO
Load
2700 MW 500-kV 300 MW
Transmission pf = 0.92
(300 miles)
SVG
SVG 500 MVA

4
~ 500 MVA
14
~
15
3000 MVA
1
2 3
SVG 500 MVA

5
Figure 15-6. Sample System for Application of Model CSVGN1

to Represent Static var Generator
15-11
Generator buses for SVGs

PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E WED DEC 06, 1995 12:04
LONG DISTANCE RADIAL TRANSMISSION WITH BUS DATA
STATIC VAR GENERATOR VOLTAGE SUPPORT
BUS# NAME BSKV CODE LOADS VOLT ANGLE S H U N T AREA ZONE
1 HYDRO 500.00 3 0 1.0250 0.0 0.0 0.0 1 1
2 RECEIVE 500.00 1 0 0.9167 -41.7 0.0 0.0 1 1
3 LOAD 230.00 -2 1 0.9643 -46.1 0.0 500.0 2 1
4 TERT1 33.000 2 0 1.0000 -40.9 0.0 0.0 2 1
5 TERT2 33.000 2 0 1.0000 -40.9 0.0 0.0 2 1
14 TRMID1 1 0 0.9908 -40.9 0.0 0.0 2 1
15 TRMDI2 1 0 0.9908 -40.9 0.0 0.0 2 1
--------------------------------------------------------------------------------
LONG DISTANCE RADIAL TRANSMISSION WITH LOAD DATA
BUS# NAME BSKV ID COD ST MVA-LOAD CUR-LOAD Y - LOAD AREA ZONE
3 LOAD 230 1 -2 1 3000.0 1300.0 0.0 0.0 0.0 0.0 2 1
--------------------------------------------------------------------------------
LONG DISTANCE RADIAL TRANSMISSION WITH GENERATING
STATIC VAR GENERATOR VOLTAGE SUPPORT PLANT DATA
BUS# NAME BSKV COD MCNS PGEN QGEN QMAX QMIN VSCHED VACT. REMOT PCT Q
1 HYDRO 500 3 1 2690.2 1029.6 1500.0 -500.0 1.0250 1.0250
3 LOAD 230 -2 1 450.0 200.0 200.0 0.0 1.0000 0.9643
4 TERT1 33.0 2 1 0.0 461.6 500.0 0.0 1.0000 1.0000 Generators
5 TERT2 33.0 2 1 0.0 461.6 500.0 0.0 1.0000 1.0000
representing
-------------------------------------------------------------------------------- SGVs
LONG DISTANCE RADIAL TRANSMISSION WITH GENERATOR
STATIC VAR GENERATOR VOLTAGE SUPPORT UNIT DATA
BUS# NAME BSKV COD ID ST PGEN QGEN QMAX QMIN PMAX PMIN MBASE Z S O R C E X T R A
N GENTAP
1 HYDRO 500 3 1 1 2690 1030 1500 -500 9999-9999 3000 0.0000 0.2000 0.0000
0.1500 1.0250
3 LOAD 230 -2 1 1 450 200 200 0 9999-9999 500 0.0000 0.2000 0.0000
0.1500 1.0125
4 TERT1 33.0 2 1 1 0 462 500 0 9999-9999 500 0.0000 1.0000
5 TERT2 33.0 2 1 1 0 462 500 0 9999-9999 500 0.0000 1.0000
--------------------------------------------------------------------------------
MBASE equal
to SVG reactor
LONG DISTANCE RADIAL TRANSMISSION WITH BRANCH DATA rating
1* 2 1 HYDRO RECEIVE 0.0050 0.0720 5.5000 1 2000 1000 2500
1* 2 2 HYDRO RECEIVE 0.0050 0.0720 5.5000 1 2000 1000 2500
1* 2 3 HYDRO RECEIVE 0.0050 0.0720 5.5000 1 2000 1000 2500
2* 14 1 RECEIVE TRMID1 0.0000 -0.0010 0.0000 F 1 1500 2000 2500
2* 15 1 RECEIVE TRMDI2 0.0000 -0.0010 0.0000 F 1 1500 2000 2500
3 14* 1 LOAD TRMID1 0.0000 0.0068 0.0000 T 1 1500 2000 2500
3 15* 1 LOAD TRMDI2 0.0000 0.0068 0.0000 T 1 1500 2000 2500
4 14* 1 TERT1 TRMID1 0.0000 0.0020 0.0000 1 0 0 0
5 15* 1 TERT2 TRMDI2 0.0000 0.0020 0.0000 1 0 0 0
--------------------------------------------------------------------------------
LONG DISTANCE RADIAL TRANSMISSION WITH TRANSFORMER DATA
FROM TO CKT TP RATIO ANGLE RG CONT RMAX RMIN VMAX VMIN STEP TABLE
CR CX
2 14 1 F 0.9250 0.00 1 3 1.0750 0.9250 1.0000 0.9800 0.00625
2 15 1 F 0.9250 0.00 1 3 1.0750 0.9250 1.0000 0.9800 0.00625
3 14 1 T 1.0000 0.00 1 0 1.5000 0.5100 1.5000 0.5100 0.00625
3 15 1 T 1.0000 0.00 1 0 1.5000 0.5100 1.5000 0.5100 0.00625
--------------------------------------------------------------------------------
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E WED DEC 06, 1995 12:04 Line-connected
LONG DISTANCE RADIAL TRANSMISSION WITH
LINE SHUNT DATA shunt reactors
X------FROM------X X-------TO-------X CKT LINE G,B (FROM) LINE G,B (TO) ST at sending end
1 HYDRO 500 2 RECEIVE 500 1 0.0000 -2.0000 0.0000 0.0000 1
1 HYDRO 500 2 RECEIVE 500 2 0.0000 -2.0000 0.0000 0.0000 1
1 HYDRO 500 2 RECEIVE 500 3 0.0000 -2.0000 0.0000 0.0000 1
Figure 15-7. Power Flow Data for CSVGN1 Application Example
15-12
Figure 15-8 shows the initial condition power flow solution with the SVGs appearing as
conventional generators at buses 4 and 5. This power flow case is ready for conversion to
a dynamics initial condition with activities CONG and CONL in the conventional manner.
Figure 15-9 shows the main elements of the dynamic simulation setup. The CSVGN1 data
is entered in appropriate records in the Dynamics Data File in the same way as for a gener-
ator. DYRE creates calls in both CONEC and CONET but, apart from this, the SVG
model is linked to the power flow case in the same way as the conventional generator
model, GENROU. The SVG reactive power output and terminal voltage are available in
the generator condition arrays, QGEN, ETERM, etc., and can be assigned directly to out-
put channels. CSVGN1 incorporates both voltage regulator and main SVG modeling, and
uses arrays VREF and VOTHSG as appropriate. The CSVGN1 leaves the values of PGEN
and EFD at zero, because it is only a reactive power device and uses no excitation voltage.
Figure 15-10 shows the initialization of a dynamic simulation run. The two SVGs at buses
4 and 5 are handled as generators in STRT and their terminal reactive flow and voltage are
displayed as channels by selecting the corresponding generator QGEN and ETERM
arrays.
Figure 15-11 shows simulation results obtained with the setup shown in Figure 15-10. A sin-
gle line-to-ground fault was applied at the receiving 500-kV bus and cleared after three
cycles by opening one of the 500-kV lines. The circuit was reclosed at t = 2.0.
Figure 15-11 also shows system response when the capacitor element of the SVG is sized at
500 MVA and the gain of the voltage regulator gain (CON(J)) is set at 100. The reactor is
nearly off initially and remains so throughout the period when the third 500-kV line is out-
of-service. The combined control action of the SVGs and the voltage regulator of the
receiving-end generator is not able to restore voltage to nominal while the line is opened.
On reclosing the line, the reactor element of the SVG is turned on in response to the volt-
age rise and the load bus voltage is restored to nominal value in 5 sec.
Figure 15-12 shows the system response when the capacitor elements of the SVGs are sized
at 800 Mvar. Here the voltage falls to a somewhat lower instantaneous minimum while the
500-kV line is opened, but is restored much more rapidly to nominal value when it is
reclosed.
15-13
R SYSTEM SIMULATOR--PSS®E WED DEC 06, 1995 12

SMISSION WITH RAT
AGE SUPPORT SE
EA CKT MW MVAR MVA %I 1.0250PU 0.00

1 2690.2 1029.6R 2880.5 96 512.50KV
1 1 896.7 343.2 960.2 47
1 2 896.7 343.2 960.2 47
1 3 896.7 343.2 960.2 47
EA CKT MW MVAR MVA %I 0.9167PU -41.66

1 458.33KV
1 1 -850.0 20.1 850.2 46
1 2 -850.0 20.1 850.2 46
1 3 -850.0 20.1 850.2 46
2 1 1275.0 -30.1 1275.4 93 0.9250LO
2 1 1275.0 -30.1 1275.4 93 0.9250LO

2 450.0 200.0H 492.4 98 221.79KV
3000.0 1300.0 3269.6
0.0 -464.9 464.9
2 1 -1275.0 -317.5 1313.9 91 1.0000UN
2 1 -1275.0 -317.5 1313.9 91 1.0000UN

2 0.0 461.6R 461.6 92 33.000KV
2 1 0.0 461.6 461.6

2 0.0 461.6R 461.6 92 33.000KV
2 1 0.0 461.6 461.6

2 KV
1 1 -1275.0 13.5 1275.1 86 0.9250UN
2 1 1275.0 443.8 1350.0 91 1.0000LK
2 1 0.0 -457.3 457.3

2 KV
1 1 -1275.0 13.5 1275.1 86 0.9250UN
2 1 1275.0 443.8 1350.0 91 1.0000LK
2 1 0.0 -457.3 457.3
---------------------------------------------------
R SYSTEM SIMULATOR--PSS®E WED DEC 06, 1995 12

SMISSION WITH
AGE SUPPORT
MW MVAR ETERM CURRENT PF MVABASE X T R

0.2 1444.5 1.0600 2880.5 0.8810 3000.0 0.0000 0
0.0 280.2 1.0252 517.1 0.8489 500.0 0.0000 0
0.0 461.6 1.0000 461.6 0.0000 500.0
0.0 461.6 1.0000 461.6 0.0000 500.0
0.2 2647.8 4500.0
Figure 15-8. Base Case for CSVGN1 Application Example
15-14
CSVGN1
Static Shunt Compensator
This device is located at system bus # 4 IBUS,

machine #_____ I,
This model uses CONs starting with #_____ J,
and STATEs starting with #_____ K,
and VAR #_____ L,
and ICON #_____ M.
The reactor Mvar base = 500 MBASE.

J 100 K
J+1 1 T1
J+2 0 T2
J+3 10 T3 (>0)
J+4 0 T4
J+5 0.05 T5
J+6 0. RMIN (reactor minimum Mvar)
J+7 1. VMAX
J+8 0. VMIN
J+9 500. CBASE (capacitor Mvar)
STATEs # Description
K First regulator
K+1 Second regulator
K+2 Thyristor
VARs # Description
L Y (model output)
ICON # Description
M Memory
IBUS, ’CSVGN1’, I, K, T1, T2, T3, T4, T5, RMIN, VMAX, VMIN, CBASE/
15-15
VREF VMAX MBASE/SBASE CBASE/SBASE

1.
– +
+ K(1 + sT1) (1 + sT2) 1
|V|  X –  Y
(1 + sT3) (1 + sT4) 1 + sT5
VMIN RMIN/RBASE
Other
Signals
VOTHSG
1 ’GENSAL’ 1 5 .05 .06 4 1 1.8 1.6 .5 .2 .1 .03 .05/

3 ’GENROU’ 1 6 .05 1.5 .05 3 1 1.5 1.45 .4 .6 .2 .1 .03 .3/
4 ’CSVGN1’ 1 100 1 0 10 0 .05 0 1. 0. 500./
5 ’CSVGN1’ 1 100 1 0 10 0 .05 0 1. 0. 500./
1 ’SEXS’ 1 .1 10 100 .1 0 3/ Note machine numbers for
3 ’SEXS’ 1 .1 10 100 .1 0 3/ CSVGN1 as for generators
Figure 15-9. Dynamic Simulation Setup Main Elements
Executing activity STRT
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E TUE DEC 12, 1995 11:45

1 HYDRO 500 1 1.0600 2.4913 2690.26 1029.57 0.9339 44.43 0.8729 0.4001
3 LOAD 230 1 1.0252 2.4828 450.01 199.99 0.9138 -5.58 0.9334 0.4451
4 TERT1 33.0 1 1.0000 0.0000 0.00 461.57 0.0000 0.00 0.6976-0.6046 SVGs are standard
5 TERT2 33.0 1 1.0000 0.0000 0.00 461.57 0.0000 0.00 0.6976-0.6046 synchronous machine
INITIAL CONDITIONS CHECK O.K. in STRT
TIME X- VALUE --X X--------- IDENTIFIER ---------X X- VALUE --X X--------- IDENTIFIER ---------X
-0.0167 0.91667 VOLT 2 [RECEIVE 500.00] 44.435 ANGL 1 [HYDRO 500.00] MC 1
3 -5.5797 ANGL 3 [LOAD 230.00] MC 1 26.903 POWR 1 [HYDRO 500.00] MC 1
5 4.5001 POWR 3 [LOAD 230.00] MC 1 10.296 VARS 1 [HYDRO 500.00] MC 1
7 1.9999 VARS 3 [LOAD 230.00] MC 1 4.6157 VARS 4 [TERT1 33.000] MC 1
9 4.6157 VARS 5 [TERT2 33.000] MC 1 1.0600 ETRM 1 [HYDRO 500.00] MC 1
11 1.0252 ETRM 3 [LOAD 230.00] MC 1 1.0000 ETRM 4 [TERT1 33.000] MC 1
13 1.0000 ETRM 5 [TERT2 33.000] MC 1 2.4913 EFD 1 [HYDRO 500.00] MC 1
15 2.4828 EFD 3 [LOAD 230.00] MC 1 0.00000 SPD 1 [HYDRO 500.00] MC 1
17 0.00000 SPD 3 [LOAD 230.00] MC 1
-0.0083 0.91667 44.435 -5.5797 26.903 4.5001 10.296

7 1.9999 4.6158 4.6158 1.0600 1.0252 1.0000
13 1.0000 2.4913 2.4828 0.18626E-09 -0.17385E-08
Figure 15-10. Initialization of Dynamic Simulation Run Using CSVGN1
15-16
11:12
POWER
TECHNOLOGIES,
INC.R
TUE DEC 12, 1995

FILE: gp
CHNL# 5: [POWR 3 [LOAD 230.00] MC 1]

8.0000 -2.000
CHNL# 11: [ETRM 3 [LOAD 230.00] MC 1]
1.2000 0.20000
CHNL# 8: [VARS 4 [TERT1 33.000] MC 1]
12.000 2.0000
5.0000
4.5000
Voltage at Load Bus
4.0000
Power Output of
Receiving End
3.5000
SVG Net VAR Output
Generator
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-11. System Response to Fault at Bus 2

Opening of One 500-kV Line and Line Reclosure, 500-Mvar Capacitor
15-17
11:26
POWER
TECHNOLOGIES,
INC.R
FILE: gp2
TUE DEC 12, 1995

CHNL# 5: [POWR 3 [LOAD 230.00] MC 1]
8.0000 -2.000
CHNL# 11: [ETRM 3 [LOAD 230.00] MC 1]
1.2000 0.20000
CHNL# 8: [VARS 4 [TERT1 33.000] MC 1]
12.000 2.0000
5.0000
4.5000
4.0000
3.5000
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-12. System Response to Fault at Bus 2

as Figure 15-11 but with 800-Mvar Capacitors in SVGs
15-18
15.4.1 Example Case 2

PSS®E 33.4
A second example case was set up to illustrate the use of the SVC model CSVGN5 and the supplementary signal model STBSVC.
The electrical network is shown in Figure 15-13, which represents a large hydro plant at the sending end of a long EHV transmission
system. SVCs are placed along the EHV transmission to provide voltage support at buses 2, 3, and 4. The receiving system is repre-
sented by buses 5 and 15 with local steam generation at bus 5. The range for the SVCs is from -300 Mvar to 600 Mvar. The solved
power flow case is shown in Figure 15-14.
The dynamic equipment data obtained from activity DOCU are listed in Figure 15-15. For the SVCs, the MVA rating is equal to the
SVC’s maximum output, making BMAX equal to 1 pu. Also, each SVC is controlling the high-side voltage of the transformer to
which it is connected. A supplementary control is included for the SVC at bus 2. The input signal is derived from the electrical
power of the hydro plant. The disturbance for the simulation runs was a line-to-ground fault at bus 11 using an admittance of –
j10,000 MVA. The fault was cleared in six cycles with the outage of one line from bus 11 to bus 12.
Changing the model status with activity DYCH model STBSVC was disabled so that no supplementary signals were produced. The
plots for the first simulation run are shown in Figures 15-16 through 15-18. Plots for the simulation supplementary control for the
SVC at bus 2 are shown in Figures 15-19 through 15-21. The plot of the rotor angle difference between the hydro unit and the steam
unit is shown in Figure 15-22 for the two runs. Employment of supplementary control increased the damping even though the system
15-19
is well damped without the device.
Static var Devices

Example Case
Figure 15-13. One-Line Impedance Diagram
THYRISTOR CONTROLLED REACTOR EXAMPLE SYSTEM
THU DEC 21, 1995 08:25
BASE CASE LOADFLOW
15-20
ISTOR CONTROLLED REACTOR EXAMPLE SYSTEM

CASE LOADFLOW
DEC 21, 1995 08:27
Figure 15-14. Solved Power Flow Case
15-21
1 PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E THU DEC 21, 1995

09:45
BASE CASE LOADFLOW
REPORT FOR ALL MODELS AT ALL BUSES BUS 1 [HYDRO 13.8] PLANT
MODELS
** GENSAL ** BUS NAME BSKV MACH C O N ' S STATE'S

1 HYDRO 13.8 1 1- 12 1- 5
4000.0 0.00000+J 0.18000 0.00000+J 0.00000 1.00000
T'D0 T''D0 T''Q0 H DAMP XD XQ X'D X''D XL

8.000 0.050 0.100 4.00 0.00 0.8000 0.5000 0.3000 0.1800 0.1200
S(1.0) S(1.2)
0.1000 0.4000
** STAB1 ** BUS NAME BSKV MACH C O N ' S STATE'S

1 HYDRO 13.8 1 69- 75 24- 26
K/T T T1/T3 T3 T2/T4 T4 LIMIT
20.000 3.000 5.000 0.050 5.000 0.050 0.100

1 HYDRO 13.8 1 95- 100 35- 36

0.300 2.500 200.0 0.050 -3.20 4.00
a. Bus 1
Figure 15-15. Dynamic Equipment Data

(Sheet 1 of 5)
1 PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E THU DEC 21, 1995 09:45
BASE CASE LOADFLOW
REPORT FOR ALL MODELS AT ALL BUSES BUS 2 [SVC-1 13.8] PLANT MODELS
** CSVGN5 ** BUS NAME BSKV MACH C O N ' S STATE'S VAR I C O N'S

2 SVC-1 13.8 1 13- 26 6- 9 1 1- 2
MBASE REMOTE BUS TS1 VEMAX TS2 TS3 TS4 TS5

600.0 12 0.000 0.150 0.100 5.000 0.000 0.000
KSVS KSD BMAX B'MAX B'MIN BMIN TS6 DV

400.0 0.0 1.000 1.000 -0.500 -0.500 0.050 0.150
** STBSVC ** BUS NAME BSKV MC C O N ' S STATE'S VAR I C O N'S

2 SVC-1 13.8 1 76- 87 27- 31 4 7- 12
I1 I2 BUS#1 TO BUS ID BUS#2 KS1 TS7 TS8 TS9 TS13

1 0 1 0 '1 ' 0 1.0000 0.0000 0.0000 0.1000 1.0000
TS14 KS3 VSCSMAX KS2 TS10 TS11 TS12

4.0000 2.0000 0.1000 0.0000 0.0000 0.0000 0.0000
b. Bus 2
Figure 0-0 (Cont). Dynamic Equipment Data

(Sheet 2 of 5)
15-22
BASE CASE LOADFLOW

3 SVC-2 13.8 1 27- 40 10- 13 2 3- 4

600.0 13 0.000 0.150 0.100 5.000 0.000 0.000

400.0 0.0 1.000 1.000 -0.500 -0.500 0.050 0.150
c. Bus 3

(Sheet 3 of 5)
BASE CASE LOADFLOW

4 SVC-3 13.8 1 41- 54 14- 17 3 5- 6

600.0 14 0.000 0.150 0.100 5.000 0.000 0.000

400.0 0.0 1.000 1.000 -0.500 -0.500 0.050 0.150
d. Bus 4

(Sheet 4 of 5)
15-23
BASE CASE LOADFLOW
REPORT FOR ALL MODELS AT ALL BUSES BUS 5 [STEAM 13.8] PLANT MODELS

5 STEAM 13.8 1 55- 68 18- 23
4000.0 0.00000+J 0.34000 0.00000+J 0.00000 1.00000

5.60 0.060 0.63 0.060 4.00 0.00 1.6000 1.5500 0.4900 0.6500 0.3400 0.2000
S(1.0) S(1.2)
0.1000 0.4000
** STAB1 ** BUS NAME BSKV MACH C O N ' S STATE'S

5 STEAM 13.8 1 88- 94 32- 34
K/T T T1/T3 T3 T2/T4 T4 LIMIT

20.000 3.000 5.000 0.050 5.000 0.050 0.100

5 STEAM 13.8 1 101- 106 37- 38

0.300 2.500 200.0 0.050 -3.20 4.00
e. Bus 5

(Sheet 5 of 5)
15-24
THU DEC 21, 1995 09:48

BASE CASE LOADFLOW
POWER
TECHNOLOGIES,
INC.R FILE: out.svc
HYDRO UNIT
CHNL# 13: [AUX 1 [HYDRO 13.800] MC 1]
0.50000 -0.5000
CHNL# 11: [SPD 1 [HYDRO 13.800] MC 1]
0.05000 -0.0500
CHNL# 9: [EFD 1 [HYDRO 13.800] MC 1]
5.0000 -5.000
CHNL# 6: [ETRM 1 [HYDRO 13.800] MC 1]
1.2000 0.70000
CHNL# 3: [VARS 1 [HYDRO 13.800] MC 1]
25.000 -25.00
CHNL# 1: [POWR 1 [HYDRO 13.800] MC 1]
50.000 0.0
5.0000
4.5000
4.0000
3.5000
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-16. L-G Fault, No Supplementary Control, Hydro Unit
15-25
THU DEC 21, 1995 09:50

BASE CASE LOADFLOW
POWER
TECHNOLOGIES,
INC.R FILE: out.svc
STEAM UNIT
CHNL# 14: [AUX 5 [STEAM 13.800] MC 1]
0.50000 -0.5000
CHNL# 12: [SPD 5 [STEAM 13.800] MC 1]
0.05000 -0.0500
CHNL# 10: [EFD 5 [STEAM 13.800] MC 1]
5.0000 -5.000
CHNL# 7: [ETRM 5 [STEAM 13.800] MC 1]
1.2000 0.70000
CHNL# 4: [VARS 5 [STEAM 13.800] MC 1]
12.500 -12.50
CHNL# 2: [POWR 5 [STEAM 13.800] MC 1]
25.000 0.0
5.0000
4.5000
4.0000
3.5000
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-17. L-G Fault, No Supplementary Control, Steam Unit
15-26
THU DEC 21, 1995 09:52

BASE CASE LOADFLOW
POWER
TECHNOLOGIES,
INC.R
FILE: out.svc
SVC-BUS 12
CHNL# 15: [AUX 2 [SVC-1 13.800] MC 1]
0.50000 -0.5000
CHNL# 8: [ETRM 2 [SVC-1 13.800] MC 1]
1.2000 0.70000
CHNL# 5: [VARS 2 [SVC-1 13.800] MC 1]
8.0000 -2.000
5.0000
4.5000
4.0000
3.5000
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-18. L-G Fault, No Supplementary Control, SVC-Bus 12
15-27
THU DEC 21, 1995 09:56

BASE CASE LOADFLOW
POWER
TECHNOLOGIES,
INC.R FILE: out.svc.2
HYDRO UNIT
CHNL# 13: [AUX 1 [HYDRO 13.800] MC 1]
0.50000 -0.5000
CHNL# 11: [SPD 1 [HYDRO 13.800] MC 1]
0.05000 -0.0500
CHNL# 9: [EFD 1 [HYDRO 13.800] MC 1]
5.0000 -5.000
CHNL# 6: [ETRM 1 [HYDRO 13.800] MC 1]
1.2000 0.70000
CHNL# 3: [VARS 1 [HYDRO 13.800] MC 1]
25.000 -25.00
CHNL# 1: [POWR 1 [HYDRO 13.800] MC 1]
50.000 0.0
5.0000
4.5000
4.0000
3.5000
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-19. L-G Fault, Supplementary Control, Hydro Unit
15-28
THU DEC 21, 1995 09:58

BASE CASE LOADFLOW
POWER
TECHNOLOGIES,
STEAM UNIT
CHNL# 14: [AUX 5 [STEAM 13.800] MC 1]
0.50000 -0.5000
CHNL# 12: [SPD 5 [STEAM 13.800] MC 1]
0.05000 -0.0500
CHNL# 10: [EFD 5 [STEAM 13.800] MC 1]
5.0000 -5.000
CHNL# 7: [ETRM 5 [STEAM 13.800] MC 1]
1.2000 0.70000
CHNL# 4: [VARS 5 [STEAM 13.800] MC 1]
12.500 -12.50
CHNL# 2: [POWR 5 [STEAM 13.800] MC 1]
25.000 0.0
5.0000
4.5000
4.0000
3.5000
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-20. L-G Fault, Supplementary Control, Steam Unit
15-29
THU DEC 21, 1995 09:59

BASE CASE LOADFLOW
POWER
TECHNOLOGIES,
SVC-BUS 12
CHNL# 15: [AUX 2 [SVC-1 13.800] MC 1]
0.50000 -0.5000
CHNL# 8: [ETRM 2 [SVC-1 13.800] MC 1]
1.2000 0.70000
CHNL# 5: [VARS 2 [SVC-1 13.800] MC 1]
8.0000 -2.000
5.0000
4.5000
4.0000
3.5000
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-21. L-G Fault, Supplementary Control, Steam Unit
15-30
THU DEC 21, 1995 09:40

BASE CASE LOADFLOW
POWER
TECHNOLOGIES,
INC.R
ROTOR ANGLES
CHNL#’S 16,17: [ANGL 1 [HYDRO 13.800] MC 1]-[ANGL 5 [STEAM 13.800] M
75.000 FILE: out.svc.2 25.000
CHNL#’S 16,17: [ANGL 1 [HYDRO 13.800] MC 1]-[ANGL 5 [STEAM 13.800] M
5.0000
75.000 FILE: out.svc 25.000
With Supplementary Control
4.5000
4.0000
3.5000
Without Supplementary Control
3.0000
TIME(SECONDS)
2.5000
2.0000
1.5000
1.0000
0.50000
0.0
Figure 15-22. Rotor Angle Difference for the Two Runs
15-31
CSVGN1
Static Shunt Compensator
This device is located at system bus #______ IBUS,

machine #______ I,
This model uses CONs starting with #______ J,
and STATEs starting with #______ K,
and VAR #______ L,
and ICON #______ M.
The reactor Mvar base = ____________ MBASE.

J K
J+1 T1
J+2 T2
J+3 T3 (>0)
J+4 T4
J+5 T5
J+6 RMIN (reactor minimum Mvar)
J+7 VMAX
J+8 VMIN
J+9 CBASE (capacitor Mvar)
K First regulator
K+1 Second regulator
K+2 Thyristor
VARs # Description
L Y (model output)
ICON # Description
M Memory
IBUS, ’CSVGN1’, I, K, T1, T2, T3, T4, T5, RMIN, VMAX, VMIN, CBASE/
15-32
Program Application Guide: Volume II Control Tuning of an SVG
VREF VMAX MBASE/SBASE CBASE/SBASE

1.
– +
+ K(1 + sT1) (1 + sT2) 1
|V|  X –  Y
(1 + sT3) (1 + sT4) 1 + sT5
VMIN RMIN/RBASE
Other
Signals
VOTHSG
Figure 15-23. PSS®E Model of CSVGN1
15.5 Control Tuning of an SVG

For this example, assume that the SVG is as shown in Figure 15-2, a controlled reactor in
parallel with a fixed capacitor bank. The SVG control function is shown in Figure 15-23.
The approach used to tune the SVS is, after it is understood, quite simple and straightfor-
ward.
1. First, determine the change in voltage for a change in SVG admittance (de/dy) for the
weakest system configuration that is likely to occur, as this will present the severest
control problem. The weakest system results in the greatest change (highest de/dy).
Represent generators as in the dynamic runs, and solve the power flow with and
without an admittance. For example, assume a 100-Mvar reactor causes a voltage
change of about 5.6% or de/dy = 0.056 on 100-MVA base.
2. Choose the SVG controller steady-state gain, which will normally be in the range of 100
to 200. Use a gain of 150, which means a 0.66% error in voltage will drive the SVS to
ceiling.
3. Assume the SVG time constant of 0.03 sec. Modern SVG controls are very fast and
this small time constant should be realizable by standard equipment.
4. Although the SVG block diagram shows a double lead-lag, which can be used to
reduce the high steady-state gain to get stable operation at the crossover frequency,
use only one of the lags (T3). The Bode plot of the controller is shown in Figure 15-24.
15-33
Control Tuning of an SVG Program Application Guide: Volume II
de
150 ------
dy
Magnitude
1.0

1- c 1
----- ---------- = 33
T3 0.03
Figure 15-24. Bode Plot of Simplified SVG Controller
15-34
Program Application Guide: Volume II Control Tuning of an SVG
The total gain for the controller and process combination is 150 de/dy. The crossover fre-
quency (c) is chosen to be safely away from the second break at 33 rad/sec (e.g., 14
rad/sec). The phase angle at crossover is approximately:
c = 90 + tan-1(0.03)(14) = 90 + 23 = 113º
This would result in an acceptable response, exhibiting reasonably fast control with a
small overshoot. The phase angle at crossover should normally be less than 120°. Phase
angles much less than 110° would mean response is more sluggish than necessary. The
magnitude at crossover determines T3.
de
150
dy
Magnitude = = 1.0
T3(14)
Therefore,
de
150
dy
T3 = sec
14
de
where is on an SVS base.
dy
de
For a 100-Mvar SVS, = 0.056 on an SVS base, and T3 = 0.6 sec.
dy
de
For a 150-Mvar SVS, = 0.084 on an SVS base, and T3 = 0.9 sec.
dy
de
For a 200-Mvar SVS, = 0.112 on an SVS base, T3 = 1.2 sec.
dy
15-35
Control Tuning of an SVG Program Application Guide: Volume II
15-36
Chapter 16
Excitation System and
Controller Models
Chapter 16 - Excitation System and Controller Models
16.1 Overview
The basic approaches to the excitation of large generators are shown schematically in
Figure 16-1.The excitation system in all cases consists of a high power source of direct cur-
rent, an intermediate power level controlling circuit, and an instrument power level volt-
age regulator. The voltage regulator determines the manipulation of the exciter, but its
dynamics are often less influential than the nonlinear characteristics of the excitation
power source in determining overall dynamic behavior of the system. The proper repre-
sentation of excitation systems, therefore, requires careful consideration of both the gains
and time constants assigned to the voltage regulators and of the characteristics of the exci-
tation power components.
16.2 The dc Exciters

The common excitation system arrangement on older large generators uses a shaft-driven
dc generator as the exciter. The dc exciter may be either separately excited, as shown in
Figure 16-2a, or shunt excited as in Figure 16-2b. In either case, the IEEE- recommended
excitation system models represent the exciter by the block diagram and nonlinear func-
tion, Se, defined in Figure 16-3.
The saturation function, Se, is a design characteristic of the exciter and is specified to the
computer program by the two values Se(E1) and Se(E2). The curve of Se versus Efd is
assumed by the present PSS®E models to have the form:
B(Efd - A)2
Se =
Efd
and to pass through the two points defined in Figure 16-3. The field voltage values, E1 and
E2, for which the saturation function is specified, should normally be chosen near the knee
of the exciter’s magnetization curve and near the excitation ceiling, respectively.
Because the saturation curve is a design characteristic of the exciter, the parameters Se(E1)
and Se(E2) are constants. The exciter parameters, Te and Ke, are not necessarily constant,
however.
16-1
Excitation System and Controller Models PSS®E 33.4
The dc Exciters Program Application Guide: Volume II
In the case of shunt excited exciters, these parameters may have to be adjusted in accor-
dance with the initial conditions of each simulation run, as outlined in Section 16.2.1.
v ref
+
Control
e term –  Gs Power
Source
Voltage Exciter
Regulator Field Exciter Main
Generator
a. Rotating dc Exciter
v ref
+
Rectifier
–
e term  Gs Bridge
Voltage
Regulator
Excitation
Control
Source
Rotating
ac Exciter
b. Rotating ac
v ref
+
e term –  Gs
Rectifier Bridge
Firing Control
Voltage Circuit
Regulator PT
CT
Excitation
Power
Controlled Transformers
Rectifier
Bridge
c. Excitation Power Fed from Generator Terminals
Figure 16-1. Excitation Power Source Alternatives
16-2
PSS®E 33.4 Excitation System and Controller Models
Program Application Guide: Volume II The dc Exciters
Field
Resistor, R
Ix
• •
Control Main Field

Voltage, Vc Voltage,
E fd
• •
Field Inductance, L
a. Separately Excited Exciter
Ix
• • •
Control
Voltage (Vc)
Main Field
Voltage (Efd)
•
Field
Resistor, R • •
Field Inductance (L 
b. Shunt-Excited Exciter
Figure 16-2. Separately and Shunt Excited Arrangements for Main Exciters
16-3
+ 1
Ic  -------- E fd
sT e
–
Ke + Se
a. Rotating Exciter Block
Se Se(E2)
Se(E1)
E1 E2 Efd
b. Saturation Factor, Se
E fd
E2
E1
I-
S e = -------
E fd
I 1 I 2 Ix
Exciter Field Current (pu)
c. Relation of Saturation Factor to Exciter Magnetization Curve
Figure 16-3. Nonlinear Transfer Function Model for Rotating Exciter
16-4
16.2.1 Separately Excited Exciter

The separately excited exciter, shown in Figure 16-2a, uses per unit characterization of
variables; therefore, the output voltage, Efd, is related to exciter field current, Ix, by the
open-circuit characteristic shown in Figure 16-3c. This characterization may be expressed
by:
Efd = (Ix - I) (1.1)
and,
I
= Se
Efd
(1.2)
so that,
Efd (1 + Se) = Ix (1.3)
The exciter field current and the output, vc, of the control source are related by:
dIx
RIx + L’ dt = vc
(1.4)
where L is the saturated value of the exciter field inductance. This inductance, L may be
taken for practical purposes as:
L
L’ =
1 + Se
(1.5)
where L is the unsaturated inductance of the exciter field. Combining (1.3) through (1.5)
gives:
dEfd
REfd (1 + Se) + L = vc
dt
Rearranging gives:
L dEfd vc (1.6)
= - Efd - Se Efd
R dt R
This equation is represented by the block diagram of Figure 16-3 if the exciter field resis-
tance, R, is constant. In this case, the quantity (vc/R) can be replaced by a newly normal-
ized variable, Ic, and the transfer function may be fitted to the equation by the following
settings:
Te = L / R (1.7)
Ke = 1 (1.8)
16-5
16.2.2 Shunt Excited Exciter

Equations (1.1), (1.2) and (1.3) apply to shunt excited exciters in the same way as they do
separately excited units. Figure 16-2 shows, though, that the control voltage, vc, is now
applied to the field winding in addition to the output voltage of the exciter itself. Equation
(1.4) must, therefore, be replaced by:
dIx
RIx + L’ = Efd + vc
dt (1.9)
The exciter field inductance continues to be characterized by equation (1.5). Combining

(1.3), (1.5), and (1.9) gives:
(1.10)
This suggests that the block diagram of Figure 16-3a can be fitted to the shunt excited
exciter by setting:
Te = L / R (1.11)
Ke = (1 - 1 / R) (1.12)
(vc / R) = Ic (1.13)
16.2.3 Exciter Operating Practice

The specification of the constants, Ke and Te, must be handled with care because the
exciter field resistance is often an adjustable quantity rather than a design constant. It is
common practice, particularly with shunt exciters, to adjust the exciter field resistance, R,
so that the control source voltage, vc, is nearly zero in steady-state operating conditions.
The steady-state solution of (1.10) is:
vc Efd
= Efd - + Se Efd
R R (1.14)
Commonly, the per unit exciter field resistance is adjustable around a nominal value that is
close to unity.
Figure 16-4a illustrates the solution of (1.14) when the per unit value of R is less than unity.
At operating point I the exciter maintains its own field current requirement exactly and
operate in steady state with zero output from the excitation control source. At operating
point II, the control source must maintain a continuous output, vc, to trim the exciter’s
field current by the amount, Ic.
Changing the exciter’s field resistance to a per-unit value greater than unity produces the
situation shown in Figure 16-4b where the control source must have a nonzero output at all
operating points to make up part of the exciter’s field current requirement that it does not
produce for itself.
16-6
Air Gap Line
Field Resistance Line

Slope = 1 Slope = R
E fd E fd with Positive Control Voltage
I
II
Operating Point with
Zero Control Voltage
E fd with Negative
Control Voltage vc
----- = I c
R
1
E fd  1 – --- = K e E fd
 R
S e E fd Ic
E fd E fd /R Ix
a. Field Resistance Below Critical Value
Field Resistance Line

E fd
Air Gap Line
kv c
-------- = I c
R
Ix
b. Field Resistance Above Critical
Figure 16-4. Effect of Field Resistance on Output of Shunt-Excited Exciter
16-7
Several widely used excitation systems take advantage of the ability of the shunt exciter
field to produce its own full current requirement with little or no output from the control
source. These systems provide slow automatic or manual adjustment of the field resistance
to bring the operating point close to point I, (Figure 16-4a) during normal steady operation
so the full positive or negative drive capability of the control source will give rapid
dynamic adjustment of exciter output during transients. In these systems, the exciter resis-
tance, R, may be assumed to vary little during the few seconds considered in the simula-
tion, but it must be assumed that this resistance has been adjusted slowly to a new value
for each new pre-event steady-state condition.
The result of this practice is the recognition that the parameters, Ke and Te, may be treated
as constants in simulation activity RUN, but should be recalculated by activity STRT for
each new initial condition to satisfy (1.11) and (1.12) and the renormalization of the control
voltage signal in accordance with (1.13). In practice, the value of R is usually quite close to
unity. As a result, Ke is strongly sensitive to small changes in R, but Te and the normaliza-
tion factor 1/R in (1.13) may reasonably be taken as constants. For most practical purposes,
Ke must be readjusted at each new simulation initial condition, but Te and the control
source normalization may be kept constant after proper normal values have been deter-
mined.
16.2.4 Automatic Calculation of dc Exciter Parameters in PSS®E Models

Several of the PSS®E excitation system models, as indicated in Table 1, allow two options:
1. Specify Ke and control source range as constants as appropriate for exciters without
adjustable field resistance. (This option is usually applicable to excitation systems built
by Westinghouse.)
2. Allow Ke and control source range to be calculated by activity STRT to recognize slow
reset adjustment of exciter field resistance. (This option is usually required to represent
GE-built dc exciter systems using an amplidyne as the control source.)
Table 1: Excitation System Models With Automatic Calculation of dc Exciter

Parameters in PSS®E Models
Exciter Type Parameters Calculated on Basis of Assumed

PSS®E Model* Normally Design and Operating Practice by
Represented Activities STRT and ESTR
IEEET1, IEET1B,
IEEEX1, EXDC2
ESDC1A, ESDC2A dc exciter Ke VRMAX VRMIN
IEEET4, IEEEX4
IEEET5, IEEET5A
IEEET2, IEEEX2
ac exciter Ke VRMAX VRMIN
IEEX2A,ESAC5A
16-8
*IEEEXn models comply with IEEE recommendation of 1979.

IEEET1 through IEEET4 models comply with IEEE recommendation of 1969.
IEEET5 and IEET5A models are variations of IEEET4.
EXDC2 model complies with IEEE recommendation of 1981.
ESDC1A and ESDC2A models comply with IEEE Std. 421.5 - 1992.
The optional calculation of these parameters is selected by specifying a zero value for Ke
and/or VRMAX, (the normalized control source maximum output), in the model’s CON
data. The models then calculate these parameters in activities STRT and ESTR as follows:
1. If the CON specifying VRMAX is zero, the model will compute VRMAX and set the CON
to this value. VRMAX is set as follows:
a. When the CON containing Ke is zero or negative, VRMAX will just allow the exciter
to reach an output voltage of E2 when Ke is zero, i.e.,
VRMAX = SE(E2) × E2
b. When the CON containing Ke is positive, VRMAX will just allow the exciter to reach
an output voltage of E2 with the specified value of Ke, i.e.,
VRMAX = (SE(E2) + Ke) × E2
In either case above, VRMIN is then set to –VRMAX.
2. If the CON specifying Ke is zero, the model will compute Ke but will not set the CON to
this value.
Ke is set to the value that will require a voltage regulator output of (VRMAX/10) to maintain
the present value of excitation voltage, Efd, i.e.,
KeEfd = VRMAX/10 – SE(Efd) × Efd
Ke = VRMAX/(10 × Efd) – [SE(Efd)]
The following should be noted:

• VRMAX may usually be treated as a constant for a given machine after it is determined
for one typical loading condition because it is only sightly affected by normal adjustment
of exciter field resistance.
• Ke is a variable depending upon the initial condition value of Efd and hence upon the
generator loading condition.
• The value for VRMAX estimated by the above logic will not necessarily produce the
required excitation system response ratio. It may have to be adjusted following the use
of the excitation system data verification activities ESTR and ERUN to produce a
required response ratio.
The recommended applications of these automatic parameter estimation options:
• Use automatic determination of VRMAX only once in activity STRT, before executing
ESTR and ERUN, to obtain an initial estimate of VRMAX. This value should be refined
and saved during the excitation system verification process.
• The value of the CON specifying VRMAX should, therefore, be nonzero after completion
of exciter data verification as outlined in Chapter 15.
16-9
The ac Excitation Systems Program Application Guide: Volume II
• Use automatic determination of Ke in all normal initialization of dynamic cases.
16.3 The ac Excitation Systems
16.3.1 Components
Virtually all modern excitation systems use an ac source of excitation power and a recti-
fier bridge in place of a dc machine. The ac power source may be either a shaft-driven
alternator, a transformer connected at the terminals of the main generator, or auxiliary
windings in the main generator. The rectifier bridge may be either uncontrolled (diodes) or
controlled, using SCRs.
These components are combined in a wide variety of ways by the different manufacturers.
The ac excitation system models covered by the IEEE recommendations and standard
range from the simple to the elaborate and reflect the breadth of opinion among manufac-
turers on the modeling of their equipment. The following sections cover the main com-
pound models combined in the various excitation system models of PSS®E.
16.3.2 The ac Exciters

Normally, ac exciters are separately excited and, because their only load is a rectifier, their
magnetic behavior can often be represented to acceptable accuracy by the same block dia-
gram and characteristic curve as used for dc exciters (Figure 16-3). The values of Ke and Te
in Figure 16-3 may normally be taken as constants, independent of operating conditions,
when modeling ac exciters.
The block diagram of Figure 16-3 models exciter output voltage as independent of the load
current (i.e., main generator field current) on the exciter. While this is quite reasonable in
the case of dc exciters, which are normally current-compounded, it is less accurate in the
case of ac exciters which, like all alternators, have a substantial synchronous reactance
and hence a substantial drop in terminal voltage as load current is increased. The IEEE
excitation system modeling recommendations of 1969 neglect this armature reaction
within the ac exciter. Newer IEEE recommendations on excitation system modeling pre-
sented do recognize this effect, however, as do the new IEEE recommendations (1981) on
data exchange and Std. 421.5 - 1992. The ac exciter armature reaction is recognized by the
additional block diagram path shown in Figure 16-5, where the parameter, KD, is related to
the exciter’s synchronous reactance.
16-10
Program Application Guide: Volume II The ac Excitation Systems
+ 1
 -------- VEX
sT e
Exciter ac
– Output Voltage
+ Se + Ke

KD L afd i fd
Main Generator
Field Current
Figure 16-5. Transfer Function for ac Exciter

Recognizing Demagnetizing Effect of Its Load Current
16-11
The excitation system models corresponding to 1969 recommendations do not represent

exciter armature reaction. These models have, however, been widely used to represent
excitation systems with ac exciters. Table 2 lists models that represent ac exciter armature
reaction as shown in Figure 16-5.
Table 2: PSS®E Excitation Systems Representing Armature Reaction
EXAC1 ac excitation power source

EXAC1A
EXAC2
EXAC3
ESAC2A
ESAC3A
EXBAS
16.3.3 Compounded Shunt ac Excitation

The alternative to a rotating ac exciter is a transformer connected at the terminals of the
main generator. This alternative may be a simple power transformer or an elaborate cur-
rent-compounding transformer in which the secondary voltage is dependent on both the
current and the terminal voltage of the main generator. The current-compounding trans-
former is shown in Figure 16-6.
The conventional primary winding, which is connected in shunt fashion between phases of
the main generator, is augmented by a winding connected in series with the main genera-
tor stator leads. It is common to place a linear reactor in series with the voltage winding to
limit its current and hence to adjust its mmf in proportion to that of the current winding.
The total primary magnetomotive force on the core of this transformer is given by:
eterm
MMF = kI igen ± j
x
where x is the total reactance in the voltage circuit due to the transformer magnetizing
reactance and the linear reactor. (The sign can be made plus or minus by reversing the
direction of the winding.) The transformer flux, and hence the secondary voltage, are pro-
portional to this MMF (assuming a constant and minimal degree of saturation), giving:
VEX = Kp eterm + jKI igen (1.15)
where VEX is the transformer secondary voltage, and Kp and KI are constants depending on
turns ratios and linear reactor impedance. The excitation required by the loaded main gen-
erator is given approximately by:
Efd = eterm + jxq igen (1.16)
16-12
Main To
Generator • Power
System
•
"Voltage" Winding "Current" Winding
Excitation
Transformer
Core
To Rectifier Bridge and

Main Generator Field
Figure 16-6. Current-Compounded Excitation Transformer Arrangement
Setting the values of Kp and KI approximately to:

Kp = 1.0
KI = xq
makes the output of the compounded transformer approximately proportional to the exci-
tation voltage that is needed by the main generator. Usually, Kp is set to a value slightly
above unity to account for saturation. The approximation (1.15) does not recognize satura-
tion of the main generator exactly or commutation voltage drop of the rectifier, so the
excitation must be adjusted by some feedback means. This feedback may be applied by
controlled saturation of a compounded transformer that feeds an uncontrolled rectifier, or
by the use of a controlled rectifier bridge receiving the output of an unmodulated trans-
former.
The use of current compounding has the advantage that the excitation source voltage, VEX,
can be strong during transmission faults because low terminal voltage is associated with
high current. Its use is not universal, however, and many excitation systems use simple
transformers and rectifier bridges with sufficiently high capability to produce adequate
excitation even when main generator terminal voltage is depressed by faults.
16-13
16.3.4 Field Current Rectifiers

The output of a shaft-driven alternator or transformer excitation power source is fed
directly to a rectifier bridge to provide the required dc field current for the main generator.
The rectifier bridge in many systems is a rotating assembly mounted on the exciter shaft,
because this avoids the need for slip rings with high current carrying capacity. When
mounted on the shaft, the rectifier bridge is uncontrolled. Its output then depends upon the
ac voltage provided by the exciter and on its own commutating voltage drop. This voltage
drop can be significant and is represented in several excitation system models.
The curve of FEX versus in has been approximated in different ways in different IEEE
modeling recommendations over the years. Table 3 lists the PSS®E models that model the
commutation drop.
Table 3: PSS®E Models with Rectifier Commulation Drop
IEEET3 EXAC1 EXST2

IEEEX3 EXAC2 EXST2A
EXAC3 EXST3
EXAC1A ESST2A
ESAC2A ESST3A
ESAC3A ESST4B
ESAC6A EXPIC1
EXBAS
This model of the commutating drop applies for the broad range of load current on the rec-
tifier. It is considered to be necessary in modeling uncontrolled rectifiers. The commutat-
ing process has the effect of reducing the dc output by a factor dependent upon the dc
output current and the level of the exciter-alternator ac voltage as shown by the block dia-
gram in Figure 16-7.
Most controlled rectifier bridges, however, operate only with currents in the lower range
covered by Figure 16-7 and can often be modeled by a linear relationship rather than by the
full curve. The rectifier commutation drop model (Figure 16-7) is combined with the alter-
nator-exciter model of Figure 16-5 and the compounded transformer model to produce the
alternator-rectifier model shown in Figure 16-8 and the transformer-rectifier model shown
in Figure 16-9. The parameters Te, Ke, Se(E1), Se(E2), KD, and Kc in Figure 16-8 are deter-
mined by inherent machine design characteristics and do not change with plant operating
conditions. This model composite alternator-rectifier is incorporated in models such as
EXAC1, EXAC2, and EXAC3.
The parameter, Kc, of Figure 16-9 is an equipment design parameter, independent of gener-
ator operating conditions. The parameters, Kp and KI, while not strictly constant in some
excitation systems that depend upon controlled saturation of the excitation transformer,
are usually treated as constants for each individual generating unit and initial operating
condition. Variations of this model are incorporated in model IEEEX3 and in models such
as EXST2, EXST3 and ESST4B, which correspond to IEEE recommended model types
ST2, ST3 and ST4B.
16-14
The excitation rectifier regulation model shown in Figures 16-7 through 16-9 uses the per-
unit measure, Lafdifd, to characterize the main generator field current. This quantity is
equal to Efd in the steady state, but not during transients when the field winding inductance
contributes to its back-emf. The quantity, Lafd ifd, is equal to unity when a generator is at
rated voltage in the steady state on open-circuit if saturation is neglected.
VEX E fd
 Field
Exciter Voltage
ac Output
Voltage FEX
FEX = f  i n 
in
K c  L afd i fd 
i n = -----------------------------
VEX L afd i fd
FEX
1.0
0.51 0.715 1.0 in
Figure 16-7. Characterization of Excitation Rectifier Commutation Voltage Drop
16-15
1-
vc +  -------  E fd
sT e VEX
Generator
– Field
Voltage
 + Se + Ke FEX = f  i n 
K C  L afd i fd 
i n = ------------------------------
-
VEX
KD L afd i fd
Generator
Field
Current
Figure 16-8. Combined Alternator-Rectifier Model

for Representation of ac Exciter with Uncontrolled Rectifier Bridge at Output
e term
V EX = K p e term+ jK I i gen  Rectifier
i gen VEX dc Output
Voltage
FEX = f  i n 
K c  L afd i fd 
i n = -----------------------------
VEX
L afd i fd
Figure 16-9. Combined Compounding Transformer-Rectifier Model

for Representation of ac Excitation System
16-16
Program Application Guide: Volume II System Voltage Regulators and Control Elements
However, it is equal to S(1.0) in steady-state, rated voltage, open-circuit, operation when

saturation is present (see Figure 14-6). All PSS®E generator models provide the excitation
system models with the instantaneous value of Lafd ifd via the array XADIFD. The quan-
tity, Lafd ifd, as shown in Figure 16-7 through 16-10, is identified in Section 14.2 as IFD. This
terminology is somewhat confusing and can lead to ambiguity in the definition of the
parameter KC of Figure 16-7. The key point, on which proposed values of KC can be tested
for reasonableness, is that the signal characterizing generator field current (whether desig-
nated Lafd ifd or IFD) should be equal to unity for steady-state, open-circuit, rated voltage
operation, in the absence of saturation.
Excitation System
L ad i fd
Gef  s  K gf  s 
Gei  s  K gi  s 
i gen
+ +
v ref + G vr  s  G ee  s  + K ge  s 
E fd et
–
Terminal Voltage Feedback
Voltage Excitation Main

Regulator Power Generator
Subsystem
Figure 16-10. Excitation System Overall General Format
16.4 System Voltage Regulators and Control Elements
16.4.1 Transfer Functions

The excitation power elements covered in Sections 16.2 and 16.3 are controlled by a volt-
age regulator to form a feedback loop around the main generator. The voltage regulator is
normally implemented with low power level components; these are usually electrome-
chanical or magnetic amplifier devices in older systems, and are purely electronic in most
more modern systems. Voltage regulator components respond to error in terminal voltage
16-17
System Voltage Regulators and Control Elements Program Application Guide: Volume II
as shown in Figure 16-10.

The excitation system’s overall transfer function includes both the dynamic characteristics
of the exciter elements and the transfer function of the voltage regulator. The elements
Gee(s), Gei(s), and Gef(s) depend upon the details of the excitation power source as shown,
for example, in Figures 16-5, 16-8, and 16-9. (These transfer functions could be derived
readily by algebraic manipulations of the figures.) The generator transfer functions vary
widely with loading conditions and system impedances. For example, Kge(s) is the simple,
time constant function 1/(1 + Tdos) when the generator is being synchronized and a high-
order transfer function with complex poles when it is synchronized and loaded.
The steady-state gain of the voltage regulator should be high to keep the voltage error as
small as possible in the steady state. High values of gain in Gvr(s) are not tolerable with
respect to dynamic behavior, though, because the generator’s field time constant ensures
the presence of sufficient phase lag to produce oscillatory behavior of the main feedback
loop. It is, therefore, common practice to design the voltage regulator to have a high
steady-state gain and a relatively low gain under transient conditions. The desired voltage
regulator characteristic is shown, in Bode plot terms, in Figure 16-11.
The gain reduction is achieved in a number of different ways in different types and vin-
tages of excitation system. Many of the more modern controlled rectifier systems handle
the shaping of their transfer function entirely within the voltage regulator by giving it a
transfer function of the form
A (1 + T1s)
Grr(s) =
(1 + T2s)
This gives a steady-state gain of A and a gain at high perturbation frequencies of

A(T1/T2). Older systems and those where the voltage regulator elements are magnetic,
rather than electronic, usually achieve their transient gain reduction by a rate feedback
arrangement, using rate of change of field voltage, or some closely related signal, as
shown in Figure 16-12. The newer systems using the gain reduction transfer function (1 +
T1s)/(1 + T2s) are, typically, set up with T1 and T2 equal to about 1 and 10, respectively.
This usually allows the steady-state gain, A, to be in the range 100 to 400, with a corre-
sponding high-frequency gain in the region of 10 to 40.
16-18
Program Application Guide: Volume II System Voltage Regulators and Control Elements
Gain
G(j)
, Perturbation Frequency
Figure 16-11. General Required Form of Voltage Regulator Transfer Function
E fd
+
v ref +   G vr  s  G ex  s  K gen  s  et
– –
sK f Rate Feedback with

----------------
- Excitation System
1 + Tf s
Main Voltage Feedback Loop
Figure 16-12. Use of Rate Feedback to Produce Transient

Gain Reduction in Excitation System
The tuning of older systems is best illustrated by considering the somewhat simplified
excitation system form shown in Figure 16-13. The excitation system time constant is nor-
mally much less than 1 sec; it is usually practical to set the rate feedback time constant to
1 or 2 sec and have it significantly greater than TEX.
When Tf is greater than TEX, the transfer function of Figure 16-13 may be approximated by:
A (1 + Tf s)
1 + (Tf + AKf) s
16-19
System Voltage Regulators and Control Elements Program Application Guide: Volume II
A typical situation might call for a steady-state gain of 200, Tf equal to 2 sec, and a high-
frequency gain of 20. For this situation:
Tf
= 0.1
Tf + AKf
2
= 0.1
2 + 200Kf
Kf = 0.09
1
 A ----------------------
+ 1 + T EX s
v
–
Kf s
----------------
-
1 + Tf s
E fd  s  A  1 + Tf s 
- = ----------------------------------------------------------------------------------
--------------- -
v  s  2
T EX T f s +  T EX + T f + AK f s + 1
Figure 16-13. Simplified Excitation System With Rate Feedback
16.4.2 Voltage Regulator Limits

The modeling of the voltage regulator must cover both its linear transfer function, Gvr(s),
and any limits that may affect its output. Systems using purely electronic methods to
derive a firing delay for SCR bridges may usually be assumed to have a large dynamic
range and no significant signal limits on the voltage regulator section. Older systems,
however, frequently have a limited dynamic range and may involve significant limits.
One of the most significant voltage regulator limits occurs in the case of dc generator type
exciters where the final stage of the voltage regulator, the control element of
Sections 16.2.1 and 16.2.2, applies current to the exciter’s field. This final stage is often an
amplidyne for which the rating and maximum output, rather than the voltage regulator
gain, are the principal factors determining the transient behavior of the exciter. In this
case, the amplidyne limits are represented by the parameters VRMAX and VRMIN in the
IEEE type 1 model and these should be specified or determined automatically in accor-
dance with Section 16.2.4.
16-20
Program Application Guide: Volume II Specific Excitation System Models
16.5 Specific Excitation System Models
16.5.1 Models IEEET1, IEET1A, IEEEX1, ESDC1A

Models IEEET1, IEET1A, IEEEX1, and ESDC1A are widely used to represent systems with
shunt dc exciters as well as systems with alternator exciters and uncontrolled shaft-
mounted rectifier bridges. When used to represent dc exciter systems, the constants Ke,
VRMAX, and VRMIN, should be entered as zero so that PSS®E will determine their values in
accordance with Section 16.2.4. When used to represent alternator-rectifier systems, the
constant, Ke, should be set to unity, and VRMAX, VRMIN should be set to zero so that their
values will be assigned by PSS®E during the response ratio test option of activity ESTR.
This option automatically initializes each generator to its rated operating condition.
16.5.2 Models IEEET2, IEEEX2, IEEX2A

These models differ from IEEET1, etc., only in the source used for the excitation system
stabilizing feedback. IEEET2 and IEEEX2 take this signal as proportional to the control ele-
ment output, while IEEX2A takes this signal to be proportional to exciter field current. (For
comparison, IEEET1, etc., takes a stabilization signal proportional to the output, Efd, of
the exciter.) Normally these models represent alternator-uncontrolled rectifier excitation
systems. Accordingly, Ke should normally be set to unity and VRMAX, VRMIN should be set
by PSS®E using the response ratio test option of activity ESTR .
16.5.3 Models IEEET3, IEEEX3

Models IEEET3 and IEEEX3 differ only in the placement of the limit, VBMAX. Both are
intended to represent SCPT excitation systems manufactured by General Electric. These
SCPT systems obtain excitation power from a compounding transformer of the general
form shown in Figure 16-6. They achieve control of the field voltage by a control winding
on this transformer that is able to saturate its core and effectively use it as a large magnetic
amplifier. The rectifier bridge is uncontrolled.
The key parameter of these models is the current compounding gain, KI, which specifies
the relative strength of the voltage and current inputs to the excitation power source. Nor-
mal design practice adjusts the gains, Kp and KI, roughly in proportion to unity and the
quadrature synchronous reactance of the generator, respectively. Refinement of such val-
ues is necessary, though, to account for rectifier commutation drop and saturation of the
generator. Refined estimates of Kp and KI can be obtained on the assumption that the com-
pounding excitation transformer is sized to provide rated excitation voltage with essen-
tially idle output, VR, from the voltage regulator. An infinite range of values of Kp and KI
will achieve this. PSS®E requires the user to specify Kp and then, if a zero CON value is
specified for KI, PSS®E computes KI so as to have an idle value of the voltage regulator
output. A reasonable basis for Kp is to set it to a value slightly greater than the per unit
field current of the main generator during unity voltage open-circuit operation. As a result,
the generator provides its own excitation naturally, with idle voltage regulator output both
at zero current output and at full load. The typical generator magnetization curves of
16-21
Specific Excitation System Models Program Application Guide: Volume II
Figure 14-5 suggest 1.1 to 1.2 as the reasonable range for Kp.
The automatic determination of KI is best handled using the response ratio test option of
activity ESTR, which always initializes each machine to rated conditions. In most cases, it
is then advisable to review the value of KI determined by PSS®E (in VAR(L)) and transfer
it to CON(J+9) so that it is, henceforth, a constant, not adjusted as the generator operating
point is varied.
16.5.4 Models IEEET4, IEEEX4, IEEET5, IEET5A

Models IEEET4, IEEEX4, IEEET5, and IEET5A represent older excitation systems where the
exciter is a dc machine and the voltage regulator is one of a wide variety of electrome-
chanical devices. All four models use the same exciter representation and all will calculate
the field resistance parameter, Ke, (see Section 16.2.4) if its value is specified as zero in the
CON array. The model block diagrams show variations of voltage regulator logic. None is
a purely continuous acting regulator. IEEET4 models a slow reset controller with a dead-
band, KR, and a parallel quick-action controller where the deadband, Kv, should be larger
than KR. IEEEX5 models a similar system but with different normalization of the reset
rate. IEEET5 is similar again, but has no deadband in its slow reset path.
16.5.5 Models SEXS and REXSYS

Models SEXS and REXSYS represents no specific type of excitation system, but rather the
general characteristics of a wide variety of properly tuned excitation systems. Model
SEXS is particularly useful in cases where an excitation system must be represented and
its detailed design is not known. The gain, K, time constant, TE, and limits EMAX, EMIN, are
a basic representation of the excitation power source. Time constants TA and TB provide
the transient gain reduction needed to allow satisfactory dynamic behavior with high
steady-state gain. Typical parameters for the SEXS model, to represent an unknown but
presumably well-tuned excitation system are:
TA = 1 sec
TB = 10 sec
K = 200 to 400
TE = 0.05 sec
EMIN =0
EMAX = 2.5 to 6, depending upon equipment rating
By proper selection of data REXSYS can be used to represent many different excitation
systems where the power source is either an ac or dc generator. Because it can represent
many exciters, typical data cannot be supplied. This model is not currently available in the
extended term dynamics.
16-22
16.5.6 Model SCRX

Like SEXS, SCRX is also a general model, not tailored to the representation of any specific
excitation system. SCRX represents the general characteristics of excitation systems
shown schematically in Figure 16-14. These systems are frequently built with rectifier
bridges for which current, in relation to the applied voltage, results in a relatively small
commutation drop, hence, allowing field voltage to be represented as independent of field
current. SCRX can represent a rectifier bridge system fed either from an independent
source, such as an externally supplied plant auxiliary bus, or from a transformer connected
directly at the generator terminals. SCRX distinguishes between excitation rectifier sys-
tems having bidirectional current capability and those that can carry current only in the
positive (normal) direction. The first six parameters of SCRX have the same meaning and
typical values as those of SEXS; CON(J+6), CSWITCH, distinguishes between systems
(bus fed) where ac supply is proportional to generator terminal bus voltage and those
(solid fed) where supply is independent of generator terminal voltage.
CON(J+7), rc/rfd, distinguishes between bidirectional and unidirectional excitation sys-
tems (Figure 16-14). Normal construction is unidirectional as shown in Figure 16-14a, with
a single rectifier bridge. The thyristors of this bridge are normally fed firing pulses in rec-
tifier mode to provide positive field voltage and carry positive current. During transients,
when generator terminal voltage is high, the thyristor firing pulses may shift to the inverter
regime to provide a negative field voltage, Efd, while the bridge continues to carry positive
current. This negative forcing of the field can reduce field current rapidly but cannot allow
negative field current. Accordingly, the bridge is often protected by a reverse bypassing
thyristor that can be fired by a crowbar circuit whenever a negative field current, attempt-
ing to flow into the diodes, produces a high positive back emf across the bridge. The
bypassing thyristor has a series resistor, rc, to limit its current. A typical value of rc is
about ten times the field resistance, rfd, in order to provide a time constant of field winding
discharge approximately 1/10 the natural time constant, Tdo on open circuit of the field
winding.
Note that the use of the reverse bypassing diode is not universal and that some controlled
rectifier excitation systems are designed to withstand any high back emf that is produced
by simply blocking reverse field current. These should be modeled by specifying a high
value of rc. Figure 16-14b shows an excitation system layout in which two rectifier bridges
are connected in reverse polarity to allow field current to flow in either direction. Only one
of the two bridges receives firing pulses at a time, and the timing of these allows the field
voltage to be either positive or negative, regardless of the direction of field current. This
arrangement, while more expensive and less common than the single bridge, is used on
hydro generators that feed radial transmission and that are susceptible to magnetic self-
excitation when the transmission is opened at the receiving end. SCRX assumes the exci-
tation system has the configuration of Figure 16-14b if the parameter, rc/rfd, in CON(J+7) is
specified as zero.
The use of a nonzero value of rc/rfd to represent an unidirectional system leads to a
small time constant in the excitation feedback loop when field current attempts to
reverse. The effective value of this time constant is not Tdo/(rc/rfd), as might at first be
expected. It is often closer to Td/(rc/rfd) and may become small enough to produce numer-
16-23
ical instability of the simulation calculation with the time step of 0.00833 sec that is nor-
mally used in PSS®E. Hence, because negative field current is a fairly uncommon
occurrence except in radial load rejection situations, it is recommended that the rc/rfd
parameter, CON(J+7), normally be set to zero. Negative field current can then be detected
readily by plotting if it is suspected, and the excitation system data can be refined, with
due care, when necessary.
Protective
Trigger
Circuit
rc Field
Excitation Winding (rfd)
ac Source
Controlled Discharge
Rectifier Resistor
Bridge
a. Unidirectional Excitation Rectifier
Field
Winding
ac Excitation
Source
Forward Reverse
Controlled Controlled
Rectifier Bridge Rectifier Bridge
b. Controlled Rectifier Excitation
Figure 16-14. Controlled Rectifier Excitation System
16.5.7 Models EXDC2 and ESDC2A

Models EXDC2 and ESDC2A differ from IEEET1, etc. in only one aspect: the voltage regu-
lator’s source of supply is the generator or auxiliary bus voltage. As a result, the regulator
output limits are proportional to VT.
16-24
16.5.8 Models EXAC1 and ESAC1A

Models EXAC1 and ESAC1A emulate a field-controlled alternator rectifier excitation sys-
tem consisting of an alternator main exciter with noncontrolled rectifiers. The exciter does
not employ self-excitation and the voltage regulator power is taken from a source not
affected by external transients. The diode characteristic in the exciter output imposes a
lower limit of zero on the exciter output voltage. These models are applicable for simulat-
ing the performance of Westinghouse brushless excitation systems. The demagnetizing
effect of load current, Ladifd, on the dynamics of the exciter alternator output voltage, VE,
is accounted for in the feedback path that includes the constant, KD. This constant is a
function of the exciter alternator synchronous and transient reactances. Exciter output
voltage drop, due to rectifier regulation, is simulated by inclusion of the constant, KC, a
function of commutating reactance, and the approximation to the rectifier regulation
curve, FEX, as described in Section 16.3.4.
In the models, a signal, VFE, proportional to exciter field current is derived from the sum-
mation of signals from exciter output voltage, VE (multiplied by the term KE + SE) and
Ladifd (multiplied by the demagnetization term, KD). The exciter field current signal, VFE,
is used as the input to the excitation system feedback. None of the parameters of these
models are calculated by PSS®E.

Models EXAC2 and ESAC2A emulate a high initial response field controlled alternator-rec-
tifier excitation system in which the alternator main exciter is used with noncontrolled rec-
tifier. The models are similar to that of EXAC1 except for the inclusion of two additional
exciter field current feedback loops simulating exciter time constant compensation and
exciter field current limiting elements, respectively. These models are applicable for simu-
lating the performance of Westinghouse high initial response brushless excitation systems.
A direct negative feedback, VH, around the exciter field time constant reduces its effective
value and thereby increases the bandwidth of the excitation system small signal response.
The time constant is reduced by the gain (1 + KB KH) of the compensation loop and is nor-
mally more than an order of magnitude lower than the time constant without compensa-
tion. To obtain high initial response with this system, a very high forcing voltage, VRMAX,
is applied to the exciter field. A limiter sensing exciter field current allows high forcing,
but limits the current. By limiting the exciter field current, exciter output voltage, VE, is
limited to a selected value, VLR, which is usually determined by the specified excitation
system response ratio. The output signals from the voltage regulator, VA, and time con-
stant compensation, VH, elements are compared with the output signal, VL, from the lim-
iter in control logic circuitry, which functions to provide a sharp transition from regulator
control to limiter control of excitation at the limit point. Excitation is controlled by the
more negative of the two control signals.
Although the current limit is realized physically as in EXAC2, the time constants associ-
ated with the loop can be extremely small. Therefore, the limit can be modeled as a posi-
tive limit on exciter voltage back of commutating reactance as done in model ESAC2A.
None of the parameters of these models are calculated.
16-25

Models EXAC3 and ESAC3A emulate a field controlled alternator-rectifier excitation sys-
tem, which includes an alternator main exciter with noncontrolled rectifiers. The exciter
employs self-excitation and the voltage regulator power is derived from the exciter output
voltage. Therefore, this system has an additional nonlinearity, simulated by the use of a
multiplier for which the inputs are the voltage regulator command signal, VA, and the
exciter output voltage, EFD, times KR. These models are applicable to systems such as the
GE ALTERREX excitation systems employing static voltage regulators.
The demagnetizing effect of load current, Ladifd, on the dynamics of the exciter alternator
output voltage, VE, is accounted for in the feedback path, which includes the constant KD,
which is a function of the exciter alternator synchronous and transient reactances. Exciter
output voltage drop due to rectifier regulation is simulated by inclusion of the constant, KC
(which is a function of commutating reactance) and the approximation to the regulation
curve, FEX. In the model, a signal proportional to exciter field current, VFE, is derived from
the summation of signals from exciter output voltage, VE (multiplied by the term KE + SE)
and Ladifd (multiplied by the demagnetization term, KD). The excitation system stabilizer
also has a nonlinear characteristic; the gain is KF for exciter output voltage less than EFDN.
When exciter output exceeds EFDN, the value of this gain becomes KN. Model ESAC3A
represents the effects of the feedback limiter operation by limits on VE. None of the
parameters of EXAC3 and ESAC3A are calculated by PSS®E.

Models EXAC4 and ESAC4A emulate an alternator-supplied rectifier excitation system.
This high initial response excitation system utilizes a full thyristor bridge in the exciter
output circuit and the voltage regulator operates directly on these elements. The exciter
alternator uses an independent voltage regulator to control its output voltage to a constant
value. These effects are not modeled, however, transient loading effects on the exciter
alternator are included. Exciter loading effects can be accounted for by using the exciter
load current and commutating reactance to modify excitation limits. The excitation system
feedback is frequently accomplished in thyristor systems by a series lag-lead network
rather than through rate feedback. The time constants, TB and TC, would be used to simu-
late this control function. The overall equivalent gain and the time constant associated
with the regulator and/or firing of the thyristors would be simulated by KA and TA, respec-
tively. Systems utilizing these simulation models include the General Electric
ALTHYREX and rotating thyristor excitation systems. The difference in these two models
is in the handling of the exciter limits.
16.5.12 Models EXST1 and ESST1A

Models EXST1 and ESST1A, of a potential source controlled rectifier-exciter excitation
system, are intended to represent systems in which excitation power is supplied through a
transformer from the generator terminals (or the unit’s auxiliary bus) and is regulated by a
controlled rectifier. The maximum exciter voltage available from such systems is directly
related to the generator terminal voltage (except as noted below).
16-26
In this type of system, the inherent exciter time constants are very small and exciter stabi-
lization as such is normally not required. On the other hand, it may be desirable to reduce
the transient gain of such systems for other reasons. The model shown is sufficiently ver-
satile to represent transient gain reduction implemented either in the forward path via time
constants, TB and TC (in which case KF would normally be set to zero), or in the feedback
path by suitable choice of rate feedback parameters, KF and TF. Voltage regulator gain and
any inherent exciter time constants are represented by KA and TA.
In many cases, the internal limiter following the summing junction can be neglected, but
the field voltage limits, which are functions of both terminal voltage (except when the
exciter is supplied from an auxiliary bus, which in turn is not supplied from the generator
terminals) and generator field current, must be modeled. The representation of the field
voltage limits as linear functions of generator field current is possible. In addition, for
most transformer fed systems, KC is quite small, permitting the term to be neglected for
many studies.
While the majority of these excitation systems employ a fully controlled bridge, the model
is also applicable to semicontrolled systems. In this case, the negative ceiling is set to zero.
Examples of type ST1 excitation systems are:
• Canadian General Electric silcomatic excitation systems.
• Westinghouse Canada solid-state thyristor excitation system.
• Westinghouse type PS static excitation system with type WTA, WHS, WTA-300, or
type TWGR regulators.
• ASEA static excitation system.
• Brown Boveri static excitation system.
• Rayrolle-Parsons static excitation system.
• GEC-Eliott static excitation system.
• Toshiba static excitation system.
• Mitsubishi static excitation system.
• General Electric potential source static excitation system.
• Hitachi static excitation system.
• Basler model SSE excitation system.
For most of these systems, the cosine characteristic relating thyristor firing angle to bridge
output is compensated for by an inverse-cosine function between regulator output and fir-
ing angle. In such systems the gain, KA, is a constant and is independent of exciter supply
voltage. In a few systems, this inverse-cosine compensation is not employed, and KA
becomes a cosine function, dependent on supply voltage. This model can be used to
approximate these systems for some types of studies, but more accurate representation
may be required for others. Model ESST1A models a field current limiter to protect the
generator rotor and exciter, which sometimes results from the very high forcing capability
of these systems. The gain of this limit is represented by KLR and a start setting of ILR.
16-27
16.5.13 Models EXST2, EXST2A and ESST2A

Some static systems utilize both current and voltage sources derived from the generator
terminal quantities as components of the power source. These compound source rectifier
excitation systems are modeled by these ST2 models. It is necessary to form the exciter
power source from a phasor combination of terminal voltage, VT, and terminal current, IT.
Rectifier loading and commutation effects are accounted for (see Section 16.3.4). The regu-
lator controls the exciter output through controlled saturation of the power transformer
components. TE represents the integration rate associated with the inductance of the con-
trol windings, and the EFDMAX limit represents the limit on the exciter voltage due to satu-
ration of the magnetic components. If the KI CON value is specified as zero, the model
calculates the proper value of KI using the same assumptions as used for IEEET3 and
IEEEX3. One example of such a systems is the GE static excitation system, frequently
referred to as the SCT-PPT or SCPT system.
EXST2A is identical to EXST2 except that in the EXST2A model the regulator output is
multiplied rather than added to the compounded transformer output. This modification
was made to more accurately model test results.
ESST2A is identical to EXST2A except for under-excitation limit input.
16.5.14 Models EXST3 and ESST3A

Some static systems utilize internal quantities within the generator, which may be
expressed as phasor combinations of generator terminal voltage and current, to form the
source of excitation power. Such compound source-controlled rectifier-excitation systems
employing controlled rectifiers in the exciter output circuit are represented by these mod-
els. The excitation system stabilizer for these systems is provided by a series lag-lead ele-
ment, represented by the time constants, TB and TC. An inner loop field-voltage regulator
is characterized by of the gains, KA and KG, and the time constants, TA. Rectifier loading
and commutation effects are accounted for (see Section 16.3.4). The EFDMAX limit is estab-
lished by the saturation level of power components. Systems of this type include the Gen-
eral Electric GENERREX and shunt-thyristor excitation systems.
Location of limits differentiate ESST3A from EXST3. Note that ESST3A is from the IEEE
standard.
16.5.15 Model ESAC5A

ESAC5A is a simplified model for a brushless excitation system where the regulator is sup-
plied from a source such as a permanent magnet generator, which is not affected by system
disturbances. The model can be used to represent small excitation systems such as those
produced by Basler and Electric Machinery.
As for the dc exciters, this model uses loaded rather than open circuit exciter saturation
data.
16.5.16 Model IEET1B

Model IEET1B is included to provide the modeling of exciters developed for Consolidated
Edison Company units; Ravenswood 1, 2, and 3; Poletti; and Arthur Kill.
16-28
16.5.17 Model EXPIC1

EXPIC1 is recommended to be used for excitation systems where voltage regulator control
element is a proportional plus integral type (PI). A specific excitation system was not used
as a starting point. Judicious choice of constants can allow this model to be used to model
a variety of manufacturers implementation of a PI type exciter. The model has been used
in PTI studies of several COGEN facilities, for exciters manufactured or supplied by the
Fuji Electric Co., Ltd., GE Canada, and the new GE EX2000 digital-based excitation sys-
tems.
16.5.18 Type ESAC6A Excitation System

Type ESAC6A excitation system model is used to represent field-controlled alternator-rec-
tifier excitation systems with system-supplied electronic voltage regulators. The maxi-
mum output of the regulator, VR, is a function of terminal voltage, VT, and the model
includes an exciter field current limiter. It is particularly suitable for representation of sta-
tionary diode systems such as those produced by C.A. Parsons.
16.5.19 Model EXBAS

Model EXBAS is used to represent Basler static voltage regulators where output provides
the entire field current of a shaft-mounted dc or ac exciter. The model should be used
mainly for separately excited exciters where the regulator is the sole source of dc excita-
tion to the excitation field, which may be either shaft mounted dc or ac. Therefore, the user
should typically enter a value of Ke near unity. The regulator power supply of this model
is auxiliary bus fed independently from the generator or is a shaft-driven permanent mag-
net generator, which gives essentially constant ac voltage.
16.5.20 Model ESAC8B

Model ESAC8B represents the Basler digital excitation control system voltage regulator as
applied to a brushless exciter. The automatic voltage regulator consists of a PID controller
implemented in a microprocessor. Because it uses a high processor rate, the design is done
as if the controller were continuous even through it is digital.
16.5.21 Model EXELI

Model EXELI is used to represent an all-static PI transformer fed excitation system. This
model combines a stabilizer and an exciter unit.
A PI voltage controller establishes a desired field current setpoint for a proportional cur-
rent controller. The integrator of the PI controller has a follow-up input to match its signal
to the present field current.
The stabilizer section can be nullified by a proper selection of the associated constants,
which can make the stabilizer signal equal to zero. The regulation of the transformer/recti-
fier unit is represented by Xe and it can be zero or positive. The controller reset time con-
stant, Tnu, cannot be zero.
16-29
Alternate Voltage Regulator Inputs Program Application Guide: Volume II
16.5.22 Model ESST4B

Model ESST4B is used to represent static systems with both potential and compound
source rectifier excitation. This model represents the following types of equipment:
• GE EX2000 Bus Fed Potential Source
• GE EX2000 Static Compound Systems
• GE Ex2000 Generrex - PPSor - CPS
Rather than the lead-lag regulation as used by the ESST3B, this model has a non windup
lead-lag regulator. Maximum excitation input for this model is into a low value gate down-
stream from the proportional and integral blocks. Minimum excitation is summed at the
reference.
16.5.23 Models REXSYS and REXSY1

Models REXSYS and REXSY1, available for state space simulation only, were added to
match a general type of exciter available in an alternate stability program and used fre-
quently. A quick review of the data shows that many different systems can be modeled
with these exciters by the proper selection of constants.
16.6 Alternate Voltage Regulator Inputs
16.6.1 Principle
Generator voltage regulators may be used to control either generator terminal voltage, a
current compensated terminal voltage, or a remote bus voltage. Current compensation is
achieved by using a compensated voltage:
vc = eterm + Zc igen (1.17)
in place of generator terminal voltage, vt, as the voltage feedback input signal to the volt-
age regulator. Remote bus voltage regulation is achieved by using a sensed voltage in
place of the terminal voltage. All PSS®E voltage regulator models use a voltage signal
from the array ECOMP as the input. If the user has called no specific model, ECOMP is
set equal to ETERM, causing the excitation system to regulate terminal voltage. The volt-
age regulator on a given generator regulates an alternate voltage by applying model
IEEEVC, COMP, or REMCMP to that machine. These models replace the terminal volt-
age in the ECOMP array with the compensated voltage or the alternate bus voltage, and
feed this signal to the voltage regulator.
16.6.2 Model IEEEVC

IEEEVC compensates the terminal voltage according to (1.17). Polarity is critical. The cur-
rent, igen, is positive when leaving the generator. Accordingly, positive values of Rc and Xc
in Zc = (Rc + jXc) cause the compensated voltage to appear to be the voltage at a point
inside the generator. If IEEEVC is used, the generator’s actual terminal voltage continues
16-30
Program Application Guide: Volume II Alternate Voltage Regulator Inputs
to be accessible in array ETERM while its compensated terminal voltage is available in

array ECOMP. Both ETERM and ECOMP values may be assigned to output channels.
16.6.3 Model COMP

Model COMP compensates generator terminal voltage in general accordance with (1.17),
but with a reversal of presumed polarity and with only a reactive component in the com-
pensating impedance. A positive value of Xe, CON(J) causes the compensated voltage to
correspond to that at a point outside the generator. COMP places the compensated voltage
in array ECOMP and leaves terminal voltage in array ETERM. Both may be assigned to
output channels.
16.6.4 Model REMCMP

Model REMCMP places the voltage of the bus specified by ICON(I) in the ECOMP array.
If the remote bus is not found in the network or if the bus specified is out-of-service, an
appropriate message is printed and terminal voltage is fed to the regulator. If during a sim-
ulation the remote bus is disconnected, a large discontinuity may be fed to the regulator
due to this switch. If REMCMP is used, the generator’s actual terminal voltage continues
to be accessible in array ETERM while the remote bus voltage is available in array
ECOMP. Both ETERM and ECOMP values may be assigned to channels.
16-31
Supplementary Excitation Controller Models Program Application Guide: Volume II
16.6.5 Model COMPCC

Model COMPCC compensates terminal voltage in accordance with equations:
(1.18)
(1.19)
This scheme is used to equalize division of reactive MVA and to prevent circulating reac-
tive current between two ac machines with individual regulators operating in parallel.
Cross-compound units are the most likely candidates to share one transformer and use this
scheme. This model assumes both units on the same bus and, therefore, the output
ETERM will be the same for both units. The ECOMP value for each unit will be different
depending on impedance input and generator currents.
16.7 Supplementary Excitation Controller Models
16.7.1 General Principles

Excitation systems with high transient gain and small time constants tend to reduce the
damping of generator rotor angle oscillations. This negative damping effect can be coun-
teracted by making the excitation system respond to rotor angle motion as well as devia-
tions of terminal voltage under transient conditions, while being sensitive only to terminal
voltage in the steady state. A detailed discussion of the destabilizing effect and its counter-
action by recognition of rotor angle motion is presented in "Concepts of Synchronous
Machine Stability as Affected by Excitation Control," F.P. de Mello, C. Concordia, IEEE
Transactions on Power Apparatus and Systems, Vol. PAS-88, pp. 316-329, 1969.
The application of supplementary stabilizing signals is illustrated by Figure 16-15. The nor-
mal excitation system path is represented here by the transfer function, E(s). The normal
reference signal of the voltage regulator, vref, is modified by the addition of a supplemen-
tary signal, vs, that is derived from a suitable measurement related to shaft motion via a
transfer function, G(s). All PSS®E excitation system models assume that the supplemen-
tary signal may be present and calculate the error signal for the voltage regulator as:
e = vref + vs - et
PSS®E always uses compensated terminal voltage in place of actual terminal voltage. The
reference signal is present in array VREF, the compensated voltage is taken from array,
ECOMP, and the supplementary signal is taken from array VOTHSG. Accordingly, the
voltage regulator error calculation in terms of PSS®E arrays is:
e = VREF(I) + VOTHSG(I) - ECOMP(I)
16-32
Program Application Guide: Volume II Supplementary Excitation Controller Models
T mech
Te K1
--------

+ +
– Ta 1 o 
  ---------- -------
Te 2H s S
+
T e
----------
-K
Eq 2 K4 K5
1 E e t
--------- ----------q- -------
dK 3  
Excitation
– System
+
K3 + –
-----------------------------
-  Es   
E q 1 + K 3 T do s E fd e t ECOMP
+
+
e t K6
----------
-
Eq Gs
Supplementary v
s
Stabilizer
VOTHSG +
 + VREF
Voltage Regulator
Input Signal
Figure 16-15. Overall Synchronous Machine Block Diagram
The three signals in ECOMP, VREF, and VOTHSG are per unit and all calculations of
supplementary stabilizing signals must reduce the value to per unit in terms of generator
voltage before placing it in the VOTHSG array. The gain values used in supplementary
stabilizer models must be specified accordingly. The user may use any meaningful logic in
computing the instantaneous value of VOTHSG(I). The usual approach is to call a PSS®E
library model, such as IEEEST, and implement an appropriate stabilizer transfer function,
G(s), to determine the supplementary signal. The model call or other calculation determin-
ing the supplementary signal must be made prior to the call of the excitation system model
that must respond to this signal.
16.7.2 Model IEEEST

Model IEEEST implements the general-purpose supplementary stabilizer representation
given in the 1979 IEEE modeling recommendations (see Section 14.2). IEEEST is able to
pick up a variety of different input signals, all of which may be presumed to be related to
16-33
shaft motion under certain circumstances. The G(s) transfer function implemented by
IEEEST is quite general; it includes a notch filter, two lead-lag blocks to introduce phase
lead into the supplementary signal, and a washout block to ensure that the supplementary
signal is zero in the steady state.
The ICONs of IEEEST specify the source of the signal inferring shaft motion. A value of
1, 3, or 4 for ICON(IC) indicates that the signal is to be a quantity defined by the machine
identified by the arguments IBUS and IM; the value of ICON(IC+1) is ignored in these
cases. When the value of ICON(IC) is 2, 5, or 6, the signal is defined by the electric net-
work and can be measured at a point other than the machine where the supplementary sig-
nal is applied. In this case, a zero value of ICON(IC+1) indicates that the signal is to be
measured at the machine terminals and a nonzero value indicates that the signal is to be
measured at a point remote from the generator terminals, namely the bus for which the
number is given.
The basic function of IEEEST is to make the phase of the supplementary signal lead that
of the input signal. This function is handled by the two lead-lag blocks and is specified by
the time constants T1, T2, T3, and T4. The required gain setting and steady-state washout
function are provided by the constants K3, T5, and T6. These seven constants, except T6,
may be set to zero to bypass an unneeded block. Setting T1 = T2 = 0 bypasses one lead-lag
block if its phase lead is not needed. If no washout action is needed, i.e., if the input is
accelerating power, the effect of the washout block may be cancelled by setting T5 = T6 =
20.
The filter block of IEEEST may be used to represent the filtration that is built into some
supplementary stabilizers. This filtration usually takes the form of a notch or band-stop fil-
ter designed to block a troublesome oscillation not related to the one that the stabilizer is
intended to act upon. This may, for example, be a torsional oscillation of the turbine gener-
ator shaft or an oscillation between two closely coupled machines that is observable in the
input signal but not related to an oscillation of the machine against other remote machines.
The filter block may be bypassed by setting the CONs A1 through A6 to zero. Nonzero
values of numerator coefficients must be associated with nonzero denominator coeffi-
cients. The values of A1 through A6 in any specific stabilizer will depend, in large mea-
sure, on its designer’s views and preferences.
The limits LSMAX and LSMIN provide simple clipping of the supplementary signal; typical
limit values are 0.1 pu to ensure that the supplementary signal does not overwhelm the
voltage regulation function during major system disturbances. The limits VCU and VCL
allow the supplementary signal to be blocked completely when the generator’s compen-
sated terminal voltage (designated as VCT on the IEEEST data sheet) falls outside the band
from VCU to VCL. These two CONs, VCU and VCL, may be set to zero to bypass this block-
ing function.
16.7.3 Model STAB1

Model STAB1 is a subset of IEEEST responding to the shaft speed of its designated gener-
ator as its only input. It’s action is the same as IEEEST when all A constants and blocking
CONs of IEEEST are set to zero.
16-34
16.7.4 Models STAB2A, STAB3, and STAB4

Models STAB2A, STAB3, and STAB4 are special representations of specific types of supple-
mentary stabilizing units. Both produce a supplementary signal by introducing phase-lead
into a signal proportional to electrical power output measured at the generator terminals.
16.7.5 Model IEE2ST

Model IEE2ST is a derivative of IEEEST. It allows two input signals to be summed to cre-
ate the signal for processing by the phase-lead blocks. Each signal is selected by ICON
values as described in Section 16.7.2. Either input may be blocked by setting the corre-
sponding ICON to zero. IEE2ST does not model the notch filter of IEEEST but includes
the same limiting and blocking of the output signal.
16.7.6 Model ST2CUT

Model ST2CUT is identical to IEE2ST except in the way in which blocking the output is
achieved. ST2CUT assumes the stabilizer is either directly wired to the exciter setpoint or
that an operator adjusts the input voltage setpoint signal to the stabilizer. Output of the sta-
bilizer is cut off when terminal voltage differs from the initial voltage by the user-speci-
fied values of VCU and VCL. VCL is normally input as a negative value.
16.7.7 Model PTIST1

PTIST1 models the microprocessor-based power system stabilizer as built by PTI. It mod-
els the inputs as derived from potential and current transformers. The algorithms convert
these sampled values into a stabilizing signal.
16.7.8 Model PTIST3

PTIST3 is an extension of the PTIST1 model for the PTI microprocessor-based stabilizer.
The updated model includes:
• Third lead-lag function.
• Two-stage torsional filter.
• Tap-averaging feature.
• Time-at-a-limit function.
• Analog output option.
• Additional signal limiters.
Each of these model additions is discussed below.
16.7.8.1 Third Lead-Lag

Two CONs and one VAR (for the STATE or STORE) have been added to accommodate
this feature. If the user wishes to disable the third lead-lag, then CON(J+13), T6 must be
set identically to zero.
16-35
16.7.8.2 Torsional Filter

The torsional filter includes two second-order stages. Six CONs and two VARs (storage
locations) have been added to accommodate each stage. If the user wishes to disable the
torsional filter, then CON(J+19), B2, must be set identically to zero. In this event both
stages are bypassed. If B2 is equal to zero, then values specified for any of the other tor-
sional filter CONs do not matter. However, if the filter is enabled, then both stages are
active and CON(J+25), B5, must also be greater than zero to avoid a divide-by-zero prob-
lem.
The actual stabilizer controlling software, except for the torsional filter, has been designed
to mimic an analog controller. The torsional filter has been implemented directly in the
digital domain (z-plane discrete time filter) via a first direct digital structure. Therefore the
torsional filter has been coded into the model using z-form for state-space dynamics as
well as extended term.
16.7.8.3 Tap-Averaging
The tap averaging function keeps a running average of the last ICON(M+1), NAV, control
outputs (clicks) up to a maximum of 16. This feature can be disabled by setting
ICON(M+1) equal to 1 (i.e., average 1 output). The last ICON(M+1) outputs are stored in
the output averaging table, VAR(L+22) through VAR(L+37). The sum of the last
ICON(M+1) outputs is stored in the output accumulator VAR(L+38). The averaging fea-
ture was added to reduce tap chattering under steady-state conditions.
If the control output at any given click exceeds the threshold specified in CON(J+26),
Athres, the averaging function will be automatically bypassed, and that output signal will
be passed directly to the limit function. This feature has been added such that under tran-
sient conditions the stabilizer reverts to normal operation.
When the absolute value of output signal is beyond the threshold, the output averaging
table entries and accumulator are zeroed out. When the signal falls below the threshold,
the table is rebuilt starting with the first signal within the threshold. If the output averaging
is enabled, then the threshold value would probably be specified in the range from 0.002
to 0.02.
16.7.8.4 Time-at-a-Limit
The time-at-a-limit function performs two options:
1. Counts the number of clicks (control outputs) that the output signal continuously
exceeds a user-defined threshold as set in CON(J+29), Lthres.
2. After the output signal has continuously exceeded the threshold for a number of user-
defined counts as specified by ICON(M+2), NCL, the function ramps the tap position to
nominal or the analog output to zero over a user-specified number of counts as defined
in ICON(M+3), NCR.
This feature is designed to prevent the stabilizer from remaining at its extreme tap position
for a prolonged period of time. The minimum control output value (absolute) that drives
the autotransformer to its maximum tap position is 0.1.
The user should be aware that the model counts the number of clicks (or control outputs)
such that if the user-defined simulation time step is greater than the physical sampling rate
16-36
(0.025 sec), the model may produce timing errors. However, it is expected that the time-at-
a-limit function would only play a role during local disturbances when the user should be
in small time step mode.
16.7.8.5 Analog Output Option

ICON(M), ISW, controls the output option in force. If ICON(M) is set equal to zero, the
stabilizer will operate in normal mode, where the output control signal is used to select an
autotransformer tap position, which, in turn, modulates the feedback terminal voltage
(digital output option). If ICON(M) is set equal to one, the stabilizer will operate in analog
mode, where the output control signal is fed into the exciter as a supplementary control
signal.
16.7.8.6 Additional Signal Limiters

CON(J+30), Pmin, can be specified such that if the machine power drops below this value
the stabilizer will instantaneously move to nominal tap or zero analog output and remain
there until the condition clears. CON(J+27) is the digital signal limiter such that the con-
trol output may be clipped to the value specified for DL prior to selecting a tap position.
CON(J+28), AL, is the corresponding output signal limiter for the analog output option.
16-37
Models MNLEX1, MNLEX2 and MNLEX3 Program Application Guide: Volume II
16.7.9 Model STBSVC

Model STBSVC provides a supplementary signal for the WSCC static var compensation
model CSVGN5. STBSVC receives one or two signals as input and its output, from the
model’s transfer function, is placed in the array VOTHSG, which is used by the static var
models. The first signal is required and for this signal the user may choose electrical
power, bus frequency, or accelerating power. The signal of electrical power is the per unit
flow of a branch or the total flow of parallel branches. The signal of accelerating power is
obtained from a remote machine. An optional second signal is derived from a remote bus
voltage, vars from the SVC into the system, the SVC current into the system; vars can be
obtained from a remote machine or from a branch. The values of these signals are per unit,
scaled by the system base quantities.
The choices for the signals are indicated from the values assigned to the ICONs in the
model. A value in ICON(IC) selects the quantity for the first signal as noted on the data
sheet. If there is an error, i.e., ICON(IC) is outside the range as listed on the data sheet, the
model will print out an error message and change the value for ICON(IC) to 3 for bus fre-
quency. A value in ICON(IC+1) selects the quantity for the second signal. By setting
ICON(IC+1) to 0 no signal will be obtained and the model will derive its stabilizing signal
output from just the first input signal. If the value entered for ICON(IC+1) is outside the
range as listed on the data sheet, then the model will print out an error message and set
ICON(IC+1) to 0 for no signal.
The other four ICONs indicate the source of the signals. ICON(IC+2), ICON(IC+3), and
ICON(IC+4) indicate the equipment source of the first input signal. For accelerating
power and bus frequency deviation, the remote bus number is stored in ICON(IC+2). The
remote machine ID number is stored in ICON(IC+4). For power flow between for a
branch, ICON(IC+2) is the from bus number; ICON(IC+3) is the to bus number; and
ICON(IC+4) is the circuit number. If set to –1, power flow is the total of the parallel
branches. A value stored in ICON(IC+5) is the remote bus number providing the bus volt-
age is obtained. By setting ICON(IC+5) to 0, the bus voltage will produce an error mes-
sage.
16.7.10 PSS2A
Model PSS2A, like IEE2ST, is a dual-input stabilizer. This model can represent a variety
of stabilizers with inputs of power, speed, or frequency.
For each of the two inputs, two washouts can be represented along with a transducer time
constant. The indices N and M allow a ramp-tracking or simpler filter characteristic to be
represented. Phase compensation is provided by the two lead-lag or lag-lead blocks.
16.8 Models MNLEX1, MNLEX2 and MNLEX3

Minimum excitation limiters or under-excitation volt-ampere limiters are provided in
excitation systems to increase EFD during high voltage to maintain steady-state stability.
The models listed in Table 4 were developed under contract for the Consolidated Edison
Company of New York. These different minimum excitation system characteristics are
represented and can be combined with excitation systems IEET1B. Care should be taken
16-38
Program Application Guide: Volume II Models MNLEX1, MNLEX2 and MNLEX3
when using these models to assure coordination between normal excitation control
through the voltage comparator and excitation control by the proper minimum excitation
limiter. Coordination is normally achieved by these methods, but currently only with exci-
tation system model IEET1B:
• Summing the output of the minimum excitation limiter with the output of a voltage error
function that is limited. Either the voltage comparator or the minimum excitation limiter
will then normally be on its limit.
• Feeding voltage comparator and minimum excitation limiter outputs into a high value
gate.
The output of these limiters is placed in the VOEL array.
Table 4: Models Developed for Consolidated Edison Company of New York
Model Units MEL Manufacturer

MNLEX1 Ravenswood 3 Allis Chalmers
MNLEX2 Poletti Westinghouse
Ravenswood 1 and 2
MNLEX3 General Electric
Arthur Kill
16-39
Models MNLEX1, MNLEX2 and MNLEX3 Program Application Guide: Volume II
16-40
Chapter 17
Speed Governor System
Modeling
Chapter 17 - Speed Governor System Modeling
17.1 Overview
The turbine-governor models are designed to give representations of the effects of power
plants on power system stability. They are not, however, intended to be used in studies of
the detailed behavior of individual plants. A functional diagram of the representation used
and its relationship to the generator is shown in Figure 17-1.
Because of the wide variety in the details of individual turbine controls, the PSS®E mod-
els do not attempt to give a high degree of exactness for any given plant; rather they repre-
sent the principal effects inherent in conventional steam turbine, gas turbine, nuclear, and
hydro plants.
Speed
Turbine-Generator
Inertia
Speed
Speed Control Governor-Controlled Turbine
Governor Mechanism Valves or Gates Mechanical
Power
Speed Governing System Turbine and

Energy System
Figure 17-1. Speed Governor and Turbine in Relationship to Generator
17-1
Speed Governor System Modeling PSS®E 33.4
Turbine-Governor Data Program Application Guide: Volume II
17.2 Turbine-Governor Data

The turbine-governor status is determined, as for generators, by the following criteria:
• Online when the terminal bus type is 2 and the machine status is 1.
• Off-line when the terminal bus type is 1 or 4 or machine status is 0.
All turbine-governor data should be specified on the same base used for the generator.
Hence, if all machine data is entered on each machine’s individual base, the governor per-
manent droop, R, of all units should be set equal to ensure equal load sharing on all units.
The turbine damping factor, Dturb, is equal to Pturbine/speed, in per unit, on a base equal to
unit rating. For steam, nuclear, and gas turbines, Dturb can normally be taken as zero; for
hydro turbines its value normally lies between 0 and 0.5 pu.
17.3 Model TGOV1

TGOV1 is a simple model representing governor action and the reheater time constant
effect for a steam turbine. The ratio, T2/T3, equals the fraction of turbine power that is
developed by the high-pressure turbine. T3 is the reheater time constant, and T1 is the gov-
ernor time constant. Representative values for typical units are:
R = 0.05 pu
T1 = 0.50 sec
T2/T3 = 0.3
T3 = 5 to 9 sec
17.4 Gas Turbine-Governors

The general subject of gas turbine modeling for power system transient stability analysis
is described in a paper by W.I. Rowen, "Simplified Mathematical Representations of
Heavy-Duty Gas Turbines," ASME 83-GT-63, Engineering for Power, October 1983, pp.
865.
17.4.1 Model GAST

GAST represents the principal dynamic characteristics of industrial gas turbines driving
generators connected to electric power systems. Speed variations from nominal are
expected to be small (approximately 5%). The model consists of a forward path with
governor time constant, T1, and a combustion chamber time constant, T2, together with a
load-limiting feedback path. The load limit is sensitive to turbine exhaust temperature, and
T3 represents the time constant of the exhaust gas measuring system.
The ambient temperature load limit [CON(J+4)] should be set to unity, when the turbine is
operating at design ambient temperature. At a higher ambient temperature, it should be set
to a lower value, as prescribed by manufacturer’s data. Hence, the dynamic behavior of
the turbine changes with ambient temperature. The constant, KT, is used to adjust the gain
of the load-limited feedback path.
17-2
PSS®E 33.4 Speed Governor System Modeling
Program Application Guide: Volume II Gas Turbine-Governors
The load reference, [VAR(L)], is set equal to shaft power PMECH, when the model is ini-
tialized during activity STRT. The load-limited feedback path only controls fuel flow to
the gas turbine through the low valve gate, when its output is lower than the original load
reference (decremented by the droop signal, 1/R). The damping coefficient, Dturb, is used
to represent speed damping introduced by the gas turbine rotor.
A distinction exists between the maximum fuel valve opening, VMAX[CON(J+6)], and
the ambient temperature load limit. The maximum fuel valve opening is an operational
control and may be adjusted by the operator to allocate load within a plant. The ambient
temperature load limit is a turbine design parameter, and load-limiting feedback serves a
protective function. Representative data for a typical gas turbine generator unit are listed
here:
R = 0.05 pu
T1 = 0.4 sec
T2 = 0.1 sec
T3 = 3.0 sec
VMAX = 1.0 pu
VMIN = – 0.05 pu
Ambient temperature load limit = 1.0 at 80F (rated); 0.9 at 105F
KT = 2.0
Dturb = 0.0
17-3
Hydro Governor Models Program Application Guide: Volume II
17.4.2 Model GAST2A

GAST2A has a more detailed representation of gas turbine dynamics than GAST. As with
GAST, it is intended for operation near rated speed. The speed governor can be configured
for droop or isochronous modes of operation by selecting 1 or 0 for CON(J+3). The tem-
perature controller assumes control of turbine power when exhaust gas temperature
exceeds its rated value. The reference cited in Section 17.4 can be used for obtaining esti-
mates of model constants if actual data is not available.
17.4.3 Model GASTWD

GASTWD has the same detail representation of gas turbine dynamics as GAST2A. The
governor system for this model is based on a Woodward governor consisting of an electric
speed sensor with proportion, integral, and derivative control.
17.4.4 Model WESGOV

WESGOV may be used to model the Westinghouse 501 combination turbine-governor.
This gas turbine model has discrete cycle sampling times, TP and TC. T1 and T2 repre-
sent the valve, piping, and combustion time constants. Tpe is the electric power measure-
ment time constant.
17.5 Hydro Governor Models

Several models are available for hydro electric plant simulation. Although they have simi-
lar governor models, they differ in hydraulic system representation:
• Linear models (IEEEG2, IEEEG3, IEESGO, WPIDHY, and HYGOV2).
• Nonlinear models (HYGOV, HYGOVM, WEHGOV).
17.5.1 Linear Models

Linear models assume the following penstock/turbine transfer function:
p 1 - TW × s
g = 1 + TW × s/2
(1.1)
where p and g are per unit mechanical power and gate position, respectively.
The water column constant, TW, is given, approximately, by:
L×Q
TW =
gv × A × H
(1.2)
where:
Q = Flow at initial loading level.
H = Head at initial loading level.
L = Centerline length of penstock plus scroll case plus draft tube.
gv = Gravitational acceleration.
17-4
Program Application Guide: Volume II Hydro Governor Models
A = Penstock cross-sectional area.
Because Q at half load is about half its full load value, while H remains fairly constant, TW
varies significantly with loading level.
These models are valid only for the small deviations of frequency and gate position that
are typical in large power systems. They also require that the user recalculate the value of
TW for each new initial loading level.
17-5
17.5.2 Model HYGOV

HYGOV represents a straightforward hydro electric plant governor, with a simple hydraulic
representation of the penstock with unrestricted head race and tail race, and no surge tank.
The hydraulic and governor models are shown in Figure 17-2.
1 e 1 + T rs c 1 g
VAR(L) + 
(nref) 1 + T fs rTrs 1 + Tgs
– Velocity and
Position Limits
Speed +  SPEED
(SPEED)
+
R Dturb
X
–
g 1 q +
 X –   X At  PMECH
Tws +
q h –
+
1. q
NL
R = Permanent droop f = Per unit flow

r = Temporary droop h = Per unit Head
Tr = Governor time constant q = No power flow
NL
tf = Filter time constant At = Turbine gain
Tg = Servo time constant Dturb = Turbine damping
g = Per unit gate opening
Figure 17-2. Hydraulic and Governor Models
17-6
Linearization of the penstock/turbine transfer function for small perturbations around a

Q0, H0 operating point results in:
p 1 - TW × s
g = 1 + TW × s/2
(1.3)
where:
Q0
T W = TW ×
H0 (1.4)
TW is calculated as in (16.2) but uses per unit base flows and heads that are independent of
initial loading level. Base flow is turbine flow when gates are fully open (g = 1 pu). Base
head is head on turbine when the flow is base flow. Q0 and H0 are per unit of the base flow
and head, respectively. By multiplying the water time constant by Q0 and 1/H0, the model
automatically accounts for dynamic changes in its effective value. The penstock/turbine
model is valid for the full range of hydro turbine operation from speed no load to maxi-
mum gate opening. It is also valid for large speed deviations and can be used to simulate
load rejection overspeed conditions if no relief valve or jet deflector action is expected.
The governor model is valid for dashpot-type mechanical governors (e.g., Woodward,
English Electric), and for dashpot-equivalent electrohydraulic governors (e.g., ASEA). No
acceleration governing (derivative action) term is included because this is used only in
specialized situations in most interconnected power systems.
The permanent droop, R, and temporary droop, r, are specified per unit on a base equal to
unit rating. The velocity limit, VELM, is the reciprocal of the time taken for the gates to
move from fully open to fully closed. The maximum gate limit, GMAX, is equal to the
gate limit setting as established by the operator at the governor console; it cannot exceed 1
pu. The minimum gate position is normally zero. The no power flow, qNL, is the flow
required to maintain rated speed with the unit off-line; qNL is expressed in per unit of base
flow. The turbine gain, At, is given by:
1
gFL - gNL
(1.5)
where:
gFL = Full load gate (0 < gFL < 1).
gNL = No load gate (0 < gNL < 1).
17-7
Representative hydro plant data values are:

R Permanent droop 0.05 pu
r Temporary droop 0.1 to 1.0 pu
Tr Governor time constant 2 to 20 sec
Tf Filter time constant 0.05 sec
Tg Gate servo time constant 0.5 sec
VELM Gate velocity limit 0.167 pu/sec
GMAX Maximum gate 1.0 pu
GMIN Minimum gate 0.0 pu
TW Water time constant 1.0 to 3.0 sec
At Turbine gain 1.2
Dturb Turbine damping coefficient 1.5 to 2.0 for Pelton (bucket)
0.0 for Kaplan (blade)
0.5 for Francis
q
NL No power flow 0.08 pu
Values of r and Tr should be set to give stable isolated load governing. A guideline for set-
ting r and Tr is:
TW
Tr = 4TW r =
H (1.6)
These guideline settings should be reviewed for each individual unit.
17.5.3 Model HYGOVM

In hydro plant layouts where a long supply conduit is required, it is fairly common prac-
tice to use a surge tank. The purpose of the surge tank is to provide a degree of hydraulic
isolation of the turbine from the head deviations generated by transients in the longest por-
tion of the conduit. Many surge tanks also include an orifice where head loss serves to dis-
sipate the energy of hydraulic oscillations and to produce a damping effect. The hydraulic
system model in HYGOVM is designed to allow detailed simulation of the representation of
the surge chamber system:
• Penstock dynamics.
• Surge chamber dynamics.
• Tunnel dynamics.
• Penstock, tunnel, and surge chamber orifice losses.
• Surge chamber level beyond maximum or minimum alarm.
Penstock dynamics are largely determined by the upper loop in Figure 17-3. The loop gain
is proportional to the inverse of the square of gate position and thereby increases signifi-
17-8
cantly for small openings. Under load rejection conditions, near total gate closure, the loop
effective time constant will tend to zero.
The model cannot handle low time constants without incurring numerical instability. It
deals with this problem by assuming an algebraic solution (i.e., an instantaneous response,
just before numerical instability would occur). This change in model response can be visu-
alized by an instantaneous drop in turbine head (VAR(L+1)) to values close to the head at
the surge chamber base (VAR(L+4)). At the time the algebraic solution is applied, power
and flows at the penstock are negligible and will not affect governor or surge chamber
studies. The governor system used in the HYGOVM model (Figure 17-4) is based on the
HYGOV governor representation, but has these additional features:
• Separate maximum opening and closing gate rate limits. The maximum gate closing
rate (MXGTCR) is usually a compromise between maximum scroll case head, surge
tank overflows, and unit overspeed under load rejection. A representative value is
0.125 pu/sec. The maximum gate opening rate (MXGTOR) determines minimum surge
chamber levels when accepting load. A value of +0.1 pu/sec is representative.
• Buffered opening and closing rates when gate opening is near full closure. Buffering
the gate closure may produce a reduction in overpressures under load rejection. This
feature will reduce impact loadings on the gate linkage and limit the magnitude of the
pressure pulsations that occur while the gates are fully closed during the decay of load
rejection overspeed. A representative value for the maximum buffered closing gate rate
(MXBGCR) is –0.05 pu/sec and 0.15 pu for the buffer limit (BUFLIM). The maximum
buffered opening rate (MXBGOR) is normally equal to MXGTOR.
• Pressure regulator (relief valve) simulation. This regulator is a bypass, generally
attached to the turbine casing. It is operated directly from the governor or the gate
mechanism of the turbine. The amount of water bypassed is sufficient to keep the total
discharge through the penstock fairly constant, hence controlling pressure rise. The
maximum relief valve opening (RVLMAX) can be set equal to GMAX. For the water-
wasting type, the maximum relief valve closing rate (RVLVCR) should be set to
0. pu/sec; for the water-saving type, a representative value for RVLVCR is –1/70
pu/sec.
• Jet deflector simulation. Long penstock impulse turbines are not allowed rapid reduc-
tions in water velocity because of the pressure rise that would occur. To minimize the
speed rise following a sudden load rejection, a governor-controlled jet deflector is nor-
mally placed between the needle nozzle and the runner. The governor moves this
deflector rapidly into the jet, cutting off the load. Typical values for maximum jet
deflector opening and closing rates (MXJDOR and MXJDCR) are +0.5 and 0.5 pu/sec,
respectively.
Turbine characteristics in HYGOVM are defined based on rated conditions:
1. Rated power = CON(J+1).
2. Rated flow = CON(J+2).
3. Rated head = CON(J+3).
4. Gate opening at rated operating point = CON(J+4).
5. No power flow = CON(J+5) ×CON(J+2).
17-9
HSCH
SCHARE
(V)
HLAKE SURGE
(V) QSCH CHAMBER
TUNNEL
TUNL/A, TUNLOS
QTUN SCHLOS
HBSCH PENSTOCK
PENL/A, PENLOS
QPEN
TURBINE HTAIL
(V)
Q2PEN
X
At PENLOS
QPEN2At +
Gate +
Relief Valve O2 O2 gv QPEN
INPUT X     s × PENL/A OUTPUT
+ – +
+ –
HBSCH HTAIL
HBSCH –
HLAKE
 –
S +
+
HSCH +
TUNLOS
gv
1
SCHLOS s SCHARE
s × TUNL/A Q2TUN
X
Q2SCH
X QSCH
 +
QTUN
QPEN –
LEGEND:
gv Gravitational acceleration At Turbine flow gain
TUNL/A Summation of length/cross section of tunnel O Gate + relief valve opening
SCHARE Surge chamber cross section HSCH Water level in surge chamber
PENLOS Penstock head loss coefficient QPEN Penstock flow
TUNLOS Tunnel head loss coefficient QTUN Tunnel flow
FSCH Surge chamber orifice head loss coefficient QSCH Surge chamber flow
PENL/A Summation of length/cross section of penstock,
scroll case and draft tube
Hydro Turbine Governor Lumped Parameter Model
Figure 17-3. Penstock Dynamics Loop
17-10
Jet Deflector
MXJDOR
+ 1 1
 s
Tg Deflector
– Position
+ MXJDCR
0.01 
+
Governor Gate Servo
Speed GMAX
MXGTOR or
MXBGOR
–
Speed + 1 1 + Trs + 1 1
 
Reference 1 + T fs r Trs Tg s Gate
Opening
– –
MXGTCR or
GMIN MXBGCR
R –
+ RVLMAX
RVLVCR 
1
s Relief Valve
Opening
0
Relief Valve
LEGEND:
R Permanent droop MXBGCR Maximum buffered gate closing rate
r Temporary droop GMAX Maximum gate limit
Tr Governor time constant GMIN Minimum gate limit
Tf Filter time constant RVLVCR Relief valve closing rate
Tg Servo time constant RVLMAX Maximum relief valve limit
MXGTOR Maximum gate opening rate MXJDOR Maximum jet deflector opening rate
MXGTCR Maximum gate closing rate MXJDCR Maximum jet deflector closing rate
MXBGOR Maximum buffered gate opening rate
Figure 17-4. HYGOVM Governor System
17-11
The following parameters are calculated by the model:

PRated
Kt (Turbine Power Gain) =
(QRated - QNo load) × HRated × MVABase
(1.7)
QRated
Tfg (Turbine Flow Gain) =
(GRated ×  HRated)
(1.8)
Turbine power is a function of turbine flow and turbine head, which, in turn, are a function
of penstock flow, gate position, and relief valve or jet deflector position. For turbines with
a relief valve:
Qpenstock × Gate Opening
Turbine Flow =
Gate Opening + Relief Valve Opening (1.9)
(Qpenstock)2 × At
Turbine Head =
(Gate Opening + Relief Valve Opening)2 (1.10)
For turbines with a jet deflector:

Turbine Flow = Qpenstock × MIN (1., Jet Position/Gate Opening) (1.11)
Qpenstock2 At
Turbine Head =
(Gate Opening)2 (1.12)
For turbines with neither relief valve nor jet deflector:

Turbine Flow = Qpenstock (1.13)
Qpenstock2 At
Turbine Head =
(Gate Opening)2 (1.14)
Turbine Power = Kt × Turbine Head

× (Turbine Flow - No Power Flow) - Damping (1.15)
Damping = DAMP × per unit speed deviation

× MIN (Jet Position, Gate Position)
(1.16)
where DAMP is DAMP1 for overspeeds under RPM1, DAMP2 for overspeeds above
RPM2, and linearly interpolated for overspeeds between RPM1 and RPM2. The
HYGOVM model should be used for dynamic analyses of hydro plants when the time
range of interest is comparable to the surge tank natural period. For shorter time periods,
the simpler HYGOV model can be used, unless relief valve or jet deflector action is
expected.
17-12
Surge Tank Natural Period = (1.17)
Figure 17-5 shows the result of simulating a 0.1-pu step load increase on an isolated hydro
plant using both HYGOV and HYGOVM. A surge tank natural period is 3 min. For the
normal 3- to 5-sec transient stability time frame of interest, simulation results are almost
identical. For longer simulation times, surge tank level starts falling, and mechanical
power recovery lags behind that of the HYGOV model, which assumes an infinite surge
tank. Detailed conversion of the HYGOVM data into HYGOV format follows these steps:
Flow Base (STATE(K+3) in HYGOV) = (HLAKE - HTAIL)

 PENLOS + TUNLOS + 1/(Tfg)2
(1.18)
Head Base (VAR(L+1) in HYGOV) = HLAKE - HTAIL - Flow Base2

× (PENLOS + TUNLOS) (1.19)
TW (CON(J+8) in HYGOV) = PENL/A × Flow Base

Head Base × Gravity (1.20)
PRated × Head Base × Flow Base

At (CON(J+9) in HYGOV =
MVABase × QRated × HRated × (1 - QNo Load)
(1.21)
Gravity = 9.81 or 32.21, depending on whether metric or English units are used. For lon-
ger term analyses, for surge chamber dynamics analyses, and for load rejection analyses
involving relief valve or jet deflector action, the HYGOVM or HYGOVT models should
be used.
17.5.4 Model WEHGOV

WEHGOV is a model of a Woodward Electronic hydro-governor with proportional, inte-
gral, and derivative control. The turbine is represented by a nonlinear model for the pen-
stock dynamics in a similar fashion as HYGOV, but the model includes look-up tables to
allow the user to represent nonlinearities in the flow versus gate position and mechanical
power versus flow during steady-state operation.
The model allows for the use of three feedback signals for droop:
• Electrical power.
• Gate position.
• PID output.
17-13
SHORT-TERM VERSUS MID-TERM GSTR/GRUN HYDRO GOVERNOR TEST

0.7 INITIAL PER UNIT LOADING - 0.1 PU STEP APPLIED AT T=0SEC
POWER
TECHNOLOGIES,
08:05
INC.R
WED JAN 03, 1996

SHORT- VS. MID- TERM MDL
CHNL# 2: [PMEC 201 [HYDRO 22.000] MC 1]
1.0000 FILE: hygovm 0.50000
CHNL# 1: [SPD 201 [HYDRO 22.000] MC 1]
0.02500 FILE: hygovm -0.0750
1.0000 FILE: hygov 0.50000
CHNL# 1: [SPD 201 [HYDRO 22.000] MC 1]
0.02500 FILE: hygov -0.0750
20.000
18.000
16.000
14.000
12.000
HYGOVM
HYGOV
TIME(SECONDS)
10.000
8.0000
6.0000
Mechanical
Power
Speed
4.0000
2.0000
0.0
Figure 17-5. Short-Term versus Midterm GSTR/GRUN Hydro Governor Test
17-14
To specify which signal will be used, refer to the following:

Feedback Signal CON(J) CON(J+1) ICON(M+)
Electrical power 0 Value for droop 0
Gate position Value for droop 0 0
PID output Value for droop 0 1
PID Controller
The derivative controller has a time constant to limit the derivative characteristic beyond a
breakdown frequency to avoid amplication of high-frequency noise. This time constant is
labeled TD (i.e., CON(J+6)), and its value is typically 0.10 sec.
The PID controller also has a limiter on the integral control to prevent windup when the
gates are at their limit. The gate position limits are set in GMAX for maximum and GMIN
for minimum. For the integral controller there is another variable called DICN,
CON(J+18) that allows the integral controller to advance beyond the values for the gate
limits. The maximum limit for the integral controller is:
GMAX + DICN
and the minimum limit is:
GMIN – DICN
The value for DICN ranges from 0 to 10% and it is set with the field tuning for the gover-
nor.
Pilot Valve
The output signal of the PID controller is fed into the pilot valve. The pilot valve also has
a set of limits that are similar to that for the integral controller. The maximum limit for the
pilot valve output is:
GMAX + DPV
and the minimum limit is:
GMIN - DPV
The value for DPV, CON(J+17), is typically about 2% to ensure that gate can be fully
opened or closed.
Distribution Valve
The output signal of the pilot valve is fed into the distribution valve. The limits of the dis-
tribution valve define the maximum rates to open or close the gates. These two rate limits
are:
1. GTMXOP (maximum gate opening rate), CON(J+10).
2. GTMXCL (maximum gate closing rate), CON(J+11).
The values for both parameters are in per unit gate position per second. Note that the value
for GTMXCL must be less than 0.
17-15
Turbine Model
The model for the penstock hydraulics is similar to that for HYGOV. However, the turbine
model includes two look-up tables to account for steady-state nonlinearities in the model.
The first table is defined by CON(J+19) through CON(J+28). This set of CONS represent
the water flow through the turbine as a function of gate position. Increasing values starting
with GATE1 and FLOW1 must be entered into this table. The second table is defined by
CON(J+29) to CON(J+48) to represent per unit mechanical power on machine MVA rat-
ing as a function of flow. Again, increasing values starting with FLOW1 and PMECH1
must be entered.
17.5.5 Model HYGOVT

In this model, a traveling-wave solution is applied to the penstock and tunnel (Figure 17-
6).These are divided into 9 to 19 segments, and the characteristics solution method is
applied to the resulting time-space lattice. Boundary conditions and head losses are fully
recognized. For accurate results, the simulation time step should be no larger than:
PENLGTH/(9  PENSPD) (1.22)
where PENLGTH is the penstock length and PENSPD is the penstock wave velocity.
Maximum accuracy is attained when simulation time step is equal to, or a submultiple of:
PENLGTH/(19  PENSPD) (1.23)
Conduit wave velocity, assuming rigid walls and accounting for water compressibility,
alone is 1420 m/sec (4659 ft/sec). This velocity is the maximum that can be physically
attained. Actual conduits do not have rigid walls, a representative value for penstock con-
duits is 1100 m/sec (3609 ft/sec).
For this model, the governor and turbine models are the same as in the HYGOVT model.
The decision to use the inelastic (HYGOVM) or elastic (HYGOVT) models relies on the
hydraulic system characteristics and the study scope. Because of time-step constraints
(16.22), traveling-wave simulation turnover may be penalized by the need to use a smaller
time step than would otherwise be required with the inelastic model. However, some error
is involved with the use of an inelastic model. This error can be quantified by the differ-
ence between the elastic and the inelastic head/flow transfer functions in the frequency
domain. This difference, per unit of the elastic case, is approximately:
2
Tp × s 2
-
3 (1.24)
where Tp is the penstock wave travel time (PENLGTH/PENSPD), and s is the Laplace
operator. Tp is typically 0.5 sec, but can be as high as 1.5 sec for long penstocks. For nor-
mal governor action, speed loop crossover, i.e., the dominant mode, occurs at about 1/2
TW rad/sec. With TW being typically 1 to 2 sec, s is in the order of 0.25 to 0.5 rad/sec. The
difference between elastic and inelastic response will usually be negligible, unless very
long penstocks are studied. A critical case run using both model assumptions may prove to
be the easiest way to assess this difference.
17-16
HSCH
SCHARE
(V)
HLAKE
SURGE
(V) QSCH CHAMBER
TUNNEL
QTUN SCHLOS
HBSCH
PENSTOCK
QPEN
TURBINE HTAIL
(V)
Time
TUNNEL
(TUNLGTH, TUNSPD, TUNARE, TUNLOS)
Tunnel Inlet
Constraints
Surge Chamber
Constraints
DELT*ICON(M+3)
Space
Flows VAR(L+46) VAR(L+45 + ICON(M+2)) = QTUN
TUNLGTH/(ICON(M+2)+1)
Heads VAR(L+66) VAR(L+65 + ICON(M+2)) = HBSCH
Surge Chamber
+ 1 +
QTUN  s SCHARE  HBSCH
QSCH HSCH
– +
QPEN X SCHLOS
Q2SCH
Time
PENSTOCK
(PENLGTH, PENSPD, PENARE, PENLOS)
Surge Chamber Turbine
Constraints Constraints
DELT*ICON(M+1)
Space
Flows VAR(L+6) = QPEN VAR(L+55 + ICON(M))
PENLGTH/(ICON(M)+1)
Heads VAR(L+26) = HBSCH VAR(L+25 + ICON(M))
Figure 17-6. Traveling-Wave Model HYGOVT
17-17
Model IEESGO Program Application Guide: Volume II
Figure 17-7 shows a case where an isolated hydro power plant with a TW of 1.83 sec and a
Tp of 0.42 sec is subject to a 0.2-pu increase in load. Except for some transient, high-
frequency effects, the difference between the elastic and inelastic solutions is negligible.
There are times, however, when traveling wave analysis is essential. Overpressures due to
load rejection are critical just before or at gate closure time, the ensuing pressure pulsa-
tions occur after the gate is totally closed. A closed or almost closed gate results in infi-
nitely small penstock time constants, and infinitely large s values in (1.24).
Figure 17-8 shows the simulation of a total gate closure on the same system as in Figure 17-
7. For gate positions at or near total closure, the inelastic simulations of scroll case head
and penstock flows are no longer applicable. Surge chamber levels and tunnel are not
affected by these high-frequency effects.
17.5.6 Model PIDGOV

The PIDGOV hydro turbine governor model represents plants with straight forward pen-
stock configurations and three term electro-hydraulic governors (i.e., Woodard elec-
tronic). This model uses a simplified turbine-penstock model that does not account for
variation of water inertia effect with gate opening. This model can be made to correspond
to other models using the classical turbine-penstock model by setting atw (factor that mul-
tiplies the water inertia time constant) to unity.
The feedback signal used by the governor can either be the gate position or the electrical
power, and can be selected by setting the feedback flag to one for gate position or zero for
electrical power.
The input to this model is the shaft speed deviation, and the outputs are turbine gate posi-
tion and mechanical power.
17.6 Model IEESGO

The IEESGO general-purpose turbine-governor model is included for its compatibility
with other widely used stability programs. With judicious selection of the time constants
and gains, this model gives either a good representation of a reheat steam turbine or an
approximate representation of a hydro plant of simple configuration.
17.7 Model HYGOV2

The nonstandard hydro turbine-governor of HYGOV2 includes the same basic permanent
and temporary droop elements as the standard PSS®E model, HYGOV, but includes a
slightly different representation of the lags within the governor hydraulic servo system and
of the speed signal filtering. The penstock turbine model of HYGOV2 is highly simplified
and is valid only for small deviations of gate position from their initial conditions. Unlike
HYGOV, HYGOV2 requires the user to recalculate the value of the water column time
constants for each new initial loading level. The water column time constants, T5, and T6,
of HYGOV2 are related to the water inertia time constant TW of Section 17.5 by T5 = PoTW
and T6 = PoTW /2. HYGOV2 was developed for a specific plant and should not be used
except in appropriate special situations. In the great majority of situations, HYGOV is to
be preferred.
17-18
Program Application Guide: Volume II Model HYGOV2

POWER
09:46
TECHNOLOGIES,
INC.R
CHNL# 4: [SCHM-HEAD]
325.00 FILE: gstrhygovt 275.00
WED JAN 03, 1996

CHNL# 3: [TURBINE HEAD]
INELASTIC VS. ELASTIC MDL

CHNL# 4: [SCHM-HEAD]
325.00 FILE: gstrhygovm 275.00
20.000
18.000
Surge Chamber Level
16.000
Mechanical Power
Scroll Case Head
14.000
12.000
TIME(SECONDS)
10.000
Inelastic
8.0000
Elastic
6.0000
Elastic
4.0000
Inelastic
2.0000
0.0
Figure 17-7. Elastic versus Inelastic Hydro Governor Models Comparison
— Example 1
17-19
Model TGOV2 Program Application Guide: Volume II

10:11
POWER
TECHNOLOGIES,
INC.R
CHNL# 4: [GATE OPENING]

1.0000 FILE: hygovtall 0.0
WED JAN 03, 1996

ELASTIC VS. INELASTIC MDL

CHNL# 4: [GATE OPENING]
1.0000 FILE: hygovmall 0.0
20.000
18.000
16.000
Head
CaseHead
14.000
Scroll Case
Model
Elastic Model
Scroll
12.000
Elastic
TIME(SECONDS)
10.000
8.0000
6.0000
Power
MechanicalPower
4.0000
Mechanical
Position
2.0000
GatePosition
Gate
0.0
Figure 17-8. Elastic versus Inelastic Hydro Governor Models Comparison

— Example 2
17.8 Model TGOV2

TGOV2 is a fast valving model of a steam turbine that represents governor action, a reheat
time constant, and the effects of fast valve closing to reduce mechanical power. In this
17-20
Program Application Guide: Volume II Model IEEEG1
model, T1 represents the governor time constant, K is equal to the fraction of the turbine
power that is developed by the turbine sections not involved in the fast valving, T3 is the
reheater time constant, and Tt represents the time constant with which power falls off after
closure of the intercept valve. TGOV2 is unique among PSS®E models in that the user
must supply the initiation time for fast valving to begin. This information is supplied in
one of two ways. The first method would be by using activity ALTR to change VAR(L+1)
to the value of TIME at the instant that the user wants the fast valving to begin. STRT sets
VAR(L+1) equal to 9999 so that VAR(L+1) can be changed only after STRT has been
run. The second and more common method of initiating fast valving is by user added
statements in CONEC. As an example, suppose a user wishes fast valving to begin when
the generator speed deviation exceeds 0.01 pu. The user would add the following to
CONEC (refer to Section 19.9.2, Accessing Machine and Load Data of the PSS®E Program
Operation Manual):
IF (VAR(720) .LT. 100. .OR. MODE.NE.2) GO TO 500
CALL GENCHK(500,'1',II,'IN CONEC')
IF (SPEED(II) .GT. CON(10000)) VAR(720)=TIME
500 CONTINUE
where VAR(720) is VAR(L+1) for machine 1 at bus 500, and CON(10000) is set equal to
0.01. Constants TA, TB, and TC define how long it takes to close the intercept valve, how
long it stays closed, and how long it takes to reopen. These three constants use VAR(L+1)
as their reference time after fast-valving is initiated. The user should use caution and in
most cases should not modify VAR(L+1) after fast valve action has commenced.
17.9 Model IEEEG1

IEEEG1 is the IEEE recommended general model for steam turbine speed governing sys-
tems. By the appropriate choice of parameters, this model can be used to represent a vari-
ety of steam turbine systems including nonreheat, tandem compound, and cross-
compound types. IEEEG1 can also approximate the behavior of hydro turbine-governors.
The user should obtain reference b listed in Section 14.2 of this manual for the explanation
of constants. Data for PMAX and PMIN should be specified in per unit on the MBASE of the
high-pressure unit in the case of cross-compound sets.
17.10 Models IEEEG2 and IEEEG3

IEEEG2 and IEEEG3 are alternative representations of hydro turbine speed governing sys-
tems. In some cases, where data may be more easily obtainable for these representations
or because they are more exact, these representations may be preferred over that of
IEEEG1. The water time constant, TW, is generally in the range of 1 to 3 sec. Typical val-
ues for IEEEG3 would be:
TR = 5.0 sec.
TG = 0.5 sec.
Tp = 0.04 sec.
17-21
d = 0.3.
s = 0.05.
17.11 Model TGOV3

TGOV3, a modification of the IEEEG1 model, is now PTI’s recommended model for fast
valving studies. Data selection should be based on the same definitions used for IEEEG1.
Several additional data items are required by this model. The PRMAX CON has been added.
This value, which limits the maximum pressure in the reheat stage, is generally about 1.1
pu. The TA, TB, and TC time constants along with the initiating of the fast valving, are
identical to those defined for TGOV2. This model does recognize the nonlinearity
between flow and valve position as shown on the data sheet.
17.12 Model CRCMGV

CRCMGV models cross-compound units in other widely used stability programs. The con-
stants used are:
T1 = The control time constant.
T3 = The servo time constant.
T4 = The steam boil time constant.
T5 = The steam reheat time constant.
F = The shaft capacity ahead of the reheater divided by the total shaft capacity.
17.13 Models DEGOV and DEGOV1

DEGOV is a model of an isochronous governor for a diesel engine. DEGOV1 is a model of a
governor for a diesel engine where droop control is used with either throttle or electric
power feedback. These models are based on a Woodward governor consisting of an elec-
tric speed sensor, a hydro-mechanical actuator, and the diesel engine. The output of the
actuator is a valve position of the fuel supply. A typical design of a diesel engine would
limit the fuel input on a per-cycle basis. The amount of energy developed per cycle is
directly proportional to the amount of fuel per cycle. Multiplying energy developed per
cycle by the rotational speed of the engine gives the power being supplied by the prime
mover to the generator. Therefore, the output of the engine is multiplied by the rotational
speed of the generator and the limits for the actuator can be expressed as torque limits.
It should be noted that the DEGOV model represents an isochronous governor and, there-
fore, use of this model should be restricted to diesel generators operating isolated from
other synchronous sources. For a diesel generator online with other synchronous sources
the DEGOV1 model should be used.
To determine the engine dead time, use the following formula:
15 60
DT = +
N Nn
where:
17-22
Program Application Guide: Volume II Model SHAF25
DT = The dead time in sec.

N = The engine speed in rpm.
n = The number of cylinders firing per revolution.
17.14 Model SHAF25

SHAF25 is a shaft spring model for up to 25 masses. The array of constants from J to J+102
must be prepared as shown in the model data sheet for SHAF25. Zeros are entered for
unused parameters. The first two constants, Xd – Xd and Tdo are used in the calculation of
exciter torque. When a shaft torsional system is stimulated by electromagnetic torques in
the generator (or motor), this model should be used in conjunction with model GENDCO
(see Section 14.4.5.5). The modeled mass-spring system is composed of rotational springs
in a radial configuration. Figure 17-9 shows a sample layout.
Exciter Generator Low-Pressure High-Pressure

Turbine Turbine
Figure 17-9. Sample Shaft System
17.14.1 Data Preparation for SHAF25

The following steps should be followed to prepare data:
1. Draw a diagram of shaft to be modeled (see Figure 17-10 with masses circled and shaft
sections boxed). Number masses and shaft sections according to the following rules:
a. Mass 1 must be at one of the ends of the shaft.

b. The remaining masses are numbered consecutively starting from mass1 and going
to the other end of the shaft.
c. Kshaft i-j connects mass i to mass j.
Figure 17-10 shows numbering for a sample layout. Note masses are circled and shaft sec-
tions are boxed.
1-2 2-3 3-4
1 2 3 4
Figure 17-10. Numbered Shaft Layout
17-23
Model SHAF25 Program Application Guide: Volume II
17.14.2 Calculation of Inertias for Masses

The mass inertias used for each component are expected to be entered in units of kW-
sec/kVA. Data supplied in the English system can be converted using the following for-
mula:
0.231 WR2 (rpm)2 10-6
H =
R
where:
H = Inertia constant in kW-sec/kVA.
WR2 = Moment of inertia in lb-ft2 (based on radius).

rpm = Rotational speed of the mass in revolutions per minute.
R = Machine kVA rating.
If given data in the metric system, the formula becomes:

1.37 GD2 (rpm)2 10-6
H =
R
where GD2 = Moment of inertia in kg-m2 (based on diameter). Note that GD2 = 4I, where
I = mass moment of inertia in kg-m2. The formula for H given above can also be written in
a more common form as given below:
2
1 I m
H = --- ------------
2 S
where:
S = Machine VA rating.
m = Mechanical angular speed in radians.
17.14.3 Shaft Stiffness Calculation

The values of Kshaft should be entered in units of per unit torque per electrical radian on
machine MVA base. When the stiffness coefficients of the shafts are expressed in terms of
Newton-meter (Nm) per mechanical radian, the following formulas should be used for
conversion.
Rating
Tbase = m
where:
Rating = Rating of machine, VA.
17-24
Program Application Guide: Volume II Model SHAF25
m = Angular velocity of the massive mechanical radians per second.

e = Angular velocity of the electrical frequency in electrical radians.
K = Stiffness coefficient in Nm per mechanical radian.
Tbase = Shaft mechanical torque base, in newton meters.
Because model results are reported per unit, this torque base value will be valuable to get
information to match typical shaft degradation curves.
17.14.4 Mass Damping

The values at D should be entered as dimensionless damping coefficients, i.e., per unit
torque divided by per unit speed. A value of D = 0.05 indicates that the reaction torque at
the mass damper at rated speed is 0.05 pu.
17.14.5 Data Calculation Check

To check the calculations, check the frequency of one mass and coefficient of stiffness,
before conversion, by one of the following:
Kq
1 =
 WR2
4K
1 =
 GD2
and compare it with after-conversion data using the formula:
Kshaft
2 = 0
 2H
The results should be identical.
17.14.6 Time Step when Using SHAF25

When using SHAF25, the integration time step should be compatible with the highest
shaft natural frequency. As a general guide, the time step should satisfy the following:
2 0.03
T < 0.1 ×  =
smax f smax
where:
smax = Maximum shaft natural frequency, rad/sec.
fsmax = Maximum shaft natural frequency, Hz.
17-25
17.15 Model TGOV4

TGOV4 is a model of a steam turbine and boiler that represents governor action, explicit
valve action for both control and intercept valves, main, reheat and low-pressure steam
effects, and boiler effects. Also incorporated into the model are a power load unbalance
(PLU) relay and an early valve actuation (EVA) relay that, when triggered, will cause fast
valving of the control or intercept valves.
17.15.1 Governor Model

Most of the model’s time constants are similar to those used in other PSS®E governor
models. They are defined according to "Dynamic Models for Steam and Hydro Turbines
in Power System Studies," IEEE Committee Report, IEEE PES 1973 Winter Meeting,
paper T73 089-0 and listed below:
K = The inverse of the governor speed droop.
T 1, T 2 = The governor controller lag and lead time constants.
T3 = The valve servomotor time constant for the control valves.
T4 = The steam bowl time constant.
T5 = The steam reheat time constant.
T6 = The crossover time constant.
K1 = The fraction of the total turbine power developed by the high-pressure turbine.
K2 = The fraction of the total turbine power developed by the
intermediate-pressure turbine.
UO = The control valve open rate limit.
UC = The control valve close rate limit.
UOIV = The intercept valve open rate limit.
UCIV = The intercept valve close rate limit.
PRMAX = The maximum pressure limit in the reheater.
TIV = The valve servomotor time constants for the intercept valves.
R = The speed droop for the intercept valve governor.
Offset = The offset applied to the intercept valve speed control to keep the valves wide open
and inhibit closing action unless speed exceeds the offset. This value is usually about
3%. PSS®E, however, adds this with speed regulated frequency so a value multi
plied by 1/R should be entered (e.g., 0.03 × 20 = 0.6).
The remainder of the time constants are explained in the following sections. Due to the per
unit system employed throughout the model, it is necessary that all machine data be
entered on the unit’s actual machine base. The logic diagram of the PLU and EVA relays
is shown in Figure 17-11, and the block diagram of the model is given in Figure 17-12.
17-26
Program Application Guide: Volume II Model TGOV4
17.15.2 Boiler Model

A boiler model is included to properly reflect the effects of boiler pressure on turbine
power. The boiler constants are:
KP = Boiler proportional control gain.
KI = Boiler integral control gain.
TFuel = Fuel lag time constant.
TFD1, TFD2 = Approximation of deadtime due to the fuel supply system.
Cb = Boiler storage constant.
Kb = Friction drop coefficient.
The effects of boiler pressure controls can be neglected by setting KP are KI to zero. How-
ever, drum pressure will still be affected by changes in main steam flow; Cb cannot be
zero.
17-27
Uo 1
+ 1 Pmech

1
T3 s
K1 + +
– 
Uc 0
1
m· SIP +
Uo
K2
+ 1 1
 T3 1 – K1 – K2
s
– PRMAX m· SIP
+
Uc 0 +
+
– m· SLP
K(1 + sT2) – 1 1 1
  KC Uo 1  KCAL    1 + sT6
(1 + sT1) 1 + sT4 T5s
+ + 1 + + Reheat
1 IV Flow Area
 T3 s Pressure
Load Flow + +
Area 
Reference –
Uc 0 Boiler
1 Pressure
Uo
+ 1 1
 T3 s 1 1
– s s
Uc 0 UCIV UCIV
Pressure – +
 1 UCIV 1 UCIV
Reference TIV TIV
–
 Kb X
Pdrum+ – –
 
1
Cbs + +
K 1 1 1 +
K p + ------I
–
1 + sTFuel 1 + sTFD1 1 + sTFD2 
s KIV
– 1
 R 
+
IV Load +
Reference 
+
Offset
Figure 17-11. Model TGOV4 Logic Diagram for PLU and EVA
17-28
Generator Power TREVAs Y

EVA > Rate
1 + TREVAs Level
TIV1 IV#1
AND
OR
–
TIV2 IV#2
EVA > Y
 Unbalance
Level
+
Reheat Pressure
TCV1 CV#1
Generator Current TRPLUs Y

PLU > Rate Timer
1 + TRPLUs Level
TCV2 CV#2
AND LATCH
TCV3 CV#3
+
PLU
– Y
 > Unbalance
Level
TCV4 CV#4
N
Figure 17-12. Model TGOV4 Block Diagram for PLU and EVA
17.15.3 Valve Model

Up to four control valves and two intercept valves can be modeled as either linear or non-
linear. The characteristics are defined by specification of a data point on the curve. The
characteristics are specified separately for the control and intercept valves. The shape of
the valve characteristic is defined by specifying the value of the per unit valve position to
get a flow area of 80% of the total flow area (0.8 pu). Figure 17-13 shows a plot of flow
area versus valve position for several different characteristics. A typical value of valve
position at 80% flow area would be 0.3. To specify a linear valve characteristic, a value of
0.8 should be given.The flow area is calculated from the valve position using the equation:
Flow Area = K(1 - e-A(Position))
The constants K and A are calculated in activity STRT and stored in VARs. Specification
of a valve position at 80% that is less than 0.2 may cause a small error in the calculation of
17-29
flow area at small values of valve position. Valves that are nearly linear (valve position at
80% very close to 0.8) should be modeled as linear.
The valve position demand may also be modeled as linear or nonlinear. A nonlinear model
would be used in cases where nonlinear demand control is used to cancel-out the nonlinear
valve characteristics to get a linear steady-state relationship between load reference and
flow area. Figure 17-14 shows a plot of valve position demand versus control signal. The
curve is defined by specifying the value of valve position demand corresponding to a con-
trol signal of 0.8. Similar to the valve characteristic, specified values less than 0.2 may
cause a small amount of error and valve control close to linear should be modeled as lin-
ear. The valve position demand is calculated using the equation:
Valve Position Demand =
The constants A and K are calculated in activity STRT. The valve position demand can
also be modeled with an offset in control valves 2, 3, and 4 and intercept valve 2. A valve
controller with an offset requires a larger value of control signal to get a given amount of
position demand. For example, an offset of 0.5 would mean that the valve position
demand would be 0 for a control signal less than 0.5, determined by the characteristic
between 0.5 and 1.5, and fully open above 1.5. The offsets must be in ascending order.
The constants KCV and KIV are calculated in activity STRT to normalize the flow area cal-
culation, accounting for the offsets, so that a one per unit control signal results in a one per
unit flow area.
The constant KCAL is used to calibrate the maximum flow area (1.0 pu) to the maximum
mechanical power, per unit on the generator MVA base. This constant can be used to
account for valves that are oversized or undersized compared to the maximum turbine out-
put. For example, if the fully opened valves result in only a mechanical power of 0.9 pu,
KCAL should be 0.9.
17-30
1.0
0.9
0.8
0.2 0.3 0.4
0.7
0.5 0.6
Per Unit Flow Area
0.7
0.6
0.8
0.5
0.4
0.3
0.2
0.1
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Figure 17-13. Per Unit Valve Position
17-31
1.0
0.9
0.8
0.7
Per Unit Valve Position
0.6
0.5
0.8
0.7
0.4
0.6 0.5 0.4 0.3 0.2
0.3
0.2
0.1
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Figure 17-14. Control Signal
17-32
17.15.4 Control of Fast Valving

The value of ICON (IC+2) determines the control mode of fast valving:
ICON(IC+2) Control Mode
0 No PLU or EVA
1 User-controlled PLU/EVA
2 PLU
3 PLU and EVA
In control mode 0, fast valving simply does not occur. Triggering of fast valving in control
modes 2 and 3 is explained in the next section. In control mode 2, the user defines what
time the control valves and/or the intercept valves are to begin closing. The valves close at
their defined closing rate, remain closed for their individually set closure times, and then
reopen at their opening rates. The following constants are used in the manual closure of
the valves:
CON Event
J+33 Time for the control valves to start to close
J+34 Control valve closing rate
J+35-J+38 Time closed for control valves 1 to 4
J+39 Time for the intercept valves to start to close
J+40 Intercept valve closing rate
J+41-J+42 Time closed for intercept valves 1 and 2
17.15.5 PLU Relay

The PLU relay consists of two parts; both must be triggered to initiate valve action. The
first part consists of a comparison circuit that calculates the per unit difference between
reheat pressure (representing mechanical power) and generator current. The second part
calculates the rate of change of current. When the rate of change of current is greater than
the PLU rate limit, a timer is started. The timer start time is stored in a VAR. If the rate of
change of current is still greater than the rate limit when the timer runs out and the differ-
ence between reheat pressure and current is greater than the unbalance limit, the PLU
relay is triggered and fast valving action is initiated. The time of activation is stored in a
VAR for both the control and intercept valves. The control and intercept valves begin
closing after their respective time delays. The valves close at the rate specified for the con-
trol and intercept valves. The time at which control valve number 1 and intercept valve
number 1 close are stored in VARs. Each individual valve stays closed for a time period
defined by the user and reopens according to its defined rate.
At the same time the PLU is triggered, the latch is set. The latch is not reset until the dif-
ference between reheat pressure and current is less than the unbalance level. While the
latch is set, the load references to the control and intercept valves are removed. Also the
control valve load reference is ramped down at a user-defined rate, but only down to a
minimum load reference. The following constants are used in the PLU circuit:
17-33
Model WPIDHY Program Application Guide: Volume II
CON Event
J+33 Delay between PLU trigger and the control valves starting to close
J+35-J+42 Same as for user control
J+39 Delay between PLU trigger and the intercept valves starting to close
J+43 Time constant for rate of change of current
J+44 Rate level
J+45 Length of timer in seconds
J+46 Unbalance level
17.15.6 EVA Relay

The EVA relay consists of two parts; both must be triggered to initiate valve action. The
first part consists of a comparison circuit that calculates the per unit difference between
reheat pressure and generator electrical power. The second part calculates the rate of
change of electrical power. If the rate of change of electrical power is greater than the
EVA rate limit and the difference between reheat pressure and electrical power is greater
than the unbalance limit, the EVA relay is triggered and fast valving action is initiated.
The time of activation is stored in a VAR for the intercept valves. The intercept valves
begin closing after their time delay. The valves close at the rate specified for the intercept
valves. The time at which intercept valve 1 closes is stored in a VAR. Each individual
valve stays closed for a time period defined by the user and reopens according to its
defined rate. The following constants are used in the EVA circuit:
CON Event
J+35-J+42 Same as for user control
Delay between EVA trigger and the intercept valves starting to
J+39
close
J+47 Time constant for rate of change of electrical power
J+48 Rate level
J+49 Unbalance level
17.16 Model WPIDHY

WPIDHY is a model of the Woodward PID hydro governor with proportion, integral and
derivative control. This electric governor has advantages over the hydraulic governor with
temporary droop in that it can provide a faster response. Care should be used in specifying
the derivative gain because studies have shown too much gain will result in excessive
oscillations. This model does allow for the nonlinearity between gate position and actual
mechanical power.
17-34
17.17 Model GGOV1

PSS®E 33.4
The GGOV1 model can be used to represent a variety of prime movers controlled by PID governors. It is suitable, for example, for
representation of:
• gas turbine and single shaft combined cycle turbines,
• diesel engines with modern electronic or digital governors,
• steam turbines where steam is supplied from a large boiler drum or a large header where pressure is substantially constant over the
period being studied, and
• simple hydro turbines in dam configurations where the water column length is short and water inertia effects are minimal.
Per unit parameters for this model are entered on base of Trate (turbine MW rating). If no value is entered for Trate, the generator
MVA base is used.
The range of fuel valve travel and of fuel flow is unity. Thus the largest possible value of Vmax is 1.0 and the smallest possible value
of Vmin is zero. Vmax may, however, be reduced below unity to represent a loading limit that may be imposed by the operator or a
supervisory control system. For gas turbines Vmin should normally greater than zero and less than wfnl to represent a minimum firing
17-35
limit. The value of fuel flow at maximum output must be less than, or equal to unity, depending on the value of Kturb.
The parameter Teng is provided for use in representing diesel engines where there is a small but measurable transport delay between
a change in fuel flow setting and the development of torque. Teng should be zero in all but special cases where this transport delay
is of particular concern.
The parameter Flag is provided to recognize that fuel flow for a given fuel valve stroke, can be proportional to engine speed. This is
Speed Governor System Modeling

the case for GE gas turbines and for diesel engines with positive displacement shaft driven fuel injectors. Flag should be set to unity
for all GE gas turbines and most diesel engines. Flag should be set to zero where it is known that the fuel control system keeps fuel
flow independent of engine speed.
The load limiter module may be used to impose a maximum output limit, such as an exhaust temperature limit. To do this the time
constant Tfload should be set to represent the time constant in the measurement of temperature (or other signal), and the gains of the
limiter, Kpload, Kiload, should be set to give prompt stable control when on limit. The load limit can be deactivated by setting the
Model GGOV1
parameter Ldref to a high value.
The parameter Dm can represent either the variation of engine power with shaft speed or the variation of maximum power capability
with shaft speed. If Dm is positive it describes the falling slope of the engine speed versus power characteristic as speed increases. A
slightly falling characteristic is typical for reciprocating engines and some Aeroderivative turbines. If Dm is negative the engine
power is assumed to be unaffected by shaft speed, but the maximum permissible fuel flow is taken to fall with falling shaft speed.
WSCC Governor Models

Speed Governor System Modeling
This is characteristic of single shaft industrial gas turbines.
This model includes a simple representation of a supervisory load controller. This controller is active if the parameter Kimw is non-
zero. The load controller is a slow acting reset loop that adjusts the speed/load reference of the turbine governor to hold the elec-
trical power output of the unit at its initial condition value (Pmwset). The load reference of the supervisory load control loop (Pmwset)
is given a value when the model is initialized and stored in VAR(L+6). The load controller must be adjusted to respond gently rela-
tive to the speed governor. Setting Kimw to 0.001 corresponds to a relatively slow acting load controller and may be used as a nor-
mal estimate.
The parameters Aset, Ka, and Ta describe an acceleration limiter.
The parameter, db, is the speed governor dead band. This parameter is stated in terms of per unit speed. In the majority of applica-
tions of GGOV1 it is recommended that this value be set to zero.
The parameters, Tsa, Tsb, are provided to augment the exhaust gas temperature measurement subsystem in gas turbines. For exam-
ple, they may be set to values such as 4., 5. to represent the radiation shield element of large gas turbines.
The parameters, Rup, Rdown, specify the maximum rate of increase and decrease of the output of the load limit controller
(Kpload/Kiload). These parameters should normally be set, 99/-99, but may be given particular values to represent the temperature
17-36
limit controls of some GE heavy duty engine controls.
17.18 WSCC Governor Models

Several new governor models were added to version 10 of the WSCC stability program. Some of these models contain provisions

for deadbands, nonlinear gains and governor blocking. This section documents equivalent PSS®E turbine-governor models.
WSIEG1 models the IEEE-recommended general model for steam turbine speed governing systems (IEEEG1) with some nonlineari-
ties.
Models WSHYDD and WSHYGP are special models required for hydro units in WSCC. The WSHYDD is a double derivative hydro
governor. The WSHYGP represents the old WSCC GP governor plus turbine model. A proportional, integral derivative controller is
modeled in WSHYGP.
PSS®E 33.4
17.19 Turbine Load Controllers
Simulations where there is large frequency deviations or power mismatches may require the modeling of the relatively slow reset
controllers that manage the action of the turbine governor. The ULCFB1 model was added to represent the adjustment made to gov-
Program Application Guide: Volume II Turbine Load Controllers
ernor’s speed or load reference. While the proper time scale for the managing controller is
slow in relation to that of the governing and supervising loops, there is, nevertheless, a
wide variations in the speed of response of turbine load reference controllers. At the more
active end of the spectrum, a load controller may be able to completely cancel a deviation
of output within as little as 30 seconds, while a reset time of a few minutes would be com-
mon in large steam plants. The load controllers of gas turbine plants would typically be
quick. The ULCFB1 model can be used with most of the standard PSS®E turbine gover-
nor models. It will modify the speed or load reference of the governor.
17-37
Turbine Load Controllers Program Application Guide: Volume II
17-38
Chapter 18
Modeling dc Transmission
This chapter describes various dc line models and their modeling in PSS®E. Although the
Chapter 18 - Modeling dc Transmission
description given indicates that dc line models are called from the connection routines
(CONEC and CONET), at PSS®E-30, several dc line models have been tableized (i.e., use
of dc line models does not require calls in CONEC and CONET).
18.1 General Considerations

Because dc transmission behavior is dominated by its controls, these controls must be
modeled, but, because the bandwidth of the controls is far greater than that of the PSS®E
simulation in general, it is usually not practical to represent the detailed dynamics of these
controls. Figure 18-1 details dc transmission controls. Each converter bridge is controlled
by a local feedback loop of bandwidth consistent with the firing delay accuracy require-
ments of the rectification/inversion process. These local loops work independently to
maintain bridge current or voltage at desired values. The desired values are provided by an
outer control loop that works in a supervisory role and coordinates the action of the sev-
eral converter bridges and the ac power system.
18-1
Modeling dc Transmission PSS®E 33.4
General Considerations Program Application Guide: Volume II
I,V
Measurement
BRIDGE CONTROL
Firing
Delay
Tap
Control
Current or Voltage
Supervisory Control Setpoint
(I, P, Start-Up, etc.)
Supervisory I,V
Measurement
Power, Current, or Voltage Setpoint
Figure 18-1. Arrangement of dc Transmission Control
The behavior of the bridges and their inner control loops is illustrated by Figure 18-2. A
rectifier bridge, for example, may be regarded as an adjustable voltage source forcing cur-
rent through transmission system resistance and inductance against the constant back-emv
of the inverter as shown in Figure 18-2a. A simple current control could be set up by apply-
ing current setpoint changes to rectifier voltage on a open-loop basis with a gain equal to
dc resistance as shown in Figure 18-2b. A step change of current setpoint would produce
the time response shown in Figure 18-2b, in which dc voltage changes in a step and the cur-
rent follows with a delay time constant determined by inductance and resistance. This time
constant would be small in relation to those of principal importance elsewhere in PSS®E,
but not negligible.
A real dc transmission rectifier is not operated by adjusting its output voltage in this open-
loop manner; rather, the local bridge control is a feedback loop that adjusts firing delay to
control the dc current to a setpoint, as shown by Figure 18-2c. The controller transfer func-
tion depends on details of the bridge design and is usually quite complex, nonlinear, and of
broad bandwidth in relation to the 0 to 30 rad/sec bandwidth over which PSS®E simula-
tion is applicable. A typical response of a bridge control loop to a step change of current
setpoint is shown in Figure 18-2c. The feedback controller is able to force the dc current
rapidly to its new value by transient overadjustment of the rectifier voltage. This rapid
forced response of the rectifier and its local control loop is completed in a time that is gen-
erally shorter than the shortest time interval that can be recognized within the bandwidth
of PSS®E.
Because the local converter bridge controls of dc transmission and the associated response
of dc current and voltage are so rapid in relation to the time scale of most PSS®E simula-
tions, several PSS®E models (CDC4, CDC6, CEELRIT, CMDWAST, CHIGATT, CMFORDT,
18-2
PSS®E 33.4 Modeling dc Transmission
Program Application Guide: Volume II General Considerations
etc.) treat dc converter pairs as if they move instantaneously to their new operating point
when any of their input signals or ac feed voltages are changed. These pseudo steady-
state, HVDC dynamic models calculate the active and reactive power loading of the
HVDC converters using steady-state converter relationships similar to those used for
power flow except that transformer taps remain fixed and the direct current and dc voltage
or margin angle may be varied to model the effects of higher level controls. These PSS®E
dc transmission models, then, are not concerned with the internal dynamic behavior of dc
converters and lines, just as the ac network model is not concerned with the internal tran-
sient behavior of transformers and three-phase transmission lines.
The pseudo steady-state, HVDC dynamic models are not able to directly represent the
mode of operation where the rectifier firing angle is not at a limit and the inverter margin
angle is also not at a limit or controlling voltage, because this condition does not occur in
the steady state. This temporary dynamic condition may occur during startup and/or a
cycle or two following a disturbance. If the recovery is expected to be slow, as may be the
case if the ac system is weak, then some pseudo steady-state models may depress direct
current and dc voltage during a disturbance and ramp up voltage and current at a user
specified rate following the disturbance to approximate the HVDC recovery characteris-
tics. Other models, such as CDC6, will only do this following a block or bypass.
HVDC models, such as CDCVUP, represent the temporary dynamic condition when nei-
ther converter is at a firing angle or a margin angle limit and both are fighting for control
of current. The dynamic interaction between the rectifier and inverter voltage-sensitive
current limits, which determines the dc voltage and current recovery following a distur-
bance, is explicitly represented in these models without having to represent the high-speed
HVDC controller and system L/R dynamics.
18-3
Idc
Rdc
Ldc
Vdcr Vdci
a. Simple dc Line Current Control Configuration, Adjustable Rectifier

Idc
Rdc Vdcr
I set V dcr
Iset
Time
b. Open-Loop Rectifier Current Control
Idc
+ G(s)
Iset Vdcr
Ydc(s) Vdcr
Idc
Iset
Time
c. Feedback Rectifier Current Control
Figure 18-2. Forced Control of dc Converter Operation

by Broad Bandwidth Feedback Loop
18-4
Program Application Guide: Volume II General Considerations
PSS®E also includes HVDC models such as CASEA1 and CDCRL that represent some
high-frequency controller effects and the L/R dynamics of the dc transmission. These
models use an internal integration time step that is shorter than that used by the network
and other PSS®E models.
All PSS®E dc transmission models are concerned with the transient behavior of the outer,
or supervisory, levels of the dc controls because the maneuvering of the dc power flow by
this control has a strong influence on the ac system.
The modeling of dc transmission recognizes three distinct types of action by the controls:
1. Normal regulation of dc converter operation to maintain specified constant current or
constant power transfer with coordination of rectifier and inverter current setpoints
2. Temporary overriding of dc converter normal operating setpoints in response to distur-

bances of ac system voltages during faults
3. Modulation of the dc power setpoint by a supplementary control device, the purpose of

which is, for example, to assist in the damping of rotor angle swings in the ac system
The modeling of the normal regulating control of the dc converters includes the basic con-
trol principle, illustrated by Figure 6-12, that underlies all dc transmission operation. Indi-
vidual dc transmission models represent different normal regime manipulations of the
current and voltage setpoints and delay angle limits defining Figure 6-12.
Normal control action is overridden by special control actions when ac or dc voltages at
the converters reach abnormal levels that may cause commutation difficulties, excessive
currents, or unacceptable harmonics. PSS®E dc transmission models execute these over-
riding control actions when the positive sequence ac voltages or dc voltages at the con-
verters reach specified levels. The execution of protective actions in model CDC4 does
not necessarily represent actions a real dc converter would take. Rather, it represents the
user’s decision to simulate the blocking, bypassing, unblocking, or unbypassing of the
converters when the specified voltage conditions are encountered. The modeling of pro-
tective action, as distinct from implementation of the user’s decision, is impossible in
PSS®E for the following reasons:
1. The protection of a dc converter is dependent on individual phase-to-ground and
phase-to-phase voltage wave forms and these are not available in PSS®E dynamic
simulations, which consider only the positive sequence fundamental frequency aspect
of system performance.
2. The protection of each bridge is determined by the internal details of the bridge firing
controls, which operate in a bandwidth far beyond the low-frequency band covered by
PSS®E and which, therefore, are not modeled in PSS®E.
The positive voltage levels at which the dc line would execute its various control actions
are quite likely to be different for different disturbances such as line-to-ground faults, line-
to-line faults, and so on.
All dc transmission dynamic models have access to all dc transmission data and variables
used in power flow (see Section 6.4.7.4). Additional parameters and variables, where
needed, are defined on the individual model data sheets.
18-5
18.1.1 Modeling Asymmetries in the ac System

Positive-sequence phasor equations are used to characterize the ac network in stability
programs. Because phasors are used, the high-frequency (>60 Hz) network behavior can-
not be represented and the response of the HVDC bridge and pole controls to such behav-
ior cannot be predicted. Also, the use of positive-sequence equations precludes
representing most asymmetries on the ac system. However, it is possible to represent the
effect on the positive-sequence flows and voltages of a few simple asymmetries, such as
faults or circuit outages, by using an equivalent shunt or branch element to represent the
impedance of the negative- and zero-sequence networks as seen from the fault.
The negative- and zero-sequence impedance as seen from the fault can be used to calcu-
late the negative- and zero-sequence voltage at the fault, but not elsewhere in the network.
Phase voltages may be obtained from the sequence quantities.
For faults at either converter, HVDC models, such as CASEA1, CDCRL, and CDCVUP, use
the sequence impedances to calculate the phase voltages at the converter terminals. Three-
phase voltages are used to flag possible commutation failures, to calculate the maximum
overlap angle – the margin angle needed to obtain the minimum margin area – and to cal-
culate the minimum firing angle required to have sufficient commutation voltage. To do
this, an ICON indicating the type of fault must be set when the fault is applied and model-
ing data must include values for the sequence and fault impedances. When a fault is not at
the converter terminals, only the positive sequence effects of faults may be represented.
The ICON indicating fault type should be set only for faults placed at the converter buses.
18.1.2 Blocking, Bypassing, and Commutation Failures

When an inverter is bypassed, the dc side is shorted and the ac side is open. The rectifier
will continue to circulate a low level of direct current through the shorted inverter at a low
voltage. The rectifier will therefore draw some vars but very little power from the ac sys-
tem and the inverter will draw no power of vars. To simulate a bypass in models such as
CASEA1, CDCRL, and CDCVUP, the appropriate ICON is set equal to 1. In models such
as CDC6 the bypassing voltage threshold can be raised to force a bypass by changing the
appropriate CON.
To simulate an inverter commutation failure, bypass the inverter and apply an ac system
shunt at the inverter with a reactance equaling the leakage reactance of the converter trans-
former. Blocking (turning off a converter) can be simulated by changing the appropriate
ICON or by raising the blocking voltage threshold to force a block. The voltage and cur-
rent recovery following a block or bypass can be represented by most of the HVDC mod-
els with varying degrees of user intervention.
18.1.3 Capacitor Commutated Converters

The dc line models discussed in the following paragraphs are appropriate for use with
noncapacitor commutated converters only.
18-6
Program Application Guide: Volume II Model CDC4
18.2 Model CDC4
18.2.1 Normal Operation

The firing angle limits of a dc transmission line generally have a wider range on a tran-
sient basis than in the steady state, as illustrated by Figure 18-3. CDC4 adjusts rectifier and
inverter firing delay angles within the dynamic limits to operate the dc transmission in
accordance with the combined characteristic shown in Figure 6-12. The scheduled dc volt-
age and scheduled dc power (or current) are as specified in the power flow working case
by the parameters VSCHED, SETVAL and MDC (see Section 6.4.7.4).
The instantaneous current setpoint, Iset, is adjusted continuously as shown in Figure 18-4, if
the line is in constant power mode (MDC = 1). The inverter current setpoint is assumed to
follow the rectifier current setpoint to always provide the current margin, DELTI, as spec-
ified in the power flow working case. Changing of the dc operating setpoints VSCHED,
SETVAL, and MDC, must be handled by changing the power flow data values via the net-
work changes section of activity ALTR or via activity LOFL and CHNG.
The dc current will fall below the instantaneous rectifier current setpoint, but not below
the corresponding inverter current setpoint, if the rectifier firing delay angle reaches its
dynamic lower limit, ALFDY. The transmitted power will then fall correspondingly, even
when the line is in power control mode (MDC = 1).
CDC4 maintains the desired constant power as long as the inverter-end dc voltage stays
above the value VCMODE, but switches to the nominal current setting Pset/Vsched, if the
dc voltage falls below this level. If the control switches out of constant power mode for
this reason, it is blocked from returning for a time delay, TCMODE, and may return to
power control if the inverter dc voltage rises above VCMODE.
Transformer taps are not adjusted automatically during dynamic simulation runs but may
be changed manually via power flow data change dialog (i.e., via network changes in
activity ALTR).
18-7
Model CDC4 Program Application Guide: Volume II
90°
Permissible
Power Flow
Range
ALFMAX Power Flow Data
Target Power
Flow Range
ALFMIN Power Flow Data
Permissible
Dynamic Range
ALFDY Dynamic Angle Limit
(Model CON Data)
Figure 18-3. Ranges of Alpha and Gamma Angles

in Power Flow and Dynamic dc Transmission Simulations
Rcomp
–
Current Setpoint
MDC = 2 I’set Bridge and V’set
+
•  Line Simulation 
(A) • Iset Low Value Logic
MDC = 1 Gate
+ +
I Vdci Vdcr Idc
V
Power Setpoint Modulating
(MW) Signal 1 1
106   106 ----------------------- ----------------------
(/SETVAL/) VAR(L) 1 + T vdc s 1 + T idc s
(MW)
SETVAL <0
Rdc 103
• +
SETVAL >0 + Voltage
 Setpoint (kV)
(VSCHED)
Figure 18-4. CDC4 dc Transmission Control Arrangements
18-8
18.2.2 Operation During Transients

All dc transmissions must be protected against commutation failure in the inverter. A com-
mutation failure results in one pair of diodes conducting permanently in the inverter
bridge. Because these are normally not on the same leg of the bridge, the dc current flows
permanently through one phase-pair of the transformer. Protection, and extinguishing of
the two conducting diodes, is handled by bypassing the inverter bridge as shown in
Figure 18-5. CDC4 models a dc transmission in which the bypass is a fast operating vac-
uum switch. CDC4 models two actions taken by the dc converters during ac system distur-
bances:
1. The rectifier and inverter are both shut down (blocked) if the ac voltage at the rectifier
falls below the per unit value, VBLOCK.
2. The inverter bypass switch is closed if the inverter end dc voltage falls below the kilovolt
value, VBYPAS. The rectifier continues to maintain dc current at scheduled value.
Low dc voltage does not cause blocking of the rectifier unless the rectifier ac voltage is
also low. Low ac voltage at the inverter does not cause blocking or bypassing unless the
inverter dc voltage is also low. If blocked, the rectifier remains blocked for a minimum of
TBLOCK seconds; it may then restart when the per-unit voltage at its ac bus rises to a
value of VUNBL. If bypassed, the inverter remains bypassed for a minimum of TBYPAS
seconds; it may then reestablish dc voltage when the voltage at its ac bus rises to the per-
unit value, VUNBY.
18-9
1 3 5
Commutation from
valve 4 to valve 6
fails
2 4 6
1 3 5 Bypass closes
to allow valves
1 and 4 to be
extinguished
2 4 6
+
Inverter operation
1 3 5
reestablished by
starting valves
3 and 6 in proper
2 4 6 sequence after
suitable delay
–
Figure 18-5. Use of Bypass Switch to Allow Shutdown of Inverter
After Commutation Failure
18-10
I set
Determined by
Normal Control
I set System
dc Voltage Vsched Vsched
(kV) (A)
Increased Current
Actual
VRAMP Actual CRAMP to Maintain
dc Voltage pu/sec dc Voltage pu/sec Constant Power
While Voltage
Programmed is Low
dc Voltage
Setpoint
RSVOLT RSCUR Programmed
dc Current
Setpoint
t t
TBLOCK TBLOCK
a. Reestablishment of Voltage and Current Following Blocking
dc Current at
Nominal Value
During Low
Iset Bypass Sequence
dc Voltage Vsched Vsched (A)
(kV)
Actual VRAMP Actual

dc Voltage pu/sec dc Voltage
Due to
Depressed
ac Voltage dc Current
Adjusted to
Programmed Hold Constant
RSVOLT dc Voltage Power
Setpoint
t t
TBYPAS TBYPAS
b. Reestablishment of Voltage with Current Maintained at Nominal Value Following Bypass
Figure 18-6. Examples of dc Voltage and Current Profiles

During ac System Disturbances
18-11
When restarting from a blocking, the dc voltage and current instantaneous setpoints follow
the programs shown in Figure 18-6a. When reestablishing voltage after a bypass, the dc
voltage setpoint follows the program shown in Figure 18-6b.The voltage and current rees-
tablishment programs are specified by the parameters RSVOLT, RSCUR, VRAMP, and
CRAMP. The current setpoints at both rectifier and inverter are overridden at all times by
a voltage-dependent current limit as shown in Figure 18-7. If the voltage-dependent current
limit comes into play during reestablishment of dc voltage or current, it will force the dc
current to a value below the instantaneous current setpoint. This effect can be largely elim-
inated by proper coordination of the voltage-dependent current limit profile, voltage rees-
tablishment rate, VRAMP, and current reestablishment rate, CRAMP.
The dc control logic of normal operation is in effect throughout reestablishment of current
and/or voltage. Accordingly, the actual current will fall below the rising instantaneous cur-
rent setpoint at any time that the rectifier encounters its dynamic minimum value of firing
delay angle, ALFDY. This situation may be fairly common: Rapid reloading of the dc
transmission presents the sending-end ac generation with a sudden, and often large,
demand for both real and reactive power, which tends to pull down the ac voltage at the
rectifier. This effect may be alleviated by reducing the dc transmission’s voltage and cur-
rent ramping rates to values that allow the voltage regulators of the ac generators to keep
up with the reapplied rectifier load.
Voltage Dependent
Current Limit
Maximum
Current (A) V3,C3
V2,C2
Permissible range in
normal operation
Possible range during

restart - C0 will
V1,C1 override
0,C0
RSCUR
Inverter
dc Voltage (kV)
Figure 18-7. Voltage-Dependent Current Limit for dc Converters
18-12
18.2.3 Modulating Control

CDC4 accepts a modulating signal via its auxiliary signal input VAR, VAR(L), as shown
in Figure 18-4. The modulating signal is always assumed to be expressed in terms of mega-
watts. Accordingly,
1. The modulating signal is added directly to the power setpoint when the line is in power
control mode (MDC = 1).
2. The modulating signal is converted to amps by dividing by measured dc voltage and

added to the current setpoint when the line is in current control mode (MDC = 2). (Mea-
sured dc voltage may be made equal to instantaneous dc voltage by setting the
transducer time constants Tvdc and Tidc in CON’s J+2 and J+3 to zero, if desired.)
The modulating signal is recognized when the dc transmission is in its normal operating
condition and ignored when the converters are blocked, bypassed, or recovering from a
block or bypass operation. The modulating signal may be applied by assigning a nonzero
value to VAR(L) by any logic that may be meaningful in a specific simulation. The nor-
mal procedure is to call a PSS®E library model such as SQBAUX to represent the produc-
tion of the modulating signal by a supplementary controller.
The model determining the modulating signal must be called before the call to CDC4 in
the CONEC subroutine. The user may combine the outputs of more than one supplemen-
tary controller. For example, the modulating signal could be set up as the sum of the out-
puts of two different controllers, both modeled by SQBAUX, by the logic:
CALL SQBAUX(I1,J1,K1,M1,NDC1)
CALL SQBAUX(I2,J2,K2,M2,NDC1)
VAR(L) = VAR(M1) + CON(999) * VAR(M2)
This logic allows the outputs of the two SQBAUX supplementary controllers to be added
together with an arbitrary weighing factor specified by CON(999). The user can make
runs comparing the effects of different weighings by changing CON(999). This is often
preferable to hard coding a weighing in a statement such as:
VAR(L) = VAR(M1) - 0.5 * VAR(M2)
The CDC4 model does not represent transmission delays that may occur when a modulat-
ing signal is derived at one end of a dc line but acted on by the converter bridge at the
opposite end. Any such transmission delays should be represented, when appropriate, in
the logic used to calculate VAR(L). Such delays are represented in SQBAUX.
18.2.4 Use of CDC4 when Bridge Firing Angle is Initialized on Power Flow
Limit
Model CDC4 does not recognize the power flow firing angle limits ALFMIN and GAM-
MIN. This is of no consequence when the power flow solution is normal (neither alpha nor
gamma is on its power flow limit) or when the dynamic limits ALFDY and/or GAMDY
are the same as the corresponding power flow limits ALFMIN and/or GAMMIN.
The distinction between static and dynamic minima does become a difficulty if the initial
condition power flow has alpha or gamma at its static minimum. In this case the network
solutions made in dynamics’ activities STRT and RUN, which recognizes ALFDY and
18-13
GAMDY, will be different from the converted power flow solution that is the intended ini-
tial condition. This difference is not proper because all network solutions prior to the
application of a disturbance should be identical to the converted power flow. The differ-
ence will cause the value of n in the STRT message, INITIAL CONDITION LOAD
FLOW USED n ITERATIONS, to be 2, 3 or more. It may also result in a nonsteady initial
condition if the incorrect solution forces excitation system or other variables onto limits.
The user must force the dynamic simulation to use ALFMIN and GAMMIN until a distur-
bance is initiated, and then ALFDY and GAMDY. Because transients are initiated by user
commands, the user must control the switch from ALFMIN to ALFDY as the disturbance
is applied. Two approaches are available:
1. Set ALFDY equal to ALFMIN and GAMDY equal to GAMMIN in all initial data setups
and initial condition snapshots, and leave it unchanged during execution of STRT and
RUN prior to the disturbance. When applying the disturbance manually, change these
values to the actual dynamic values by the appropriate changes of CONs. Also, change
the values of ALFMIN and GAMMIN to ALFDY and GAMDY. This change is quite easy
to do, when making simulation runs, with the aid of PSAS.
2. Automate the adjustment of dynamic firing angle limits by making the initial data setup
with the dynamic values of the limits in ALFDY and GAMDY, and replacing them with
the static limits by CONEC logic during the predisturbance work. The following state-
ments in CONEC may be used ahead of the CDC4 calls. Spare refers to any unused
CON; TST is the time (usually 0) when the disturbance is to be initiated.
IF(MODE.NE.4) GO TO 1
CON(spare) = CON(J)
CON(spare+1) = CON(J+1)
CON(J) = ALFMIN(line number)
CON(J+1) = GAMMIN(line number)
1 IF(ABS(TIME-TST).GE.0.1*DELT
.OR.KPAUSE.NE.2)GO TO 2
CON(J) = CON(spare)
CON(J+1) = CON(spare+1)
11 ALFMIN(line number) = CON(J)
GAMMIN(line number) = CON(J+1)
2 CONTINUE
Both approaches suggest changing ALFMIN and GAMMIN to the dynamic limit values
when a disturbance is applied. This change is necessary only if the SQBAUX model is
applied to the line in question. In this case, the change should be made whether or not the
initial condition power flow has alpha or gamma on its limit.
18.2.5 Calling Sequence

Model CDC4 must be invoked by coordinated calls in subroutines CONEC and CONET
as follows:
CONEC: CALL CDC4 (I,J,K,L,M)
CONET: CALL TDC4 (I,J,K,L,M)
18-14
The five calling arguments must have iden- assignments made by DYRE with the indi-
tical values in the two call statements vidual data items for the two lines. Assign-
because the simulation logic in subroutine ment of dc line variables to output
CONEC communicates with the logic in channels requires reference to Figure 18-5
subroutine CONET through the CON, through 18-13 to find the VARs and
VAR, ICON, and STATE arrays. STATEs containing the desired quantities.
Activity DYRE inserts the required call Plotting alpha and gamma angles for dc
statement into both CONEC and CONET line 1, for example, requires that
when a data record is read with the model activity CHAN be used to place VAR(23)
name CDC4. The model name TDC4 is not and VAR(24) in output channels.
recognized by activity DYRE. Figure 18-14 shows the behavior of the dc
transmissions following a fault at bus
18.2.6 Example Cases 1200, the ac side of the inverter station.
Figures 18-8 through 18-14 illustrate the
The fault is represented by a positive-
application of model CDC4 in the sample sequence admittance of -j35 per unit,
problem setup shown by Figures 6-15, 6-16, which approximates the effect of a high-
and 6-17. Figure 18-8 shows the generator impedance, single-phase fault on one of
the four 230-kV circuits between buses
data pertaining to the two synchronous
condensers at buses 1401 and 1402, and 1200 and 1201. Figure 18-14 also shows
that the terminal voltage of the synchro-
the power flow data for the dc transmission
nous condensers falls to 0.72 per unit as a
lines; these tabulations reflect the initial
result of the fault. The dc line current ini-
condition power flow solution. Figure 18-8
tially rises in the attempt to hold constant
also shows that the initial alpha and
power with reduced dc voltage. The dc
gamma values are above the static minima
current is then reduced with a time con-
(adjustments discussed in Section 18.2.4 are
stant of about 0.05 sec as the measured dc
not needed) and that the predisturbance
voltage at the inverter falls and brings the
current is comfortably within the operating
voltage-dependent current limit into play.
region allowed by the voltage-dependent
Removal of the fault allows the dc current
current limit. Figure 18-9 shows the
dynamic simulation data used to represent to rise with a similar time constant. The
the two lines. Of particular interest are the line returns to nominal operating condi-
dynamic alpha and gamma minimum val- tions as the inverter terminal ac voltage is
ues that fall below the power flow minima, returned to nominal value by the voltage
and the voltage dependent current limit regulators of the synchronous condensers.
function specified by CONs J+15 through
J+20. Figure 18-10 shows the permissible
operating region drawn to scale, while
Figure 18-11 shows the output produced
when activity DYRE reads the dc line data
records shown in Figure 18-9. DYRE also
produces the CONEC and CONET state-
ments shown in Figure 18-11. Figures 18-12
and 18-13 show CDC4 data sheets filled in
to correlate the CON, VAR, and STATE
18-15
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E WED JAN 03, 1996 14:38
SAMPLE SYSTEM FOR PSS®E MANUAL GENERATOR
1100KV DC CASE UNIT DATA
BUS# NAME BSKV COD ID ST PGEN QGEN QMAX QMIN PMAX PMIN MBASE Z S O R C E X T R A N GENTAP
100 NUCLEAR 345 2 1 1 1000 6 400 -100 1050 330 1100 0.0000 0.2000 0.0000 0.1500 1.0250
100 NUCLEAR 345 2 2 1 1000 6 400 -100 1050 330 1100 0.0000 0.2000 0.0000 0.1500 1.0250
200 HYDRO 345 3 1 1 1237 -125 1000 -300 1750 0 1750 0.0000 0.2500 0.0000 0.1000 1.0250
1100 CATNIP 230 -2 1 1 500 250 250 0 500 150 500 0.0000 0.2500 0.0000 0.1000 1.0250
1401 WCOND 18.0 2 1 1 0 654 800 -100 9999-9999 800 0.0000 0.1800 Synchronous
1402 ECOND
1600 MINE
18.0 2
765 3
1 1 0 654 800 -100 9999-9999 800 0.0000 0.1800
1 1 1000 278 667 -333 1010 320 1067 0.0000 0.2000 0.0000 0.1200 1.0250
condensers at
1600 MINE 765 3 2 1 1000 278 667 -333 1010 320 1067 0.0000 0.2000 0.0000 0.1200 1.0250 inverter station
1600 MINE 765 3 3 1 1000 278 667 -333 1010 320 1067 0.0000 0.2000 0.0000 0.1200 1.0250
Steady-state ,  are Static ,  limits

clear of static limits
SAMPLE SYSTEM FOR PSS®E MANUAL DC LINE DATA
1100KV DC CASE
DC# MDC RDC RCOMP DELTI SETVAL VSCHED DCVMIN VCMODE DCAMPS VCOMP METER
1 1 8.200 0.000 0.1500 1500.0 554.4 0.0 443.5 2605.2 554.4 INV
X---CONVERTER BUS--X ALF/GAM MIN MAX PAC QAC VDC Predisturbance

R: 1600 [MINE 765] 6.76 5.00 40.00 1500.0 417.7 575.8
I: 1403 [WDUM 18.0] 18.17 18.00 30.00 -1444.3 590.9 554.4 dc line operating
condition
NB EBASE RC-OHMS XC-OHMS TR TAP TAPMIN TAPMAX TAPSTP
R: 2 500.0 0.000 3.880 0.44000 1.03125 0.90000 1.10000 0.00625
I: 2 230.0 0.000 3.047 0.96500 1.00000 0.90000 1.10000 0.00625
X---MEASURING BUS--X X--TAPPED SIDE BUS-X X--UNTAPPED SIDE --X CKT RATIO
R: 0 [ ]
I: 0 [ ]
DC# MDC RDC RCOMP DELTI SETVAL VSCHED DCVMIN VCMODE DCAMPS VCOMP METER
2 1 8.200 0.000 0.1500 1500.0 554.4 0.0 443.5 2605.2 554.4 INV
X---CONVERTER BUS--X ALF/GAM MIN MAX PAC QAC VDC

R: 1600 [MINE 765] 6.76 5.00 40.00 1500.0 417.7 575.8
I: 1404 [EDUM 18.0] 18.17 18.00 30.00 -1444.3 590.9 554.4
NB EBASE RC-OHMS XC-OHMS TR TAP TAPMIN TAPMAX TAPSTP

R: 2 500.0 0.000 3.880 0.44000 1.03125 0.90000 1.10000 0.00625
I: 2 230.0 0.000 3.047 0.96500 1.00000 0.90000 1.10000 0.00625
X---MEASURING BUS--X X--TAPPED SIDE BUS-X X--UNTAPPED SIDE --X CKT RATIO
R: 0 [ ]
I: 0 [ ]
Figure 18-8. Power Flow Data for dc Lines and Synchronous Condensers
in System of Figure 6-15 Predisturbance Condition
18-16
CDC4
dc Line Model
CALL CDC4 (I,J,K,L,M) from CONEC

CALL TCD4 (JI,J,K,L,M) from CONET
This is dc line #______ I,
This model uses CONs starting with #______ J,
and ICONs starting with #______ M.

J 5 ALFDY, minimum alpha for
dynamics (degrees)
J+1 15 GAMDY*, minimum gamma
for dynamics (degrees)
J+2 0.05 TVDC, dc voltage transducer
time constant (sec)
J+3 0.05 TIDC, dc current transducer
time constant (sec)
J+4 0.6 VBLOCK, rectifier ac blocking
voltage (pu)
J+5 0.65 VUNBL, rectifier ac
unblocking voltage (pu)
J+6 0.1 TBLOCK, minimum blocking
time (sec)
J+7 0.6 VBYPAS, inverter dc
bypassing voltage (kV)
J+8 0.65 VUNBY, inverter ac unby-
passing voltage (pu)
J+9 0.1 TBYPAS, minimum
bypassing time (sec)
J+10 200 RSVOLT, minimum dc
voltage following block (kV)
J+11 500 RSCUR, minimum dc current
following block (amps)
J+12 5. VRAMP, voltage recovery
rate (pu/sec)
J+13 5. CRAMP, current recovery
rate (pu/sec)
J+14 400 C, minimum current
demand (amps)
J+15 300 V1, voltage limit point 1 (kV)
18-17

J+16 1000 C1, Current limit point 1
(amps);
>C
J+18 3000 C2, current limit point 2
(amps)
(amps)
J+21 0.1 TCMODE, minimum time
stays in switched mode (sec)
*Ignored if in gamma control

(i.e., GAMMAX = GAMMIN in power
flow).
K Measured inverter dc voltage (V)
K+1 Measured inverter dc current
(amps)
VARs # Description
L Other signals, MW
L+1 RESTR, time unblocks or unby-
passes (sec)
L+2 VRF, voltage ramping factor
L+3 CRF, current ramping factor
L+4 VCOMP, compensating dc voltage
(V)
L+5 PACR, rectifier ac real power (pu)
L+6 QACR, rectifier ac reactive power
(pu)
L+7 PACI, inverter ac real power (pu)
L+8 QACI, inverter ac reactive power
(pu)
L+9 VDCI, inverter dc voltage (V)
L+10 VDCR, rectifier dc voltage (V)
L+11 DC, dc current (amps)
L+12 ALFA, alpha (degrees)
L+13 GAMA, gamma (degrees)
L+14 TIME, reswitches mode
18-18
ICONs # Description
M Bypass control flag*:
0 = Not bypassed
1 = Bypassed
2 = Unbypassing
M+1 Blocking control flag*:
0 = Not blocked
1 = Blocked
2 = Unblocked
M+2 Switched mode control flag*:
0 = Normal
1 = Mode switched
*Not intended to be changed by the

user.
Note: Statement must be added to CONEC to sum any supplementary signals into VAR(L).
I, ’CDC4’, ALFDY, GAMDY, TVDC, TIDC, VBLOCK, VUNBL, TBLOCK, VBYPAS, VUNBY,
TBYPAS, RSVOLT, RSCUR, VRAMP, CRAMP, C, V1, C1, V2, C2, V3, C3, TCMODE/
Maximum dc Current (amps)

V3, I3
V2, I2
dc Current
V1, I1
C
Minimum dc Current (amps)
Voltage-Dependent Upper Current Limit
n 'CDC4' 5 15 .05 .05 .6 .65 .1 .6 .65 .1 200 500 Data record to be

5. 5. 400 300 1000 500 3000 500 3000 .1/
included in input file
for activity DYRE
Figure 18-9. Permissible dc Line Operating Domain
18-19
V3,C3
3000
V2,C2
Predisturbance
Operating Point
2000
dc Current (A)
1000 V1,C1
RSCUR
C0
0
0 100 200 300 400 500 600
dc Voltage (kV)
Figure 18-10. dc Line Dynamics Data to Complement Power Flow Data of Figure 18-8
18-20
ACTIVITY? dyre
ENTER DYNAMICS DATA SOURCE FILENAME: scdc1.dyr File of dynamics raw data
ENTER FILENAME FOR SUBROUTINE CONEC: cce Place CONEC here
ENTER FILENAME FOR SUBROUTINE CONET: cte Place CONET here
ENTER FILENAME FOR PSSPLT RELAY CHARACTERISTIC DATA:

CON STATE VAR ICON
1 1 1 1
ENTER STARTING CON, STATE, VAR, ICON OR CARRIAGE RETURN Fresh data setup
GENERATOR MODELS USE:

CONS 1- 120
STATES 1- 51
EXCITER MODELS USE:

CONS 121- 200
STATES 52- 75
VARS 1- 2
GOVERNOR MODELS USE:

CONS 201- 254
STATES 76- 91
VARS 3- 10
OTHER MODELS USE:

CONS 255- 298
STATES 92- 95 Assigned to two CDC4 calls
VARS 11- 40
ICONS 1- 6
GENS: GENROU GENSAL

6 3
EXSYS: SCRX SEXS IEEEX1

3 4 2
GOVS: TGOV1 HYGOV

6 1
MISC: CDC4
2

CON STATE VAR ICON
299 96 41 7
COMPILE AND CLOAD4 BEFORE RUNNING SIMULATIONS
ENTER FILENAME FOR COMPILING FILE (0 TO EXIT): cpe
Figure 18-11. Use of DYRE to Set Up CONEC and CONET Calling CDC4
(Sheet 1 of 2)
18-21
SUBROUTINE CONEC File cce.flx made by DYRE

C
$INSERT COMON4.INS
C
CALL CDC4 ( 1, 255, 92, 11, 1)
CALL CDC4 ( 2, 277, 94, 26, 4) CDC4 calls generated by DYRE
C
RETURN
END
INTEGER MC,SLOT,IT
SELECT (IT)
FIN
RETURN
END
SUBROUTINE CONET File cte.flx made by DYRE

C
$INSERT COMON4.INS
C
CALL TDC4 ( 1, 255, 92, 11, 1) TDC4 calls generated by DYRE
CALL TDC4 ( 2, 277, 94, 26, 4)
C
C
C
C
9000 CONTINUE
C
RETURN
END
Figure 0-0 (Cont). Use of DYRE to Set Up CONEC

and CONET Calling CDC4 (Sheet 2 of 2)
18-22
CDC4
dc Line Model

This is dc line # 1 I,
This model uses CONs starting with # 255 J,
and STATEs starting with # 92 K,
and VARs starting with # 11 L,
and ICONs starting with # 1 M.

J+15 270 V1, voltage limit point 1
J 255 ALFDY, minimum alpha for (kV)
dynamics (degrees)
J+1 256 GAMDY*, minimum gamma (amps);
for dynamics (degrees) >C
J+2 257 TVDC, dc voltage trans- J+17 272 V2, voltage limit point 2
ducer time constant (sec) (kV)
J+3 258 TIDC, dc current transducer J+18 273 C2, current limit point 2
time constant (sec) (amps)
J+4 259 VBLOCK, rectifier ac J+19 274 V3, voltage limit point 3
blocking voltage (pu) (kV)
J+5 260 VUNBL, rectifier ac J+20 275 C3, current limit point 3
unblocking voltage (pu) (amps)
J+6 261 TBLOCK, minimum J+21 276 TCMODE, minimum time
blocking time (sec) stays in switched mode
J+7 262 VBYPAS, inverter dc (sec)
bypassing voltage (kV) *Ignored if in gamma control
J+8 263 VUNBY, inverter ac unby- (i.e., GAMMAX = GAMMIN in power
passing voltage (pu) flow).
J+9 264 TBYPAS, minimum
K 92 Measured inverter dc voltage (V)
K+1 93 Measured inverter dc current
J+11 266 RSCUR, minimum dc cur-
(amps)
rent following block (amps)
J+12 267 VRAMP, voltage recovery
rate (pu/sec) VARs # Description
J+13 268 CRAMP, current recovery L 11 Other signals, MW
rate (pu/sec)
L+1 12 RESTR, time unblocks or unby-
J+14 269 C, minimum current passes (sec)
demand (amps)
L+2 13 VRF, voltage ramping factor
18-23
VARs # Description
L+3 14 CRF, current ramping factor
L+4 15 VCOMP, compensating dc voltage
(V)
L+5 16 PACR, rectifier ac real power (pu)
L+6 17 QACR, rectifier ac reactive power
(pu)
L+7 18 PACI, inverter ac real power (pu)
L+8 19 QACI, inverter ac reactive power
(pu)
L+9 20 VDCI, inverter dc voltage (V)
L+10 21 VDCR, rectifier dc voltage (V)
L+11 22 DC, dc current (amps)
L+12 23 ALFA, alpha (degrees)
L+13 24 GAMA, gamma (degrees)
L+14 25 TIME, reswitches mode
ICONs # Description
M 1 Bypass control flag*:
0 = Not bypassed
1 = Bypassed
2 = Unbypassing
M+1 2 Blocking control flag*:
0 = Not blocked
1 = Blocked
2 = Unblocked
M+2 3 Switched mode control flag*:
0 = Normal
1 = Mode switched

user.
18-24

V3, I3
V2, I2
dc Current
V1, I1
C
1 'CDC4' 5 15 .05 .05 .6 .65 .1 .6 .65 .1 200 500 Data record to be

5. 5. 400 300 1000 500 3000 500 3000 .1/
included in input file
for activity DYRE
18-25
CDC4
dc Line Model

This is dc line # 2 I,
This model uses CONs starting with # 277 J,
and STATEs starting with # 94 K,
and VARs starting with # 26 L,
and ICONs starting with # 4 M.

J 277 ALFDY, minimum alpha for
dynamics (degrees)
J+1 278 GAMDY*, minimum gamma
for dynamics (degrees)
J+2 279 TVDC, dc voltage trans-
ducer time constant (sec)
J+3 280 TIDC, dc current transducer
time constant (sec)
J+4 281 VBLOCK, rectifier ac
blocking voltage (pu)
J+5 282 VUNBL, rectifier ac
unblocking voltage (pu)
J+6 283 TBLOCK, minimum blocking
time (sec)
J+7 284 VBYPAS, inverter dc
bypassing voltage (kV)
J+8 285 VUNBY, inverter ac unby-
passing voltage (pu)
J+9 286 TBYPAS, minimum
J+11 288 RSCUR, minimum dc cur-
rent following block (amps)
J+12 289 VRAMP, voltage recovery
rate (pu/sec)
J+13 290 CRAMP, current recovery
rate (pu/sec)
J+14 291 C, minimum current
demand (amps)
18-26

(amps);
>C
(amps)
(amps)
J+21 298 TCMODE, minimum time
stays in switched mode
(sec)
*Ignored if in gamma control

(i.e., GAMMAX = GAMMIN in power
flow).
K 94 Measured inverter dc voltage (V)
K+1 95 Measured inverter dc current
(amps)
VARs # Description
L 26 Other signals, MW
L+1 27 RESTR, time unblocks or unby-
passes (sec)
L+2 28 VRF, voltage ramping factor
L+3 29 CRF, current ramping factor
L+4 30 VCOMP, compensating dc voltage
(V)
L+5 31 PACR, rectifier ac real power (pu)
L+6 32 QACR, rectifier ac reactive power
(pu)
L+7 33 PACI, inverter ac real power (pu)
L+8 34 QACI, inverter ac reactive power
(pu)
L+9 35 VDCI, inverter dc voltage (V)
L+10 36 VDCR, rectifier dc voltage (V)
L+11 37 DC, dc current (amps)
L+12 38 ALFA, alpha (degrees)
L+13 39 GAMA, gamma (degrees)
18-27
VARs # Description
L+14 40 TIME, reswitches mode
ICONs # Description
M 4 Bypass control flag*:
0 = Not bypassed
1 = Bypassed
2 = Unbypassing
M+1 5 Blocking control flag*:
0 = Not blocked
1 = Blocked
2 = Unblocked
M+2 6 Switched mode control flag*:
0 = Normal
1 = Mode switched

user.

V3, I3
V2, I2
dc Current
V1, I1
C
2 'CDC4' 5 15 .05 .05 .6 .65 .1 .6 .65 .1 200 500 Data record to be

5. 5. 400 300 1000 500 3000 500 3000 .1/
included in input file for
ti it DYRE
18-28
SAMPLE SYSTEM FOR PSS/E MANUAL
MON JAN 08, 1996 14:57

1100KV DC CASE
POWER
TECHNOLOGIES,
INC.R FILE: outscdc1
DC LINE 1
CHNL# 5: [DC 1]
3000.0 2000.0
CHNL# 4: [GAMA 1]
50.000 0.0
CHNL# 3: [ALFA 1]
50.000 0.0
CHNL# 2: [QACR 1]
10.000 0.0
CHNL# 1: [PACR 1]
20.000 10.000
1.0000
0.90000
0.80000
0.70000
0.60000
TIME(SECONDS)
0.50000
0.40000
0.30000
0.20000
0.10000
0.0
Figure 18-14. Response to High Impedance L-G Fault at Bus 1200

(Sheet 1 of 3)
18-29
MON JAN 08, 1996 15:01

1100KV DC CASE
WCOND
POWER
TECHNOLOGIES,
CHNL# 13: [EFD 1401 [WCOND 18.000] MC 1]

10.000 -10.00
CHNL# 10: [ETRM 1401 [WCOND 18.000] MC 1]
1.1000 0.60000
CHNL# 8: [VARS 1401 [WCOND 18.000] MC 1]
10.000 0.0
CHNL# 6: [POWR 1401 [WCOND 18.000] MC 1]
1.0000 -1.000
1.0000
0.90000
0.80000
0.70000
0.60000
TIME(SECONDS)
0.50000
0.40000
0.30000
0.20000
0.10000
0.0
Figure 0-0 (Cont). Response to High Impedance L-G Fault at Bus 1200 (Sheet 2 of 3)
18-30
MON JAN 08, 1996 15:04

1100KV DC CASE
POWER
TECHNOLOGIES,
MINE PLANT
CHNL# 14: [EFD 1600 [MINE 765.00] MC 1]
10.000 0.0
CHNL# 11: [ETRM 1600 [MINE 765.00] MC 1]
1.1000 0.90000
CHNL# 12: [ETRM 1401 [WCOND 18.000] MC 1]
1.2000 0.0
CHNL# 9: [VARS 1600 [MINE 765.00] MC 1]
10.000 0.0
CHNL# 7: [POWR 1600 [MINE 765.00] MC 1]
15.000 5.0000
1.0000
0.90000
0.80000
0.70000
0.60000
TIME(SECONDS)
0.50000
0.40000
0.30000
0.20000
0.10000
0.0
Figure 0-0 (Cont). Response to High Impedance L-G Fault at Bus 1200 (Sheet 3 of 3)
18.3 Model CDC6

The model CDC6 is recommended for studying new proposed dc lines. It incorporates
most of the control logic of CDC4 (The first 21 CONs of this model are identical.), plus
18-31
Model CDC6A Program Application Guide: Volume II
some additional protection from the proposed modern lines. The dc control arrangements
for this line are shown in Figure 18-15. The blocking, bypassing, and mode switching logic
is summarized in Figure 18-16. This model differs from CDC4 in power control and when
converting the megawatt modulating signal. It will divide by a measured rectifier voltage
rather than measured current times dc resistance if power output is set at the rectifier end
(SETVAL>0).
This model allows for delayed automatic blocking of the line based on low rectifier ac
voltage. A timer is started when the rectifier ac voltage goes below VDEBLK. If the volt-
age is still below this value, TREBLK seconds later, the line blocks. Following this block,
the line cannot restart until TREBLK seconds after the ac rectifier voltage goes above
VUNBY.
This model also allows a low inverter ac voltage to send a signal via a communication
channel to the rectifier to shut the line down. This action is modeled by the instantaneous
inverter ac blocking voltage, VINBLK, and the delay, TCOMB. The line will stay blocked
until TINBLK seconds after the ac inverter voltage goes above VUNBY. TINBLK should
include the communication delay in getting the signal from the inverter to the rectifier.
Also, if the ac voltage goes below VACBYP on the inverter, a timer is started. If after
TDEBYP seconds the ac inverter voltage is still below VACBYP, the inverter will be
placed into full electronic bypass. The line will unbypass TUNBY seconds after the ac
inverter voltage goes above VUNBY. The blocking, bypassing and mode switching logic
is summarized in Figure 18-16.
18.4 Model CDC6A

This model is identical to model CDC6 except for the items listed below.
1. VAR(L+18), Imeasured current, VAR (L+19), Idesired before VDCL, and VAR(L+20),
VDCL output are for output purposes only. They are used by the Celilo-Sylmar margin
switching unit, MSU1.
2. VAR(L+21), GAMMOD allows the user to impose a gamma modulation signal.

This value will be used only with a constant gamma model in the power flow
(i.e., GAMMAX = GAMMIN).
3. VAR(L+22) and VAR(L+23), current modulation inputs downstream from the VDCL.
This is commonly referred to as low level modulation.
If the user does not set VAR(L+21) through VAR(L+23), the model will behave identi-
cally to CDC6.
18-32
Program Application Guide: Volume II Model CDC6A
Rcomp
–
Current Setpoint
MDC = 2 I’set Bridge and V’set
+
•  Line Simulation 
(A) • Iset Low Value Logic
MDC = 1 Gate
+ +
I Vdci Vdcr Idc
V
Power Setpoint Modulating
(MW) Signal 1 1
106   106 ----------------------- ----------------------
(/SETVAL/) VAR(L) 1 + T vdc s 1 + T idc s
(MW)
SETVAL <0
Rdc 103
•
Voltage
SETVAL >0 Setpoint (kV)
(VSCHED)
Figure 18-15. CDC6 dc Transmission Control Arrangements
Instantaneous Unblock if:

If VACR < VBLOCK
Block VACR > VUNBL for TREBLK
CON(J+4)
CON(J+5) CON(J+6)
If VACR Stays < VDEBLOCK for TDEBLK and TBLOCK Later
Block
CON(J+22) CON(J+23) CON(J+6)
Block Unblock if:

If VACI < VINBLK VACI > VUNBY for TINBLK
TCOMB Later
CON(J+25) CON(J+8) CON(J+29)
CON(J+26)
and TBLOCK Later
CON(J+6)
Bypass and Unbypassing Conditions
Bypass Unblock if:

If VACI Stays < VACBYP for TDEBYP
CON(J+27) CON(J+28) VACI > VUNBY for TINBYP
CON(J+30)
Instantaneous and TBYPAS later
If VDCI < VBYPAS*1000. CON(J+9)
Bypass
Mode Switch and Reswitch
Only in power control and not bypassed or blocked,

Mode
If VDCI < VCMODE*1000. Reswitch TCMODE if above
Switch CON(J+21)
Figure 18-16. CDC6 Blocking and Unblocking Conditions
18-33
18.5 CDC7T Model for the HVDC Cable and/or Overhead Line
CDC7T Model for the HVDC Cable and/or Overhead Line Program Application Guide: Volume II
Modeling dc Transmission
18.5.1 Introduction
The PSS®E CDC7T model simulates a pole of the 2-terminal HVDC transmission. There is a significant difference between this
model and generic HVDC models available in the PSS®E dynamic library, such as CDC4T or CDC6T.
The latter models assume an instantaneous response of the dc system to disturbances coming from the adjacent grids. They use the
same algebra as the power flow model to update converter parameters (alpha, gamma) at every each integration step. They assume
that converters are capable of maintaining the steady-state control algorithm using the same set points as in power flow. There is a
special algorithm of simulating blocking and by-passing HVDC that uses some threshold voltages on ac and dc sides. Setting up
these thresholds requires experience because they are very sensitive to how strong the interconnection to the system is.
The CDC7T model simulates the dynamics of the dc line and converter controls as it can be seen from the description below. At the
same time, this is a generic HVDC model in terms that a configuration of the dc circuit and algorithm of converter controls are typ-
ical for a conventional HVDC.
18-34
18.5.2 DC Circuit arrangement simulated by the CDC7T model

A dc circuit arrangement that can be simulated by the CDC7T model is shown in Figure 18-17. A dc line may comprise overhead
lines from both rectifier and inverter sides and a cable. An overhead dc line is represented by its dc resistance and some equivalent
inductance. A cable is represented by its dc resistance, equivalent inductance and capacitance. A small resistance representing the
cable damping is placed in series with the cable capacitance.
The model does not include frequency dependent sub-models for the overhead line or cable, therefore for the equivalent inductance
and capacitance respective values for the fundamental frequency are recommended.
PSS®E 33.4
Program Application Guide: Volume II CDC7T Model for the HVDC Cable and/or Overhead Line
LOHR LDCC LDCC LOHI

ROHR RDCC RDCC LOHI
LF1
LRR
LRI LF3
VDCC LF2
RF1
IDCR CDCC IDCI
RRR
RRI RF3
VDCcomp
RF2
RCDCC
Rectifier Inverter
VDCR VDCI
Figure 18-17. A DC Circuit Arrangement Simulated by the CDC7T Model
Although the dc line can be represented by a T-circuit with the lumped Rdc and Ldc
parameters, for the sake of flexibility the model uses resistances and inductances of over-
head lines on rectifier (ROHR, LOHR) and inverter (ROHI, LOHI) sides, resistance,
inductance, and capacitance of the dc cable (RDCC, LDCC, CDCC), and resistance and
inductance of smoothing reactors on both sides (RRR, LRR and RRI, LRI).
By changing ICON(M+3) from 1 to 2 the model switches from the mixed overhead-cable
dc line to just overhead line. It is not recommended to make this transition by reducing the
cable capacitance because of potential numeric problems. Also, if the user wishes to
model a pure overhead-line, it is recommended that RDCC & LDCC be set to zero.
The model allows to simulate faults in the dc system. Three fictitious LR-shunts have been
added to the model as it is shown in Figure 18-17. During normal operation, parameters of
these shunts have been set to very big numbers to have almost zero currents in these
shunts. To simulate faults in any of three points of the dc circuit, namely on the terminals
of the overhead line from the rectifier side, on the terminals of the overhead line on the
inverter side, and in the middle point of the cable, both inductance and resistor of a respec-
tive shunt should be reduced. To avoid numerical instability, the ratio of fault inductance
to fault resistance (i.e., the L/R ratio) should be kept constant at 0.1.
18-35
18.5.3 Controls Algorithm

Analysis of operation of numerous existing available HVDC projects reveals the follow-
ing 3 major configurations of converter controls.
Control configuration 1:
• Rectifier in dc current control
• Inverter in gamma control
• Rectifier in dc current control
• Inverter in dc voltage control
• Rectifier in dc voltage control
• Inverter in dc current control
The control configuration 1 was used for historically first HVDC projects. The second
configuration is typical for modern HVDC transmissions with overhead dc lines or com-
paratively short cables. For HVDCs with a long cable the control configuration 3 is very
likely to be implemented.
Choice of the control configuration depends on many factors, beginning with the value of
a short circuit ratio in points of interconnection of HVDC terminals to the system. That is
why the CDC7T model has a provision for choosing the control configuration. Accord-
ingly to the selected control configuration, some parameters of controls must be adjusted
as it is shown below.
For HVDC lines with a long HVDC cable a special control is needed to monitor the so
called compounded dc voltage VDComp of the cable terminal on the rectifier side and
keep it not greater than nominal voltage. In the description of converter controls below
characteristics of controls are “tied up” to this point.
Converter controls
Controls of both converters can be simulated based on the same structure, which includes
3 controllers, namely
• DC current controller
• DC voltage controller
• Gamma controller.
The basic converter control configuration is shown in Figure 18-18. The outputs of dc cur-
rent, dc voltage, and gamma controllers are used as inputs for the mode selector, which is
a maximum signal selector for the rectifier and minimum signal selector for the inverter.
The selected signal is processed by the PI-controller whose output is the alpha angle order
( order).
To avoid steady-state instability in situations when, due to ac voltage fluctuations on both
sides, rectifier sits on its minimal alpha limit and the inverter is in gamma control, addi-
tional signals proportional to the dc current error are added to the outputs of the dc voltage
and gamma controllers. The Current Error Control (CEC), shown in Figure 18-18, is sup-
18-36
posed to favorably change the slope of the converter characteristic around the steady-state
operating point. The CEC contribution is determined by non-linear gains that are repre-
sented by hard-coded look-up tables in the model.
CDC7T model can accept one auxiliary signal input (auxiliary signal index 1). The auxil-
iary signal has to be in units of MW. The current order (Iorder) is calculated by dividing the
sum of rectifier side power order and any auxiliary signal input by the rectifier side fil-
tered dc voltage.
Values for margins of controlled variables should be set depending on the selected control
configuration as shown in Tables 1 and 2.
Table 1: Rectifier: margins of controlled variables for different control
configurations
Control configuration 1 Control configuration 2 Control configuration 3

DC current margin 0.0 0.0 -0.1
DC voltage margin N/A N/A 0.0
Gamma margin N/A N/A N/A
Table 2: Inverter: margins of controlled variables for different control

configurations
Control configuration 1 Control configuration 2 Control configuration 3

DC current margin 0.1 0.1 0.0
DC voltage margin N/A 0.0 N/A
Gamma margin 0.0 N/A 10 degrees
Gamma order 12 – 18 degrees N/A 25 – 30 degrees
18-37
Figure 18-18. Basic Converter Control Configuration
18-38
The Voltage Dependent Current Order Limit (VDCOL) algorithm implemented in the
CDC7T model is shown in Figure 18-19. This is based on the generic Vd-Id characteristics
of Figure 18-20. The compounded dc voltage (Vdcomp) is processed through a lag block
and multiplied by a non-linear gain. The time constant of the former is set to up or down
value depending on whether the dc voltage is increasing or decreasing. The non-linear
gain (which represents the Vd-Id characteristics) is simulated using lookup tables, one for
rectifier and another for inverter, whose 5 pairs of coordinates are provided as last 20
CONs of the model. The output of the VDCOL (which is a current limit) is compared with
the current order, and the minimum of the two is used as an input to the dc current control-
ler.
Figure 18-19. VDCOL Controller for CDC7T
Figure 18-20. Typical VDCOL Control
18-39
18-40
Figure 18-21 depicts an example of how the current orders of the rectifier and inverter may
change with the Vdcomp.
VDCL DC Current Order

VDCOL Action as Coded in CDC7T
2
1.8
1.6
1.4
1.2
IDCorder
IDCrect
Iorder
1
IDCinv
0.8
0.6
0.4
0.2
0
0 0.5 1 1.5 2
VDComp
Vdcomp
Figure 18-21. VDCOL DC Current Order for Rectifier and Inverter
This dynamic simulation model is initialized from the power flow HVDC model and uses
all the pertinent information from the power flow.
The CDC7T model datasheet is in the PSS®E Model Library. The model uses 75 CONs, 17
STATEs, and 26 VARs, many of which can be used for plotting.
The model uses 4 integer variables (ICONs) to control the simulation and select the con-
trol configuration option. The first ICON can be used to simulate blocking/unblocking of
the HVDC. Normally it is set as zero. At any instant of the simulation setting ICON(M) =1
will start blocking the HVDC. It is simulated by reducing the DC current order at a rate
given by CON(J+52). Unblocking can be simulated by setting ICON(M)=2 by increasing
the DC current order at a rate given by CON(J+53).
18-41
ICON(M+2) is used to select the control configuration by setting its value respectively (1,
2 or 3). Accordingly, with the controllable variable margins and gamma order, these
parameters are:
CONs # Value Notation Description
J+26 IMargR Current margin, rectifier, pu
J+27 IMargI Current margin, inverter, pu
J+28 VMargR Voltage margin, rectifier, pu
J+29 VMargI Voltage margin, inverter, pu
J+30 GMargR Gamma margin, rectifier, pu
J+31 GMargI Gamma margin, inverter, pu
…
J+49 GAMA_ORDER1 Control configuration 1
J+50 GAMA_ORDER2 Control configuration 3
must be consistent with the selected control configuration.

By setting ICON(M+1) = 1 and VAR(L+24) (dc power order, p.u.) to any desirable value,
it is possible to simulate change in HVDC power including simulation of its overload
capability. The model allows obtaining the MW-input from the Auxiliary System model
DC2SIG and placing it into VAR(L).
The internal integration is embedded into the model that allows the use of normal integra-
tion step for PSS®E simulation.
18-42
0sec<T<1sec – unperturbed run

T=1.0 sec – dc line fault on inverter side of the pole 1
T=1.5 sec. – blocking pole 1 with the rate of 2000 A/sec.
T=5 sec. – overloading pole 2 by 50%
T=10 sec. – unblocking pole1 with the rate of 500 A/sec.
T=15 sec. – restore the system
Figure 18-22. Example of Complex Simulation for a dc Line with the Long Cable
Figure 18-23. Another Example of Complex Simulation for a dc Line with the Long Cable
18-43
18.5.4 Parameters of the DC circuit

Several CONs in the beginning of the model datasheet correspond to parameters of the dc
circuit and also provide the information about the HVDC rating:
CONs # Value Notation Description
J+2 LRR rectifier smoothing reactor inductance, mH
J+3 RRR rectifier smoothing reactor resistance, Ohm
J+4 LRI inverter smoothing reactor inductance, mH
J+5 RRI inverter smoothing reactor resistance, Ohm
J+6 LOHR inductance of O/H DC line from rectifier side, mH
J+7 ROHR resistance of O/H DC line from rectifier side, Ohm
J+8 LOHI inductance of O/H DC line from inverter side, mH
J+9 ROHI resistance of O/H DC line from inverter side, Ohm
J+10 LDCC inductance of DC Cable line, mH
J+11 RDCC resistance of DC Cable Line, Ohm
J+12 CDCC DC line capacitance, mkF
J+13 LF1 DC shunt 1 inductance, mH
J+14 RF1 DC shunt 1 resistance, Ohm
J+19 RCDCC Losses associated with the dc cable capacitance, Ohm
J+20 IDCRated Rated DC current, A
J+21 VDCRated Rated DC voltage, kV
The program calculates parameters of the equivalent circuit (Fig. 2.1) as:
• Equivalent resistance and inductance on the rectifier side
RDCR = ROHR+RDCC/2.
LDCR = LOHR+LDCC/2.
• Equivalent resistance and inductance on the inverter side
RDCI = ROHI+RDCC/2.
LDCI = LOHI+LDCC/2.
18-44
Program Application Guide: Volume II Models PAUX1, PAUX2, SQBAUX, CPAAUX, HVDCAU
18.6 Models PAUX1, PAUX2, SQBAUX, CPAAUX, HVDCAU

It is not practical for the dynamic simulation model library to accommodate every con-
ceivable form of controller that could be used to modulate the power or current setpoint of
a dc transmission system. The library does, however, have models of the most common
types used and proposed. Any combination and number of these can be connected to a sin-
gle dc line. (See Section 18.2.3 for discussion on how to apply the model’s output to the dc
line.) Models CPAAUX, SQBAUX and PAUX1 all give modulation based on deviation of
frequency from nominal. PAUX2 modulates based on angle deviation. Model HVDCAU
allows modulation based on branch current or power flow, frequency difference between
two buses, or voltage or frequency at a single bus. Though initially intended to be used
only as a GAMMA modulation signal, appropriate selection of CONs (gain) can allow it
to be used to modulate power or current. All the models will print an appropriate message
and be ignored if the bus specified by the user is not found. The timing and constants for
these models will vary based on the system the models are connected to and the intended
use of the modulations.
18.7 Models MTDC01 and MTDC03
18.7.1 General Description

The modeling philosophy of multiterminal models MTDC01 and MTDC03 are similar to
that of the PSS®E two-terminal models CDC4 and CDC6. All these models assume that
the local converter bridge controls and the associated response of dc current and voltage
are rapid in relation to the time scale of PSS®E simulations. Like the power flow modeling
of the multiterminal dc, this stability model assumes that the user can limit a certain
amount of communication by choice of data. Figure 18-24 shows the block diagram of the
dc line controls as modeled at all the converters except the voltage-controlling bus. The
control mode (MDC) is set in the power flow and should not be changed during a simula-
tion. Blocking of the entire line should be made by setting the appropriate ICON to 1. All
the other switches are set automatically by the model. Figure 18-25 shows the voltage-
dependent current limits (VDCL) logic modeled. Depending on the input data specified,
the curve will use either measured ac or dc voltage as an input. If V3 is set to zero, the
voltage-dependent current limits are ignored for this converter.
Like the two-terminal models, current following a block is controlled by a ramp based on
a user-supplied initial restarting current, RSCURi, and current ramp rate, CRMPi
(Figure 18-6). Unlike the two-terminal lines, which have several types of logic for block-
ing, an automatic block occurs in MTDC01 and MTDC03 only when a current reversal
occurs or voltage is so bad at a converter that it cannot produce the desired current. No
other automatic unblocking is run by this model; it is left to the user to control this func-
tion either by blocking manually via a change in the appropriate ICONs with ALTR or by
a user-written code in CONEC. Measured ac voltages and measured dc voltages are stored
in STATES and are available to the user. An example of a user-written blocking code is
shown in Section 18.7.2.
18-45
Models MTDC01 and MTDC03 Program Application Guide: Volume II
Measured ac
VDCL
or dc Voltage
Current Setpoint
MDC = 2 1. or + M M Bridge and
+
•  Ramp
I1  I I
I2
Simulation
Factor N N Logic
• + +
0. Ramping C0
• (A)
Redistributed
Current
Power
Setpoint 106   106 Modulating
(MW) Signal (MW)
VNominal
Measured dc
Mode Switch • • Voltage
or Ramping
Figure 18-24. Converter Controls
Voltage Dependent
Maximum Current Limit
Current (A) V3,C3
V2,C2
Permissible range in
normal operation
Possible range during

restart - C0 will
V1,C1 override
0,C0
RSCUR
Measured dc Voltage (kV)

Measured ac Voltage (pu)
Figure 18-25. Voltage-Dependent Current Limit for dc Converters
18-46
Program Application Guide: Volume II Models MTDC01 and MTDC03
Figure 18-24 shows a redistributed current exceptions: Taps are always blocked, and
value that a central controller would pro- each converter has a minimum angle speci-
vide. The central controller in this model is fied by model’s input data. Modulation is
assumed to know when a converter is assumed to be local. The user may, by writ-
blocked. If a converter is blocked or ramping a code, simulate any master controller
ing up, the difference between I1 (see modulation. In MTDC01, a modulation
Figure 18-24), where I1 must be greater than signal cannot be used directly by the volt-
or equal to Co, and desired current based age-controlling converter. However,
on set values is redistributed to all other because it acts, essentially, as a slack bus
converters according to the distribution by applying modulating signals to the other
factors (DCPF). The central controller also converters, the signal can be applied at this
is assumed to recognize a rectifier not able converter. Any of the dc modulating mod-
to make its desired current and will again els can be used. As with the two-terminal
cause a current redistribution based on dis- model, the user will have to edit CONEC
tribution factors. to place the VARs in the appropriate loca-
Figure 18-26 shows the control logic simu- tions. The voltage-dependent current limits
lated for the voltage-controlling bus. Its are ignored at the voltage-controlling bus
voltage recovery logic following a block is and this converter is not permitted to
like the two-terminal model shown in block. The IMIN value is checked and
Figure 18-7. The bridge and simulation warning messages will be printed when-
logic for this dynamic model is identical to ever current goes below this value at the
that used in the power flow with two voltage controlling bus.
Voltage Setpoint 1. or Vdesired Bridge and

Ramp Factor Simulation Logic
Figure 18-26. Voltage-Controlling Converter dc Control
(SETVAL) if they wish power/current to

MTDC01 is restricted to a model of up to 5 be redistributed to the other poles.
converters. This model will not allow rep-
resentation of negative converters or 18.7.2 Example
ground paths. Calls for the MTDC01/TTDC01 model are
MTDC03 allows any representation
automatically generated by DYRE and the
allowed in the power flow. During simula- procedure for setting this model up is simi-
tion, if the voltage controlling bus on all lar to that for the 2-terminal model. For the
rectifiers on a single pole are blocked, all
sample system described in Section 6.4.9,
rectifiers on a single pole are blocked and
Figure 18-27 shows the resulting DOCU.
all converters on that pole are blocked. The
Inspection of the data indicates that all
user must set all the appropriate ICONs on
measuring transducers are assumed to be
that pole to 2 to start ramping up of all set
instantaneous, i.e., they were entered as a
points. Note that redistribution of currents
zero. The user wished to block the entire
on this model is done on an individual pole
line, if the measured ac voltage dropped
basis. Users will have to modify set points
below a certain value at either rectifier, and
18-47
a block of just the inverter if the ac voltage next available CON would be 4392, and
at the inverter dropped below a particular double checked that data using RSTR.
value. The user noted from the DOCU that Figure 18-29 shows the code written to run
STATEs 64 through 78 and ICONs 16 the blocking and subsequent unblocking.
through 22 were being used by the model This code was placed in CONEC immedi-
and filled in the data sheet appropriately ately following the model.
(Figure 18-28). The user also noted that the
ADDITION OF DC LINE TO BASE CASE DEMO LOADFLOW
FOR PSS®E MANUAL
REPORT FOR CONEC MODELS BUS 200 [HYDRO 500]
*** CALL MTDC01( 1, 200, 64, 22, 16, 17) ***
MTDC# C O N ' S STATE'S V A R ' S ICON'S V A R ' S

1 200- 275 64- 78 22- 57 16- 22 17- 21
# X-- CONVERTERS --X

1 200 HYDRO 500
STATEs being 2 201 HYDRO 500 ICONs being used
used by model 3 160 INVJCT 18.0
4 161 INVJCT 18.0
DY1 TVAC1 TVDC1 TIDC1 RSVLT1 RSCUR1 VRMP1 CRMP1

3.000 0.000 0.000 0.000 0.00 50.00 0.000 3.000
C0-1 V1-1 C1-1 V2-1 C2-1 V3-1 C3-1

200.0 350.0 300.0 400.0 500.0 475.0 650.0

5.000 0.000 0.000 0.000 0.00 50.00 0.000 3.000
C0-2 V1-2 C1-2 V2-2 C2-2 V3-2 C3-2

200.0 350.0 300.0 400.0 500.0 475.0 650.0

10.000 0.000 0.000 0.000 0.00 50.00 0.000 3.000
C0-3 V1-3 C1-3 V2-3 C2-3 V3-3 C3-3

200.0 350.0 200.0 400.0 500.0 475.0 650.0

10.000 0.000 0.000 0.000 100.00 0.00 5.000 0.000
C0-4 V1-4 C1-4 V2-4 C2-4 V3-4 C3-4

200.0 0.0 0.0 0.0 0.0 0.0 0.0

0.000 0.000 0.000 0.000 0.00 0.00 0.000 0.000
C0-5 V1-5 C1-5 V2-5 C2-5 V3-5 C3-5

0.0 0.0 0.0 0.0 0.0 0.0 0.0
TCMODE DCMODE VCMODE

0.00000 2 400.000
ALF/GAM MIN MAX PAC QAC NB EBASE RC XC TR TAP

1 9.39 8.00 10.00 320.9 138.5 4 500.0 0.000 19.000 0.2200 0.0100
2 9.39 8.00 10.00 320.9 138.5 4 500.0 0.000 19.000 0.2200 0.0100
3 19.19 18.00 20.00 -300.0 141.4 4 230.0 0.000 10.000 0.4520 0.0100
4 19.19 18.00 20.00 -300.0 141.4 4 230.0 0.000 10.000 0.4520 0.0100
REPORT FOR CONET MODELS BUS 200 [HYDRO 500]
*** CALL TTDC01( 1, 200, 64, 22, 16, 17) ***
MTDC# C O N ' S STATE'S V A R ' S ICON'S V A R ' S

1 200- 275 64- 78 22- 57 16- 22 17- 21
# X-- CONVERTERS --X

1 200 HYDRO 500
2 201 HYDRO 500
3 160 INVJCT 18.0
4 161 INVJCT 18.0
Figure 18-27. Example DOCU for Four-Terminal Line
18-48
MTDC01
Multiterminal dc Line Model
CALL MTDC01 (I,J,K,L,M,LAUX) from CONEC

CALL TTDC01 (I,J,K,L,M,LAUX) from CONET
This is multiterminal dc line #______ I,
using CONs starting with #______ J,
and ICONs starting with #______ M,
and VARs starting with #______ LAUX.

J+14 650 C3-1, current limit point 3,
J 3.0 DY1, minimum angle con- converter 1(2) (amps)
verter 1 (degrees)
J+15 5. DY2, minimum angle con-
J+1 0. TVAC1, ac voltage transducer verter 2 (degrees)
converter 1 (sec)
J+16 0. TVAC2, ac voltage transducer
J+2 0. TVDC1, dc voltage transducer converter 2 (sec)
converter 1 (sec)
J+17 0. TVDC2, dc voltage transducer
J+3 0. TIDC1, current transducer converter 2 (sec)
converter 1 (sec)
J+18 0. TIDC2, current transducer
J+4 0. RSVLT1, minimum dc voltage converter 2 (sec)
following block, converter 1(1)
(kV) J+19 0. RSVLT2, minimum dc voltage
J+5 50. RSCUR1, minimum dc current (kV)
(amps) J+20 50. RSCUR2, minimum dc current
J+6 0. VRMP1, voltage recovery (amps)
rate, converter 1(1) (pu/sec)
J+21 0. VRMP2, voltage recovery
J+7 3. CRMP1, current recovery rate, converter 2(1) (pu/sec)
J+22 3. CRMP2, current recovery
J+8 200. C0-1, minimum current rate, converter 2(2) (pu/sec)
demand converter 1(4)
(amps) J+23 200. C0-2, minimum current
J+9 350. V1-1, voltage limit point 1, (amps)
converter 1(2) (kV)
J+24 350. V1-2, voltage limit point 1,
J+10 300. C1-1, current limit point 1, converter 2(2) (kV)
converter 1(2) (amps)
J+25 300. C1-2, current limit point 1,
J+11 400. V2-1, voltage limit point 2, converter 2(2) (amps)
converter 1(2) (kV)
J+12 500 C2-1, current limit point 2, converter 2(2) (kV)
J+13 475 V3-1, voltage limit point 3, converter 2(2) (amps)
converter 1(2) (kV)
18-49

converter 2(2) (kV)
verter 3 (degrees)
converter 3 (sec)
converter 3 (sec)
converter 3 (sec)
J+34 0. RSVLT3, minimum dc voltage
(kV)
J+35 50. RSCUR3, minimum dc current
(amps)
J+38 200. C0-3, minimum current
(amps)
J+39 350. V1-3, current limit point 1,
converter 3(2) (kV)
converter 3(2) (kV)
converter 3(2) (kV)
verter 4 (degrees)
converter 4 (sec)
converter 4 (sec)
18-50

converter 4 (sec)
J+49 100. RSVLT4, minimum dc voltage
(kV)
J+50 0. RSCUR4, minimum dc current
(amps)
(amps)
converter 4(2) (kV)
converter 4(2) (kV)
converter 4(2) (kV)
verter 5 (degrees)
J+61 0. TVAC5, ac voltage trans-
ducer converter 5 (sec)
J+62 0. TVDC5, dc voltage trans-
ducer converter 5 (sec)
converter 5 (sec)
J+64 0. RSVLT5, minimum dc
voltage following block,
converter 5(1) (kV)
J+65 0. RSCUR5, minimum dc cur-
rent following block,
J+66 0. VRMP5, Voltage recovery
18-51

(amps)
converter 5(2) (kV)
converter 5(2) (kV)
converter 5(2) (kV)
J+75 0. TCMODE (sec)
K 64 Measured ac voltage, converter 1*
K+1 65 Measured ac voltage, converter 1*
K+4 68 Measured ac voltage, converter 2
K+6 70 Measured ac voltage, converter 3†
*Rectifier voltages
†Inverter voltage
18-52
VARs # Description
L VAC bus converter 1
L+1 PAC bus converter 1
L+2 QAC bus converter 1
L+3 VDC converter 1
L+4 IDC converter 1
L+5 Angle converter 1
L+6 VAC bus converter 2
L+9 VDC converter 2
L+10 IDC converter 2
L+15 VDC converter 3
18-53
VARs # Description
L+30
L+31
L+32 Internal VARs required
L+33 by model
L+34
L+35
ICONs # Description
M* 16 Entire line blocking flag:
0 = Not blocked
1 = Blocked
2 = Unblocking
M+1 17 Converter 1 flag(3):
0 = Normal operation
1 = Blocked
2 = Unblocking
M+2† 18 Converter 2 flag(3)
M+3 19 Converter 3 flag(3)
M+6 22 Mode switch flag:
0 = Normal
1 = Mode switch
*ICON to be changed to block entire

line.
†ICON to be changed to block con-
verter 3 (inverter)
VARs # Description
LAUX Auxiliary signal converter 1(5)
LAUX+1 Auxiliary signal converter 2(5)
18-54
Notes:
(1) Used only at voltage controlling converter.
(2) Used at all except voltage controlling
converter.
(3) Set to zero if this is voltage controlling
converter (i.e., not used).
(4) Used as minimum current allowed even at voltage controlling bus.
(5) At voltage controlling inverter, only used if in gamma control (i.e., ANGMX = ANGMN in
power flow).
I, ’MTDC01’, DY1, TVAC1, TVDC1, TIDC1, RSVLT1, RSCUR1, VRMP1, CRMP1, C0-1, V1-1, C1-
1, V2-1, C2-1, V3-1, C3-1, DY2, TVAC2, TVDC2, TIDC2, RSVLT2, RSCUR2, VRMP2, CRMP2, C0-
2, V1-2, C1-2, V2-2, C2-2, V3-2, C3-2, DY3, TVAC3, TVDC3, TIDC3, RSVLT3, RSCUR3, VRMP3,
CRMP3, C0-3, V1-3, C1-3, V2-3, C2-3, V3-3, C3-3, DY4, TVAC4, TVDC4, TIDC4, RSVLT4,
RSCUR4, VRMP4, CRMP4, C0-4, V1-4, C1-4, V2-4, C2-4, V3-4, C3-4, DY5, TVAC5, TVDC5,
TIDC5, RSVLT5, RSCUR5, VRMP5, CRMP5, C0-5, V1-5, C1-5, V2-5, C2-5, V3-5, C3-5, TCMODE/
Figure 18-28. Example Case Data Sheet
CALL MTDC01( 1, 200, 64, 22, 16, 17)

C
C BLOCK ENTIRE LINE ON LOW RECTIFIER VOLTAGE Only check if blocked
C
IF (ICON(16).EQ.0)
Next available CON
. IF (STATE(64).LT.CON(4392) .OR. STATE(67).LT.CON(4392))
. . ICON(16)=1
. . WRITE(LPDEV,9992) Either voltage blocks
9992 . . FORMAT(' BLOCKED MULTI-TERMINAL BECAUSE OF LOW RECTIFIER ',
#. . 'VOLTAGE')
. ...FIN
...FIN
C
C UNBLOCK ON VOLTAGE RECOVERY
C
IF (ICON(16).EQ.1) Only check if blocked
. IF (STATE(64).GT.CON(4392) .AND. STATE(67).GT.CON(4392))
. . ICON(16)=2
. . WRITE(LPDEV,9993) Both voltages must recover
9993 . . FORMAT(' UNBLOCKED DC LINE - VOLTAGE RECOVERED') Print message is good idea
. ...FIN
...FIN
C
C BLOCK INVERTER ON LOW AC VOLTAGE
C
IF (ICON(19).EQ.0 .AND. STATE(70).LT.CON(4393))
. ICON(19)=1
. WRITE(LPDEV,9994)
9994 . FORMAT(' BLOCKED INVERTER ON LOW AC VOLTAGE')
...FIN
C
C UNBLOCK INVERTER ON RECOVERY
C
IF (ICON(19).EQ.1 .AND. STATE(70).GT.CON(4393))
. ICON(19)=2
. WRITE(LPDEV,9995)
9995 . FORMAT(' UNBLOCK INVERTER ON VOLTAGE RECOVERY')
...FIN
C
Figure 18-29. Example Code for Automatic Blocking
18-55
After reloading (linking) PSS®E, the user can then change the CONs to appropriate values
via ALTR. For an example run, the CONs were set to 0.8 per unit. Figure 18-30 shows the
converter power flows that result from a fault near the inverter. The plots show an initial
change when the fault is applied and removed and a one time step delay before blocking
and unblocking. The one time-step delay is the result of using measured values. The plots
show that immediately following the block, the inverter goes to a value and does not start
ramping up, because a RSCUR was specified as a value less than IMIN and the IMIN
overrides until the current ramps up to it. This example shows that the user will have to
take care in specifying the VDCL, IMIN, and restarting values for all converters to simu-
late reasonable response. In this example, the DCPFs were all equal so power was redis-
tributed evenly. The user does have the ability to change these values to simulate the
central coordinator performance.
18-56
Program Application Guide: Volume II Models CASEA1 and CDCRL
ADDITION OF DC LINE TO BASE CASE DEMO LOADFLOW
THU JAN 11, 1996 08:31

FOR PSS/E MANUAL
POWER
TECHNOLOGIES,
FILE: outfig17-23
POWERS
INC.R
CHNL# 45: [P INVERTER 4]

0.0 -5.000
CHNL# 44: [P INVERTER 3]
0.0 -5.000
CHNL# 43: [P RECTIFIER 2]
3.5000 1.0000
CHNL# 42: [P RECTIFIER 1]
0.40000 0.50000
3.5000 1.0000
0.45000
Voltage Controlling Inverter
0.35000
0.30000
Inverter Switch Blocked
TIME(SECONDS)
0.25000
Rectifiers
0.20000
0.15000
Unblocks
Blocks Here
0.10000
Removed
Fault
0.05000
Applied
Fault
0.0
Figure 18-30. Example Case
18.8 Models CASEA1 and CDCRL

These two terminal HVDC models include the L/R dynamics of the dc system and high-
speed controller dynamics. CDCRL is a generic model and is not intended to specifically
represent a particular HVDC system. It provides this level of modeling with a relatively
18-57
Models CASEA1 and CDCRL Program Application Guide: Volume II
simple controller and may be used to study future HVDC additions or to study installa-
tions where only limited information is available on the actual controls. CASEA1 was
developed for the Los Angeles Department of Water and Power based upon information
provided by ASEA to model the IPP HVDC system. A detailed definition of this model is
contained in ASEA report TR-PSC84009. Both of these models must be tuned properly to
give meaningful results, and their response may require considerable interpretation. They
are therefore not recommended for the inexperienced user. The level of representation pro-
vided by these models is only required in very special situations.
18.8.1 Internal Integration Used by Model

These models represent the L/R dynamics of the dc line and smoothing reactors as well as
the high-frequency firing angle controller dynamics. It is therefore necessary to use time
steps in the numerical integrations associated with these processes, which are much
shorter than those normally used by PSS®E. Normally in PSS®E, the ac network equations
are solved to update the voltages and currents at each integration time step. This process
would be very expensive and time consuming if the short time steps associated with the dc
process were used. Instead, the dc model uses an internal integration, which takes many
short time steps between each ac network solution.
Because the response of the HVDC system depends upon the ac voltage at the rectifier and
inverter terminals, and these voltages in turn depend upon the operation of the HVDC sys-
tem, it is necessary to approximate the ac voltages for times between the ac network solu-
tions. To do this, a two-port Thevenin equivalent of the ac system is used. At each internal
integration time step, the ac currents at the invertor and rectifier are calculated as a func-
tion of the ac voltages and the dc control and network dynamics. The Thevenin equivalent
is then used to calculate new ac voltages for the next time increment. This equivalent is a
two–port equivalent, which not only represents the complex self-impedance at the rectifier
and invertor, but also represents the complex coupling impedance. At each solution of the
ac network, the internal Thevenin voltages are updated.
Two additional refinements are required for this procedure. First, because the ac system is
dynamic, the internal Thevenin voltage may be moving with system swings. If this move-
ment is not accounted for, there will be discontinuities at each network solution. The
Thevenin complex voltage is therefore changed during the course of the internal integra-
tion at a rate that is determined by the discontinuities observed at previous network solu-
tions. The adjustment on the rate of change is subject to a relatively long time constant to
prevent numerical problems.
VTHEV = VTHEV + (dVTHEV) t
where:
VTHEV = Complex Thevenin voltage.
dVTHEV = Complex rate of change in Thevenin voltage.
Secondly, the Thevenin impedances may change as a result of faults or outages in the ac
network. In most cases, such changes will not be important. However, when a fault is rep-
resented at the terminals of the inverter or rectifier, the change would be large enough to
affect the solution results. Therefore, data must be entered for the ac network negative-
18-58
Program Application Guide: Volume II Models CASEA1 and CDCRL
and zero-sequence impedances at the inverter and rectifier and for the impedance of the
fault itself. When ICON M+11 is set and indicates a fault of a specified type at the inverter
or rectifier the program automatically adjusts the Thevenin impedance to account for this
fault. The user must still enter the equivalent fault admittance as a bus shunt in the net-
work. The HVDC model will check the value of the shunt to make sure that it is consistent
with the network sequence and fault impedances.
18.8.2 Initializing Thevenin Impedances

A better estimate for Thevenin impedances will result in fewer network iterations and less
time to run a simulation. The CASEA1 model requires entering initial values for the
Thevenin self- and mutual-impedances of the rectifier and inverter in CONs J+18 through
J+23. Equivalent CONs for model CDCRL exist. The model will automatically calculate
these values perturbing the current injected at each terminal and observe the change, V,
in voltage at both terminals. The complex impedances are of course V/I.
To activate this calculation, the user initializes the system using activity STRT, sets ICON
of M+1 to 1, and runs the simulation for at least eight time steps. The program will print
the desired Thevenin impedances. This activity should not be done as part of a normal
simulation because the perturbations introduce extraneous disturbances and it should be
done for only one HVDC system at a time.
The Thevenin reactance could also be estimated using the power flow as follows:
• Solve the system for the base case condition.
• Convert generators and load.
• Run ORDR, FACT, TYSL.
• Observe the voltage and var output at each converter.
• Change converter transformer tap position at one converter. This will have little effect
on the power or var output of the other converter if no control angle limits are
encountered.
• Solve using TYSL.
• The Thevenin reactance will be:
VMAG
XTHEV =
(VAR/VMAG)
• Repeat for other converter.
18.8.3 Simulating dc Faults

The CASEA1 and CDCRL models allow the user to simulate a fault on the dc line. The dc
fault resistance and the fault location on the dc line must be entered as modeling parame-
ters. If this is done, a dc fault can be applied by setting an appropriate ICON to 1. When
the fault is applied, the inverter current will quickly extinguish if its minimum firing angle
is set to keep it from operating as a rectifier. The rectifier will continue to supply fault cur-
rent unless its current setpoint is modified. To do this, set the VAR in which MYDCR is
contained to a negative value. After the fault current has extinguished and the deionization
18-59
Model CDCVUP Program Application Guide: Volume II
period has passed, the HVDC recovery may be indicated by resetting the ICON and VAR
to zero.
18.8.4 Simulating ac System Faults at the Converter Terminals

All ac system faults at the converter terminals are simulated by CASEA1 and CDCRL
using a procedure similar to that described for the CDCVUP model that follows.
18.9 Model CDCVUP

When either the inverter or rectifier are at control angle limits, this model uses equations
similar to those used in the power flow to calculate the power and var requirements of the
converters. It is similar in this respect to HVDC dynamic models such as CDC4 and
CDC6. The CDCVUP model, however, is also able to represent the dynamics of the tempo-
rary condition following a disturbance or during start-up when neither the inverter nor the
rectifier are at control angle limits. This model provides a better representation of the cur-
rent controller response under such conditions. The details of this model are described in
"HVDC Models Used in Stability Studies" presented at the 1988 PAS Summer Power
Meeting.
Figure 18-31 shows the output from DOCU for this model.
18.9.1 Simulating ac System Faults at the Converter Terminals

When an ac system fault is simulated at one of the converter terminals, the appropriate
ICON should be set to tell the model the fault type, i.e., line-to-ground, etc. This allows
the model to vary its response based upon the individual phase voltages instead of just
basing it on the positive-sequence voltage.
An ac system line-to-ground fault at the inverter using CDCVUP was simulated. Such a
fault would probably cause commutation failures, which are modeled by setting the appro-
priate ICON to 1 and changing the shunt admittance to represent a converter transformer
shorted on the dc side. Figure 18-32 is a plot of the inverter and rectifier current order and
the direct current for this simulation.
18.10 Model DCPOW

Model DCPOW can be used with the CDCVUP and CDCRL two-terminal HVDC models
to represent the dynamics of HVDC power controllers. It also represents current margin
makeup, which raises both the inverter and rectifier current setpoints to achieve the
desired level of current when the inverter assumes current control. This controller is nor-
mally much slower than the HVDC pole controls. Typical values for the model parameters
follow: TVDCP = 0.1, VDCP-MAX = 1.0, VDCP-MIN = 0.85, and RESET = 4.0.
Figure 18-33 shows a block diagram for the model.
18-60
Program Application Guide: Volume II Model DCPOW
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E THU JAN 11, 1996 13:41
SAMPLE SYSTEM FOR PSS®E MANUAL
1100KV DC CASE
REPORT FOR CONEC MODELS BUS 1403 [WDUM 18.0]
*** CALL CDCVUP( 1, 7, 299, 96, 41) ***
DC# ICON'S C O N ' S STATE'S V A R ' S

1 7- 11 299- 339 96- 97 41- 74
X-- RECTIFIER ---X X--- INVERTER ---X MDC RDC RCOMP SETVAL VSCHED VCMODE DELTI
1600 MINE 765 1403 WDUM 18.0 1 8.20 0.00 1600.0 554.4 0.0 0.1500
ALF/GAM MIN MAX PAC QAC NB EBASE XC TR TAP

R: 7.54 5.00 40.00 16.000 4.678 2 500.0 3.880 0.44000 1.02500
I: 18.95 18.00 30.00 -15.370 6.558 2 230.0 3.047 0.96500 0.99375
BLFLAG FLTYP BYFLAG BALIM INVCT

ICON( 7) ( 8) ( 9) ( 10) ( 11)
0 0 0 0 0
DC RATINGS
DC(AMPS) VDC(KV)
CON( 299) ( 300)
6000.00 500.00
RECTIFIER VOLTAGE SENSITIVE CURRENT LIMIT

C0 C1 V1 C2 V2 V3
CON( 301) ( 302) ( 303) ( 304) ( 305) ( 306)
0.150 0.290 0.000 0.290 0.160 0.550
TV1 TV2
CON( 307) ( 308)
0.010 0.010
INVERTER VOLTAGE SENSITIVE CURRENT LIMIT

C0 C1 V1 C2 V2 V3
CON( 309) ( 310) ( 311) ( 312) ( 313) ( 314)
0.150 0.290 0.000 0.290 0.160 0.550
TV1 TV2
CON( 315) ( 316)
0.010 0.040
FIRING ANGLE CONTROLLER

GI GPR GPI MX DA/DT
CON( 317) ( 318) ( 319) ( 320)
3600.0 0.5 0.5 6000.00
RECTIFIER MINIMUMS INVERTER MINIMUMS

GAMA ALFA GAMA ALFA
CON( 321) ( 322) ( 323) ( 324)
15.0000 6.0000 17.0000 101.0000
INVERTER TRANSITION SLOPE CON( 325)

D(DCV)/D(IDC) = 0.430
POSITIVE SEQUENCE IMPEDANCES

RESISTANCE REACTANCE RESISTANCE REACTANCE RESISTANCE REACTANCE
CON( 326) ( 327) CON( 328) ( 329) CON( 330) ( 331)
REC 0.00000 0.00980 INV 0.00240 0.00830 MUTUAL 0.00000 0.00000
ZERO SEQUENCE IMPEDANCES

RESISTANCE REACTANCE RESISTANCE REACTANCE
CON( 332) ( 333) CON( 334) ( 335)
RECTIFIER 0.00001 0.00367 INVERTER 0.00000 0.00827
1-PHASE FAULT IMPEDANCES

RESISTANCE REACTANCE RESISTANCE REACTANCE
CON( 336) ( 337) CON( 338) ( 339)
RECTIFIER 0.00000 0.00706 INVERTER 0.00000 0.00000
Figure 18-31. Example DOCU for CDCVUP

(Sheet 1 of 2)
18-61
Model DCPOW Program Application Guide: Volume II
TV1 TV2
CON( 85) ( 86)
0.010 0.040
FIRING ANGLE CONTROLLER
GI GPR GPI MX DA/DT

CON( 87) ( 88) ( 89) ( 90)
3600.0 0.5 0.5 6000.0
ICON( 5)
FLAG FOR TYPE OF BRIDGE ANGLE LIMITS = 0
FOR MINIMUM FIRING ANGLE LIMIT FLAG .GE. 0

FOR MINIMUM COMMUTATION VOLTAGE FLAG .LT. 0
FOR MINIMUM MARGIN ANGLE LIMIT ABS(FLAG) .NE. 1
FOR MINIMUM MARGIN AREA LIMIT ABS(FLAG) .EQ. 1
MINIMUM ALPHA
RECTIFIER INVERTER
CON( 92) ( 94)
6.0000 101.0000
MINIMUM GAMA
RECTIFIER INVERTER
CON( 91) ( 93)
15.0000 17.0000
INVERTER TRANSITION SLOPE CON( 95)

D(DCV)/D(IDC) = 0.430
POSITIVE SEQUENCE IMPEDANCES
RESISTANCE REACTANCE
CON( 96) ( 97)
RECTIFIER 0.0000E+00 0.9800E-02
CON( 98) ( 99)

INVERTER 0.2400E-02 0.8300E-02
CON( 100) ( 101)

MUTUAL 0.0000E+00 0.0000E+00
ZERO SEQUENCE IMPEDANCES

CON( 102) ( 103)
RECTIFIER 0.1000E-04 0.3670E-02
CON( 104) ( 105)

INVERTER 0.0000E+00 0.8270E-02
FAULT IMPEDANCES
CON( 106) ( 107)
1-PHASE RECTIFIER FAULT 0.0000E+00 0.7060E-02
CON( 108) ( 109)

1-PHASE INVERTER FAULT 0.0000E+00 0.0000E+00
ICON( 3)
FLAG FOR FAULT LOCATION = 0
FOR DC LINE-TO-GROUND FAULT FLAG .EQ. 1

FOR 1-PHASE RECTIFIER FAULT FLAG .EQ. 2
FOR 1-PHASE INVERTER FAULT FLAG .EQ. 3
ICON( 6) EQUAL TO 1 INDICATES THAT THE INVERTER VOLTAGECONTROL IS ACTIVE
Figure 0-0 (Cont). Example DOCU for CDCVUP (Sheet 2 of 2)
18-62
Program Application Guide: Volume II Model DCPOW
14:37
1100KV DC CASE
POWER
TECHNOLOGIES,
INVERTER SLG FAULT & MULTIPLE COMMUTATION FAILURES
INC.R POWER CONTROL OFF & VDC CONTROL OFF
THU JAN 11, 1996

DC & DESIRED DC
FILE: outfig17-25
CHNL# 3: [DESDCI]
0.50000 0.0
CHNL# 2: [DESDCR]
0.50000 0.0
CHNL# 1: [DCCUR1]
0.50000 0.0
1.0000
0.90000
0.80000
0.70000
0.60000
TIME(SECONDS)
0.50000
0.40000
0.30000
0.20000
0.10000
0.0
Figure 18-32. Response to Inverter Fault With Commutation Failures

(Sheet 1 of 2)
18-63
Model DCPOW Program Application Guide: Volume II
14:40
1100KV DC CASE
VDC
POWER
TECHNOLOGIES,
INVERTER SLG FAULT & MULTIPLE COMMUTATION FAILURES
INC.R POWER CONTROL OFF & VDC CONTROL OFF
FILE: outfig17-25
THU JAN 11, 1996

CHNL# 5: [VDCI]
2.0000 0.0
CHNL# 4: [VDCR]
2.0000 0.0
1.0000
0.90000
0.80000
0.70000
0.60000
TIME(SECONDS)
0.50000
0.40000
0.30000
0.20000
0.10000
0.0
Figure 0-0 (Cont). Response to Inverter Fault With Commutation Failures (Sheet 2 of 2)
18-64
Program Application Guide: Volume II Model CDCAB1
MYSET(L)
+
+

SETVAL(I)/DCRATE
+
DELDCP(I)
+
• (2)
+ P0 MDC(I) +

P0 VDCP
• •  DCREF(I)
SETVAL(I)/PRATE (1)
+ DCREF(I)*DELT(I) +
VDCP(K)
MYSET(L) +
RESET(K)
VPMAX  RESETG*
S
–
KA
0.0
1 + sTVDCP DCR(I)
1.0
VPMIN
SETVAL(I)>0 • • SETVAL(I)<0
VDCR(I) VDCI(I)
*Integrator reset to zero if bypassed, blocked, or on VDCL.
Figure 18-33. Block Diagram Showing Power Control for dc System (DCPOW)
18.11 Model CDCAB1

Model CDCAB1 was developed by the Technical University of Denmark in cooperation
with NESA A/S and ABB Power Systems, sponsored by the Academy of Technical Sci-
ences of Denmark. A detailed technical write-up is available from the Technical Univer-
sity of Denmark, Publication No. 9703. The following is excerpted from this report.
CDAB1 is considered to be a general detailed monopole dc line model though it is based
on a vendor-specific design (ABB). The model includes the converters, converter trans-
formers, dc line, and the HVDC control system. In addition to the general features, spe-
18-65
Model CDCAB1 Program Application Guide: Volume II
cific control functions pertinent to the Kontek HVDC interconnection are included (i.e.,
power modulation, frequency modulation, and emergency power control). The model also
includes an inverter voltage controller for long dc cables as well as several dynamics
and/or transient control functions. CDCAB1 features a system independent integration
time step, which allows important details of the dc control system to be included without
severe consequences to overall simulation time.
The model including the converter representation, the basic control features, and the extra
control functions are described in the following sections. A block diagram illustrating the
components of the HVDC system is shown in Figure 18-34.
18.11.1 Master Controller

The HVDC system can be either in power control or in current control. When the HVDC
system is in current control, the current order is limited by a maximum limit
IOMAX_MASTER before being sent to the VDCOL. The modulation signals are ignored
when the control system is in current control.
The block diagram for the master controller is shown in Figure 18-35 for the HVDC link in
power control. The current order is found by dividing the ordered power by a filtered
value of the absolute value UDC_ORDER of the measured dc voltage.
In power control, the master controller sets the total power order, which consists of the
actual ordered value, and the modulation signals from the frequency controller, the damp-
ing controller, and the emergency power controller (EPC). The current order correspond-
ing to the power order is calculated in the lead station. The calculation of the current order
is based on a local dc voltage measurement and is delivered to the rectifier and the inverter
VDCOLs.
Models of frequency control, damping control, and emergency power control are
described in the following sections. When the link is in power control, the modulation sig-
nals are added to the power order. In current control, they are ignored. When the emer-
gency power control is operating, the other modulation signals are suspended.
A local frequency less than the nominal frequency (fnom) of the connected ac network
results in a reduction of the power for rectifier operation and an increase of power for
inverter operation, and conversely if the frequency in the converter station is greater than
the nominal frequency.
18-66
Figure 18-34. Diagram of the HVDC System with Connections to the ac Systems
Figure 18-35. Master Controller in Power Control
18-67
18.11.2 Frequency Control

A model of the frequency controller used in the Kontek control system is shown in
Figure 18-36. This frequency controller is of the proportional type, meaning that a station-
ary frequency deviation follows from a change in load. The droop of this frequency con-
troller corresponds to the proportional gain K_FREQ_(R,I).
The variables DELT_PO_FREQ_(R,I) are always added to the power order signal accord-
ing to Figure 18-35. Because the frequency control does not contribute to the modulation
signal when the emergency power controller (EPC) is operating, the value of
DELT_PO_FREQ(R,I) in Figure 18-35 is equal to the output from the frequency controller
when the EPC is not operating, and 0 when the EPC is operating. In this way, the output
from the frequency controller (state 19 and 20) always corresponds to the actual frequency
deviation.
18.11.3 Damping Control

A model of the damping controller for the Kontek interconnection is shown in Figure 18-
37, where T2DAMP > T1DAMP.
When the output from the damping controller is at the upper or lower limit, the values of
the internal variables, named state 21, 22, 23, and 24 in Figure 18-37, are kept unchanged to
avoid an uncontrollable increase of these states.
The variables DELT_PO_DAMP_(R,I) are always added to the power order in the master
controller in Figure 18-35. As for the frequency controller, the damping control signal is
added to the power order only if the emergency power controller is not active. In this case,
DELT_PO_DAMP(R,I) are set equal to the values from the damping controllers and when
the EPC is operating the DELT_PO_DAMP_(R,I) values are 0.
18.11.4 Emergency Power Control

The activation of the emergency power controller (EPC) for the Kontek Interconnection
immediately suspends other power modulation signals. When the HVDC link is in current
control, the EPC and other modulation signals are ignored.
The EPC of the Kontek control system is activated by different criteria. Some are invoked
when the frequency deviation has exceeded certain values for a specified time, some when
the ac voltage of the converter bus is less/greater than specified values for a specified time,
and finally some are activated by external signals.
Depending on the priority of the activated criteria (entries) the power order is ramped
up/down with a specified rate. As soon as an entry of a higher priority is activated, the cur-
rent entry is temporarily stopped on its current value until the entry with the higher prior-
ity has increased/decreased the power order with the specified value. In this way, the
active entry of the highest priority determines the rate of change of the power order.
A model of the EPC with power steps activated by a specified frequency deviation for a
specified time is used.
18-68
Figure 18-36. Frequency Controller
Figure 18-37. Converted Block Diagram for the Damping Controller
18.11.5 Voltage Dependent Current Order Limiter (VDCOL)

A block diagram of the VDCOL is shown in Figure 18-38. The VDCOL restricts the cur-
rent order when the dc voltage is low, according to Figure 18-39. Note that the maximum
current order IOMAX also depends on the actual current order IOO from the master con-
troller in this specific VDCOL.
In situations where the ac voltage falls (e.g., because of an ac fault), the dc voltage also
falls due to the slow reaction of the tap changers. In power control, the dc current order
without the VDCOL would increase, and this would result in a larger reactive power con-
sumption from the ac network, which would reduce the ac voltage furthermore. In these
situations, the VDCOL limits the dc current increase and prevents a voltage collapse in the
ac system. Also in situation where the bypass pair is fired, the VDCOL reduces the current
so the thyristors in the bypass pair are not overloaded.
The measured dc voltage is used in the VDCOL. A decreasing dc voltage results in a faster
reduction of the maximum limit for the current order from the VDCOL than the increasing
of the current order for an increasing dc voltage. The time constant T_UP is therefore
greater than T_DOWN. This results in a quick reaction during an ac fault and gives, on the
18-69
other hand, the ac system time to recover form a fault. The time constant is initially set to
T_DOWN.
18.11.6 Current Control Amplifier (CCA)

The CCA converts the current order from the VDCOL to an alpha order value. The differ-
ence between the current order and the measured dc current (and the current margin CUR-
MARG for the inverter) defines the current error I_ERROR that is fed into a PI controller
giving the alpha order.
The CCA takes care of the rectifier current control, the inverter voltage control, and the
inverter current control according to the static Ud/Id characteristic as shown in Figure 18-
40. The CCA always tries to increase the alpha order if the measured dc current is greater
than the current order (minus the margin for the inverter) and the other way around if it is
less.
In normal operation, the rectifier controls the current because the current order given to
the rectifier is greater (by the margin) than the current order to the inverter. This means
that the inverter normally controls the dc voltage. Hence, the inverter alpha order will
always be in the upper limit determined by the alpha-max segment. If the dc voltage in the
rectifier cannot be sufficiently increased (either by decreasing alpha or changing the tap
changers of the converter transformer) the inverter will take over control of the current
and alpha order for the rectifier will be the lower limit, because the current order is higher
than the actual dc current. This phenomenon is known as mode shift. In this situation, the
rectifier is, in a sense, voltage controlling because the dc voltage is determined by the ac
voltage and the minimum alpha.
The CCA is basically a PI controller. The integral part of the CCA is of the nonwindup
type; meaning, that the output from the integrator is kept on its upper or lower limit in case
one of these limits are reached. The proportional part of the controller is formed by a lin-
earization around the steady-state operating point because small changes in the firing
angle mainly come from the proportional part.
18-70
Figure 18-38. Voltage Dependent Current Order Limiter
Figure 18-39. VDCOL Characteristic (Current Order as a Function of the dc Voltage)
18-71
Figure 18-40. Ud/Id Characteristic for the dc Model without the VDCOL
18-72
In steady state, there is no contribution to alpha order from the proportional part for the
rectifier because the current error is zero. The contribution for the inverter proportional
part of the CCA is on the contrary nonzero because the actual dc current in steady state is
greater than the current order from the inverter VDCOL minus the current margin. How-
ever, because the D current in the inverter is too large compared to the ordered value, the
integral part of the alpha order will be in the upper limit determined by the alpha-max seg-
ment.
From Figure 18-41, the input to the CCA is supplied with an optional input. This input item
is used when the Inverter Gamma 0 Start function for the inverter is active. This function
is described later.
18.11.7 Alpha-Max Segment

Figure 18-42 shows the block diagram of the -max segment. This model is almost identi-
cal with what is done in the real control system of Kontek. The alpha-max segment is the
section of the inverter characteristic with the positive slope as seen from the Ud/Id charac-
teristic shown in Figure 18-40. The nominal operating point is lying somewhere on this
segment, which is composed of two parts. The first part keeps the firing angle  constant
and is obtained from the fundamental converter equations, which give:
2 Idc Xc
cos = - cos
Eac
To represent the Ud/Id characteristic as shown in Figure 18-40, the current inserted in the
above equation is a filtered value of the current order and the ac voltage is a measured
voltage that is multiplied with the actual transformer ratio (TR in the figure).
Figure 18-41. Current Control Amplifier
18-73
Figure 18-42. max Segment
In the model, the  value (named GAMREF in Figure 18-42) is obtained from the power
flow solution. Its value is kept constant during the dynamic simulation. In the real control
system, the value of  inserted is the nominal value. The second part contributes to an
additional increase in the positive slope and is proportional to the difference between the
current order and the measured dc current (K1 in Figure 18-42).
The upper limit for the output from this controller ALPHA_MAX_ORDER_(R,I) is (180o
-GAMMAMIN_(R,I), where GAMMAMIN_(R,I) is the minimum value of inverter
gamma for the HVDC link in nominal operation. ALPHA_MAX_ORDER_(R,I) is lim-
ited downward by a constant value named MIN_AMAX_(R,I). The output from the alpha-
max segment is used as an upper limit for the firing angle in the CCA.
18.11.8 Converger Firing Control (CFC)

The block diagram for the CFC including dynamic current compound and phase shift cor-
rection is shown in Figure 18-43. The CFC model is responsible for sending the firing
pulses to the valves. The model of the CFC determines the value of alpha put into the con-
verter equations. The firing of the valves are assumed to be equidistant.
The variable DELTA_LIM in Figure 18-43 represents the change in firing angle per com-
mutation. Limits on the change in alpha per commutation are imposed and are determined
from Figure 18-44. Some special considerations for the inverter must be made because it is
operated close to the minimum gamma limit in order to reduce the reactive power con-
sumed. After the conduction period, the carriers must be cleared out of the thyristor junc-
tions to make the thyristors capable of withstanding the forward blocking voltage.
Therefore, the valves must be biased by a specific negative voltage-time area to be able to
commutate properly and in this way correct commutation depends on the remaining volt-
18-74
age-time area of the commutation voltage from the end of the overlap (at time / - tu)
and until the voltage becomes positive (at time /). The commutation voltage is propor-
tional to Udiosin(t) so the voltage-time area A(V-t) is proportional to:


Udio
A(V - t) 

 Udio sintdt =

(1 - costu)
- tu

The nominal commutation margin nom at nominal voltage and frequency holds a margin
() to the minimum allowed commutation margin. For a given voltage and frequency, the
minimum commutation margin min is found from the minimum voltage-time area that is
equal to the voltage-time area at nom- at nominal voltage and frequency. Because tu
expresses the commutation margin, the above equation can be used to find a reasonable
value for min for the CFC. This gives:
Udio UdioN
 (1 - cosmin) = nom (1 - cos(nom - ))
Udio
cosmin = 1 - (1 - cos(nom - ))
Udionom
Figure 18-43. CFC with Dynamic Current Compound and Phase Shift Correction
18-75
Figure 18-44. Limits on the Change in  per Commutation
Thus, the ordered firing angle from the CFC must not exceed the value corresponding to
the calculated min value. The upper limit of the inverter CFC is thus determined from:
ALPHA_MAX_I = 180o -min - u
where u is the actual overlap angle found from the measured inverter dc current and mea-
sured inverter ac voltage.
The firing of the valves are related to the commutation voltage at the connected ac bus. If
the phasor voltage angle changes, this represents a displacement in time of the ac phase
voltages. In the real control system, the firing of the valves is advanced/delayed according
to this displacement, and it takes some time before the control system phase locks the fir-
ing pulses to the new phase voltages. This effect is modeled by the phase shift correction
both in the rectifier and in the inverter control systems and it is shown in the upper part of
Figure 18-43. The phase shift correction outputs an increment (DANG) in the firing angle,
which is added to the firing angle from the CFC. DANG increases to positive values for
restart in a weak inverter network and decreases to negative values in a weak rectifier net-
work. This means that the firing angle during restart is increased for the inverter and
reduced for the rectifier leading to increased dc voltages, by which the dc current increase
in the inverter is slowed down, where it on the contrary is speeded up in the rectifier. This
effect is very essential in weak networks where the phase angle change is significant.
DANG should also be subtracted from the limits ALFA_MAX_(R,I) and
ALFA_MIN_(R,I) on STATE(K+11) and STATE(K+12) adjusting the limits to the new
zero-crossing of the phase voltages. The firing angle sent to the control system can in this
way never be out of the limits from ALFA_MIN_(R,I) to ALFA_MAX_(R,I) correspond-
ing to a protective firing at these limits.
The lower part of Figure 18-43 shows a model of the dynamic current compound function
in the inverter end, which exists in the control system for Kontek. A contribution to the fir-
18-76
ing angle is added just before the firing. The aim of this function is to improve the small
signal stability of the inverter control system. In steady state the inverter is operated close
to the constant gamma part of the Ud/Id characteristic, which has a negative slope, mean-
ing that the rectifier sees the inverter as a negative resistance because the dc voltage
decreases with increasing dc current. For small disturbances around the operation point
the dynamic current compound function will transiently make the inverter look like a pos-
itive resistance.
18.11.9 Converters and Converter Transformers

The input required for and the output generated by the converters and converter transform-
ers is:
Input: • Firing angle ALFA_(R,I) (state) from CFC.
• dc current.
Output: • dc voltage (state).
The commutations of the single valves are not represented in the model. The converters
are represented by the fundamental equations as seen in Figure 18-45 based on the rms
value of the commutation ac voltage. The commutation resistance and inductance are used
in the converter equations together with the tap changer position, which is not changed
during the dynamic simulation as mentioned earlier.
18-77
Rectifier:
32
(R1)Udio =
 Eac
3Xc Idc
(R2)Udc = N Udio cos - - 2Rc Idc

(R3)Eac = TR U
ac, net
TAP
2 Idc Xc
(R4)cos(u + ) = cos -
Eac
2u + sin2 - sin(2u + 2)

(R5)tan =
cos2 - cos(2 + 2u)
18 I (cos2 - cos(2 + 2u))2 + (2u + sin2 - sin(2 + 2u))2

(R6)Pac,rect = dc Eaccos
 4(cos - cos( + u))

(R7)Qac,rect = dc Eacsin
 4(cos - cos( + u))
Inverter:
32
(I1)Udio =
 Eac
3Xc Idc
(I2)Udc = N Udio cos  - - 2Rc Idc

TR U
(I3)Eac = ac, net
TAP
2 Idc Xc
(I4)cos(u + ) = cos - = - cos 
Eac
2u + sin2 - sin(2u + 2)

(I5)tan =
cos2 - cos(2 + 2u)

(I6)Pac,inv = dc Eaccos
 4(cos - cos( + u))

(I7)Qac,inv = dc Eacsin
 4(cos - cos( + u))
Figure 18-45. CDCAB1 Algebraic Equations

(Sheet 1 of 2)
18-78
Symbols for Figure 18-45:

Uac,net rms value of the ac voltage on the network side
Eac rms value of the ac voltage on the valve side of the converter-transformer
Udc Average value of the DC voltage at the converters
Udio Ideal no-load DC voltage
UdioN Nominal ideal no-load DC voltage
Idc DC current
Xc Commutating reactance per bridge
Rc Commutating resistance per bridge
TR Transformer ratio
TAP Tap setting
N Number of bridges
 Firing angle
 Extinction angle or commutation margin
u Overlap angle
 Power angle
Pac,Qac Active and reactive power consumed from the ac network
Figure 0-0 (Cont). CDCAB1 Algebraic Equations (Sheet 2 of 2)
It could, however, be an advantage if the user is allowed to change the tap changer posi-
tion manually during the dynamic simulation as in PSS®E. Only symmetrical conditions
are represented. Therefore, the model is only valid for equidistant fixing of the values.
The converter equations assume single overlap (i.e., an overlap angle less than 60). This
is valid when the dc current is not too high (10 pu at normal ac voltage) or if the ac com-
mutating voltage is not too low (0.1 pu at normal dc current). Even during bypass or
blocking the overlap angle only exceeds 60 in a very short time interval. The error intro-
duced by using the same converter equations when the overlap angles are greater than 60
is therefore very limited. Another way could be to set the overlap angle equal to 60 in the
calculations when it goes beyond 60. By keeping the value of the overlap angle on 60,
the equations give a dc voltage that is too big and changes more slowly than the actual
value does, because it is not sensible to the actual change in the overlap angle. If the over-
lap angle greater than 60 instead is used in the equations the dc voltage is able to respond
to changes in the overlap angle although the assumptions of the equations are not com-
pletely fulfilled. The latter is therefore a better solution through still an approximation.
18-79
18.11.10 Cable Model and Current Return Path

The input required for and the output generated by the cable model and current return path
is:
Input: • dc voltage from the converters UDC_(R,I).
Output: • dc currents IDC_(R,I).
• Measured dc currents.
• Measured dc voltages.
• Voltage over the cable capacitance and sheath resistance
(state 6).
The cable is represented by discrete elements in a t-equivalent. The line capacitance is rep-
resented by a capacitance (CC) in the middle of the cable. In reality, the cable capacitance
is primarily a capacitance from the conductor to the sheath. To represent the damping of
the cable, a resistance (RC) equal to the resistance of the sheath is therefore placed in
series with the cable capacitance to the return. Half of the total line resistance (R_(R,I))
and inductance (L_(R,I)) is placed on each side of the cable capacitance. The resistance
(RS_(R,I)) and inductance (LS_(R,I)) of the smoothing reactor are also included.
Instead of a t-equivalent, the cable could be represented by a pi-equivalent. It must then be
noted that as the number of differential equations equals the number of inductances and
capacitances, it introduces an extra state variable. In this way, the calculation time will be
increased. A pi-equivalent and a t-equivalent is considered to give the same accuracy.
The current return path is in this case assumed to be ideal but can, of course, be repre-
sented by discrete elements like for the dc line, if, for example, it is an electrode line or a
metallic return. As mentioned above, the number of capacitances and inductances must
carefully be considered and only included if it is necessary to model the return with higher
accuracy. In the case of Kontek the voltage drop over the return is not more than 1 kV
(0.25%) and it will not significantly improve the results to model the current return path.
The correlations between the voltages and the currents are described by a differential
equation for each capacitance and inductance. These equations are given below. The com-
mutation inductance LC_(R,I) enters into the calculation of the current derivative, only
when the converter is not bypassed. LC_(R,I) depend on the actual overlap angle, which
means that its value basically lies between 1.5 and 2 times the commutation inductance
per phase.
Current direction from inverter to rectifier (negative dc voltage in steady state):
18-80
ICC = IDC_I - IDC_R
diDC_R (UCC + RCICC) - UDC_R - (R_R + RS_R)IDC_R

=
dt L_R + LS_R + LC_R
diDC_I UDC_I - (UCC + RCICC) - (R_I + RS_I)IDC_I

=
dt L_I + LS_I + LC_I
dUCC ICC
=
dt CC
Current direction from rectifier to inverter (positive dc voltage in steady state):
ICC = IDC_R - IDC_I
diDC_R UDC_R - (UCC + RCICC) - (R_R + RS_R)IDC_R

=
dt L_R + LS_R + LC_R
diDC_I (UCC + RCICC) - UDC_I - (R_I + RS_I)IDC_I
=
dt L_I + LS_I + LC_I
dUCC ICC
=
dt CC
18.11.11 Special Control Functions

The HVDC control system is affected by changes in the ac voltages. Some of the most
important control actions as a result of too low ac voltages from the real control system
should be included in the model. The rectifier transient controller and inverter gamma 0
start were found to be important for the simulation results and are described in the follow-
ing paragraphs.
18.11.11.1 Rectifier Transient Controller

This function is active during and after a fault in the rectifier ac network. During the fault,
the dc voltages and the cable voltage are reduced and accordingly the firing angles are at
or close to their minimum values. When the ac voltage suddenly comes back, the rectifier
dc voltage increases immediately and an overcurrent arises due to the difference between
the large rectifier voltage and the low cable voltage. The smoothing reactor is also a deci-
sive factor for the overcurrent. The overcurrent causes unwanted disturbances in the recti-
fier ac network so a control action that limits the overcurrent is essential to prevent
situations like this.
The required effect is achieved by the rectifier transient controller function where the min-
imum value of the firing angle in the CCA is increased during and after the fault according
to Figure 18-46. The function is activated and respectively deactivated when the rectifier ac
voltage is below and respectively above a limit that is named UAC_TRCONR in
Figure 18-46. The minimum limit of the firing angle to the CCA is increased to
18-81
MAX_AORDER_MIN_R during the fault and is ramped down after the fault to the nor-
mal minimum limit value with a rate of D_AORDER_MIN_R. With this function, the rec-
tifier dc voltage after the fault is cleared, is limited due to the greater minimum value of
alpha, and, in this way, the risk of a large overcurrent after the fault clearing is reduced.
Also, the rectifier transient controller tends to prevent peak rectification because of the
delaying of the firing to the valves. Peak rectification is a phenomenon that occurs when
the valves start to conduct later than defined by the firing angles because the cable voltage
is greater than the converter voltage at the given alpha. The valves conduct only at the
peaks of the commutating voltages.
Figure 18-46. Transient Controller Rectifier
18.11.11.2 Inverter Transient Controller

During a fault in the inverter ac network, there is a considerable risk of commutation fail-
ure. Immediately after the fault removal, the recovery voltage is more or less distorted,
which increases the risk of repetitive commutation failures when the inverter valves start
to conduct. This risk is reduced by increasing the gamma value giving larger margin for
the valves to commutate. This is the idea behind the inverter transient controller; the prin-
ciple in this function is shown in Figure 18-47.
The function is activated when the ac voltage in the inverter falls below a limit, named
TRCONI_ACVOLT_ACTIVE, and is deactivated when the ac voltage exceeds
TRCONI_ACVOLT_DEACTIVE. During the fault, the reference gamma is increased
with TRCONI_DGAMA with a small time constant TRCONI_TUP. After removal of the
18-82
fault, gamma decreases to the initial value with a time constant TRCONI_TDOWN,
which is in the order of the time required for the ac voltage to stabilize after a fault.
18.11.11.3 Inverter Gamma 0 Start

During a fault in the inverter ac network, the firing angle in the inverter is decreased to its
minimum limit while the firing angle in the rectifier is increased to around 90. After fault
clearing, the rectifier starts to conduct and the cable voltage is built up. After a while the
inverter is able to commutate and current in the inverter starts to flow. Ignoring the
VDCOL, the same current order is given to the rectifier and the inverter. Because the CCA
in both ends have almost the same parameter values, the changes in the firing angles are of
about the same speed. The current margin is subtracted from the current order in the
inverter CCA while the inverter firing angle is adjusted to give the current order minus the
current margin and the rectifier firing angle is adjusted to give the current order. In this
way, only the current margin is used to charge the cable and, therefore, the voltage is built
up very slowly. Basically, it is the inverter that controls the current.
The VDCOL helps to speed up the recovery of the dc system. Because the magnitude of
the dc voltage in the rectifier is larger than in the inverter, the current order given to the
rectifier is larger than the current order to the inverter. Thus, a current larger than the cur-
rent margin is used to charge the cable.
Figure 18-47. Transient Controller Inverter
The increase of the current order stops when the dc voltage has reached a reference (about
70%). When both rectifier and inverter have reached this limit, the cable is charged from
the current margin only and, therefore, the voltage is built up slower than before.
18-83
The control function inverter gamma 0 start improves the recovery of the dc system con-
siderably. When this function is activated the firing angle to the inverter is forced to a
higher value, which has two effects: 1) it delays the start-up of the inverter current, and 2)
it increases the inverter dc voltage so the inverter current is reduced. In this way, a larger
proportion of the rectifier current is used to build up the cable voltage.
In the control system,  from the CCA is forced to a high value after removal of the fault
in the inverter ac system. In the model this is represented by adding an additional current
to the optional input in the CCA block diagram. This function is active when the dc volt-
age in the inverter has been less than a reference (VOLT_EN) for a specified time
(TIME_EN) and the ac voltage has exceeded a specified value (GAMST_LIM). The func-
tion is disabled when the dc voltage is greater than a value (VOLT_DIS for a specified
time (TIME_DIS). The model of the inverter gamma 0 start function is illustrated in
Figure 18-48.
18-84
18.11.11.4 Inverter Voltage Control

For long cables, the cable capacitance prevents the rectifier to immediately respond to
what happens in the inverter. Therefore, the risk of a commutation failure in the inverter
end is larger than when the cable capacitance is smaller. In such cases, the problem could
be reduced by adding a voltage control to the inverter control system. Also, the nominal
extinction angle nom should be increased giving a larger margin. One way to implement a
voltage controller is illustrated in Figure 18-49.
The output DALFA_VC from this controller should be added to the upper limit
ALPHA_MAX_ORDER_I for the firing angle for the CCA allowing the inverter to
increase the dc voltage and in this way counteracting for the current increase.
Figure 18-48. Inverter Gamma Start
18-85
CDCAB1 Example Case Program Application Guide: Volume II
Figure 18-49. Voltage Control for Long Cables
18.12 CDCAB1 Example Case
18.12.1 Power Flow Data (RAWD File) for CDCAB1 Example Case
0 100
hvdc test case
nesa/tud/abb dynamic model
1 ‘dummy2’ 400. 1 0. 0. 1 1 1.025 0.
2 ‘dummy2’ 400. 1 0. 0. 1 1 1.025 0.
3 ‘rec_gen’ 400. 3 0. 0. 1 1 1.025 0.
4 ‘inv_gen’ 400. 3 0. 0. 1 1 1.025 0.
5 ‘load5’ 400. 1 0. 0. 1 1 1.025 0.
6 ‘load6’ 400. 1 0. 0. 1 1 1.025 0.
0
5 1 1 1 1 600. 300. 0. 0. 0. 0.
6 1 1 1 1 600. 300. 0. 0. 0. 0.
0
3 1 1202.664 219.01 9999. -9999. 1.025 0 3000. .08625 .9957 0. 0. 1. 1 100. 9999.,-9999.
4 1 7.389 230.63 9999. -9999. 1.025 0 3000. .08625 .9957 0. 0. 1. 1 100. 9999.,-9999.
0
1 3 1 0. .0001 0. 0. 0. 0.,,, 0. 0. 0. 0. 1 0.
1 -5 1 0. .0001 0. 0. 0. 0.,,, 0. 0. 0. 0. 1 0.
2 4 1 0. .0001 0. 0. 0. 0.,,, 0. 0. 0. 0. 1 0.
2 -6 1 0. .0001 0. 0. 0. 0.,,, 0. 0. 0. 0. 1 0.
0
0
0
1 2 2.1 1500. 400. 0. 2.1 0.1 R 0.
1 2 17. 13. .296 9.25 400. .41 1.0325 1.1625 .8625 .00125 0 0 0 1
2 2 19. 15. .296 9.125 400. .41 1.0400 1.1625 .8625 .00125 0 0 0 1
0
1 0 1. 1. 0 326. 1 326.
2 0 1. 1. 0 326. 1 326.
0
0
0
0
0
0
18.12.2 Dynamics Data (DYRE File) for CDCAB1 Example Case

3 ’GENROU’ 1 5.0000 0.60000E-01 0.20000 0.60000E-01
3.0000 0.0000 1.6000 1.5500 0.70000
0.85000 0.35000 0.20000 0.90000E-01 0.38000 /
3 ’SEXS’ 1 0.10000 10.000 100.00 0.10000
0.0000 4.0000 /
4 ‘GENROU’ 1 5.0000 0.60000E-01 0.20000 0.60000E-01
18-86
Program Application Guide: Volume II CDCAB1 Example Case
3.0000 0.0000 1.6000 1.5500 0.70000

0.85000 0.35000 0.20000 0.90000E-01 0.38000 /
4 ‘SEXS’ 1 0.10000 10.000 100.00 0.10000
0.0000 4.0000 /
1 ‘CDCAB1’ 0 0 0 0 0 0 0 0 0 0 0 0 0
0.10000E-03 0.0000 0.0000 0.20000E-01 10.000
0.40000 1500.0 150.00 0.20000E-02 0.10000E-02
0.50000E-01 200.00 21.400 0.20000E-02 0.10000E-02
0.50000E-0 1 200.00 21.400 113.00 27.000
1650.0 1.0000 0.10000 0.28000E+06 0.10000E-02
0.25000E-01 280.00 50.00 150.00 3000.0
0.10000E-02 0.65000E-01 280.00 450.00 150.00
3000.0 0.10000E-01 164.00 5.0000 15.000
1.0000 0.60000 0.38350E-01 1.5330 180.00
-180.00 5.0000 0.10000E-01 164.00 5.0000
15.000 1.0000 0.60000 0.38350E-01 1.5330
180.00 -180.00 105.00 0.10000E-03 0.12000E-01
0.10000 0.10000E-03 1.1500 0.70000 -0.50000E-01
105.00 15.000 0.10000E-03 0.12000E-01 0.10000
0.10000E-03 1.1500 0.70000 -0.50000E-01 105.00
15.000 1000.0 180.00 0.10000E-02 1000.0
105.00 2.0000 5.000 50.000 70.000
100.00 110.00 140.00 6000.0 6000.0
5000.0 1000.0 500.00 -3000.0 -3000.0
-7000.0 -2500.0 .50000E-01 0.50000E-01 10.000
0.20000E-02 0.14400E-02 1.0000 -3.0000 25.000
0.50000E-01 0.70000E-01 0.60000 187.50 40.000
0.10000E-01 0.80000E-01 0.40000 0.50000 0.50000
2625.0 0.50000E-01 0.0000 0.16000E+06 0.20000E+06
0.0000 200.00 0.50000 50.000 -50.000
0.0000 200.00 0.50000 50.000 -50.000
0.70000E-01 1.0000 250.00 50.000 -50.000
0.70000E-01 1.0000 250.00 50.000 -50.000
51.500 0.0000 0.0000 0.0000 0.0000
48.500 48.700 49.500 49.800 0.0000
51.500 0.0000 0.0000 0.0000 0.0000
48.500 48.700 49.500 49.800 0.0000
1.0000 0.0000 0.0000 0.0000 0.0000
0.10000 1.0000 0.10000 0.10000 0.0000
1.00000 0.0000 0.0000 0.0000 0.0000
0.10000 1.0000 0.10000 0.10000 0.0000
100.00 0.0000 0.0000 0.0000 0.0000
50.000 50.000 50.000 50.000 0.0000
100.00 0.0000 0.0000 0.0000 0.0000
50.000 55.000 50.000 50.000 0.0000
0.50000 0.80000 12.000 -12.000 /
18.12.3 DOCU Report

PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E Tue, Oct 14 1997 15:09
HVDC TEST CASE
NESA/TUD/ABB DYNAMIC MODEL
REPORT FOR CONEC MODELS AT ALL BUSES
*** CALL CDCAB1( 1, 1, 41, 17, 1) ***
DC# C O N S S T A T E S V A R S I C O N S
1 41-244 17-50 1-48 1-13
X--- RECTIFIER ---X X--- INVERTER ----X MDC RDC RCOMP SETVAL VSCHED VCMODE DELTI
1 DUMMY2 400 2 DUMMY2 400 2 2.10 2.10 1500.0 400.0 0.0 0.1000
CUR_DIR FRC_R FRC_I PMOD_R PMOD_I EC_R EC_I VC_I PSC_R PSC_I
0 0 0 0 0 0 0 0 0 0
DELTMAX MANBYP_R MANBYP_I T_BYP_MIN GAMMACF VAC_NO_CF IdN

0.0001 0.0000 0.0000 0.0200 10.0000 0.4000 1500.0000
CURMARG TIDC_R TVDC_R RS_R LS_R L_R TIDC_I

150.0000 0.0020 0.0010 0.0500 200.0000 21.4000 0.0020
TVDC_I RS_I LS_I L_I CC RC IOMAX_MSTR

0.0010 0.0500 200.0000 21.4000 113.0000 27.0000 1650.0000
18-87
TMSTR_HIGH TMSTR_LOW UMSTRLIM T_DOWN_R T_UP_R Udbr_R(kV) IOMAXM_R
1.0000 0.1000 280000.0000 0.0010 0.0250 280.0000 450.0000
IOMIN_R IMAX_R T_DOWN_I T_UP_I Udbr_I(kV) IOMAXM_I IOMIN_I

150.0000 3000.0000 0.0010 0.0650 280.0000 450.0000 150.0000
IMAX_I T_IOF_R A_MAX_R A_MIN_R A_NOM_R LIN_MAX_R LIN_MIN_R
3000.0000 0.0100 164.0000 5.0000 15.0000 1.0000 0.6000
KP_R KI_TI_R APROP_MAX_R APROP_MIN_R AORDR_MIN_R T_IOF_I A_MAX_I

0.0384 1.5330 180.0000 -180.0000 5.0000 0.0100 164.0000
A_MIN_I A_NOM_I LIN_MAX_I LIN_MIN_I KP_I KI_TI_I APROP_MAX_I

5.0000 15.0000 1.0000 0.6000 0.0384 1.5330 180.0000
APROP_MIN_I AORDR_MIN_I K1_R T1_R T2_R K1_MAX_R U_NORM_MX_R
-180.0000 105.0000 0.0001 0.0120 0.1000 0.0001 1.1500
U_NORM_MN_R S1_MIN_R MIN_AMAX_R GAMMAMIN_R K1_I T1_I T2_I

0.7000 -0.0500 105.0000 15.0000 0.0001 0.0120 0.1000
K1_MAX_I U_NORM_MX_I U_NORM_MN_I S1_MIN_I MIN_AMAX_I GAMMAMIN_I T_CFC_R

0.0001 1.1500 0.7000 -0.0500 105.0000 15.0000 1000.0000
ALFA_MAX_R TALFA_MAX_I T_CFC_I ALFA_MIN_I DELTGAM ALFA1 ALFA2
180.0000 0.0010 1000.0000 105.0000 2.0000 35.0000 50.0000
ALFA3 ALFA4 ALFA5 ALFA6 DALFA_MAX1 DALFA_MAX2 DALFA_MAX3
70.0000 100.0000 110.0000 140.0000 6000.0000 6000.0000 5000.0000
DALFA_MAX4 DALFA_MAX5 DALFA_MIN1 DALFA_MIN2 DALFA_MIN3 DALFA_MIN4 T_PSC_R

1000.0000 500.0000 -3000.0000 -3000.0000 -7000.0000 -2500.0000 0.0500
T_PSC_I K_CC T_CC1 T_CC2 DA_CC_LIMU DA_CC_LIML MX_AOR_MN_R
0.0500 10.0000 0.0020 0.0014 1.0000 -3.0000 25.0000
TEN_TRCONR TDIS_TRCONR UAC_TRCONR D_AORD_MN_R TRCONI_DGAM TRCONI_TUP TRCONI_TDWN
0.0500 0.0700 0.6000 187.5000 40.0000 0.0100 0.0800
TRCONI_V_A TRCONI_V_D GAMST_LIM GAMST_IORD TIME_EN TIME_DIS VOLT_EN
0.4000 0.5000 0.5000 2625.0000 0.0500 0.0000 160000.0000
VOLT_DIS DB_R K_FREQ_R T_FREQ_R UL_FREQ_R LL_FREQ_R DB_I
200000.0000 0.0000 200.0000 0.5000 50.0000 -50.0000 0.0000
K_FREQ_I T_FREQ_I UL_FREQ_I LL_FREQ_I T1DAMP_R T2DAMP_R KDAMP_R

200.0000 0.5000 50.0000 -50.0000 0.0700 1.0000 250.0000
ULDAMP_R LLDAMP_R T1DAMP_I T2DAMP_I KDAMP_I ULDAMP_I LLDAMP_I

50.0000 -50.0000 0.0700 1.0000 250.0000 50.0000 -50.0000
EC_R_FL_H1 EC_R_FL_H2 EC_R_FL_H3 EC_R_FL_H4 EC_R_FL_H5 EC_R_FL_L1 EC_R_FL_L2

51.5000 0.0000 0.0000 0.0000 0.0000 48.5000 48.7000
EC_R_FL_L3 EC_R_FL_L4 EC_R_FL_L5 EC_I_FL_H1 EC_I_FL_H2 EC_I_FL_H3 EC_I_FL_H4

49.5000 49.8000 0.0000 51.5000 0.0000 0.0000 0.0000
EC_I_FL_H5 EC_I_FL_L1 EC_I_FL_L2 EC_I_FL_L3 EC_I_FL_L4 EC_I_FL_L5 EC_R_TIME1

0.0000 48.5000 48.7000 49.5000 49.8000 0.0000 1.0000
EC_R_TIME2 EC_R_TIME3 EC_R_TIME4 EC_R_TIME5 EC_R_TIME6 EC_R_TIME7 EC_R_TIME8

0.0000 0.0000 0.0000 0.0000 0.1000 1.0000 0.1000
EC_R_TIME9 EC_R_TIME10 EC_I_TIME1 EC_I_TIME2 EC_I_TIME3 EC_I_TIME4 EC_I_TIME5

0.1000 0.0000 1.0000 0.0000 0.0000 0.0000 0.0000
EC_I_TIME6 EC_I_TIME7 EC_I_TIME8 EC_I_TIME9 EC_I_TIME10 EC_R_PS1 EC_R_PS2

0.1000 1.0000 0.1000 0.1000 0.0000 100.0000 0.0000
EC_R_PS3 EC_R_PS4 EC_R_PS5 EC_R_PS6 EC_R_PS7 EC_R_PS8 EC_R_PS9

0.0000 0.0000 0.0000 50.0000 50.0000 50.0000 50.0000
EC_R_PS10 EC_I_PS1 EC_I_PS2 EC_I_PS3 EC_I_PS4 EC_I_PS5 EC_I_PS6

0.0000 100.0000 0.0000 0.0000 0.0000 0.0000 50.0000
EC_I_PS7 EC_I_PS8 EC_I_PS9 EC_I_PS10 KP_VC TI_VC DALF_MX_VC

55.0000 50.0000 50.0000 0.0000 0.5000 0.8000 12.0000
18-88
Program Application Guide: Volume II CDCAB1 Example Case
DALF_MN_VC DCCUR KVDCR KVDCI

-12.0000 1500.0000 400.0000 396.8500
ALF/GAM MIN MAX PAC QAC NB EBASE RC XC TR TAP

R: 13.11 13.00 17.00 602.7 261.5 2 400.0 0.296 9.250 0.4100 1.0325
I: 15.24 15.00 19.00 -592.6 273.1 2 400.0 0.296 9.125 0.4100 1.0400
18-89
18-90
Chapter 19
Transmission Line and
Other Relay Models
Chapter 19 - Transmission Line and Other Relay Models
19.1 Principles
The model library includes a group of detailed branch relay models. These models, as
described in this section, are intended to represent the general principles of relays that are
sensitive primarily to apparent impedance. The models are not intended to exactly repre-
sent any particular product of any relay manufacturer. Rather, they model the effects of
actual relays given properly specified data and proper interpretation by the engineer. The
detailed relay models, which are demanding in terms of data, are constructed from a set of
basic elements described below. The relay models themselves are best understood on the
basis of a good understanding of these elements.
19.2 Relay Elements
19.2.1 Impedance Detection Unit

The basic element of these relay models is the impedance detection unit. This unit accepts
the positive-sequence current flowing into the terminals of a transmission line and the pos-
itive-sequence voltage at the same terminals. It derives an apparent impedance by dividing
the voltage by the current. This apparent impedance is compared to a relay characteristic
and, if it is within the zone of the R-X (Figure 19-1) plane defined by the characteristic, a
zone flag is set.
All relays are reset by activity STRT, and this resets all zone flags. The flag for a zone is
set when the apparent impedance enters the zone and reset when it leaves the zone. The
relay zones are delineated by the logical product of the basic relay zone and the restrictive
zone defined by any blinders that are in effect. The picking up and resetting of the zone
flag for one relay zone is illustrated in Figure 19-1. The apparent resistance and reactance
detected by the unit are available in VARs for placement in output channels so that the
user can plot loci of apparent impedance.
19.2.2 Zone Timer

The zone timer has a clock, two inputs, and a single output. The output is zero until the
clock times out. The first input to the zone timer is the zone flag, the second is a reset sig-
19-1
Transmission Line and Other Relay Models PSS®E 33.4
Relay Elements Program Application Guide: Volume II
nal. The reset input instantaneously initializes the clock but does not start it. The start
input starts the clock; when the clock times out the output has a value of 1. After the clock
has been started, additional start signals are ignored unless a reset signal has been given.
The timer delay setting may be zero in which case the output is enabled instantaneously.
The action of the timer is depicted in Figure 19-2. If the delay time specified is not a multi-
ple of the time step, action will take place at the nearest time.
19-2
PSS®E 33.4 Transmission Line and Other Relay Models
Program Application Guide: Volume II Relay Elements
X Inactive Area
t
t + t Locus of Apparent
Impedance
t + 2t
t1
t1 + t
Active Area
Zone Flag Set at t + 2t

Zone Boundary Zone Flag Reset at t1 + t
a. Without a Blinder
Locus of Apparent
t + t t Impedance
t + 2t
t + 3t
t1 + 2t
t1
t1 + t
Active Zone Flag Set at t + 2t
Area Zone Flag Reset at t1 + t
Blinder
b. With a Blinder
Figure 19-1. Setting and Resetting of Zone Flags
19-3
Relay Elements Program Application Guide: Volume II
Delay
Start Signal
Start
Reset Signal
Timer Output
Reset Output
Simulation Step Size
a. Times Out Before Reset
Start Signal
Delay
Reset Signal
Output
b. Resets Before Times Out
Figure 19-2. Operation of Zone Timer
19-4
Program Application Guide: Volume II Relay Elements
19.2.3 Circuit Breaker Timer

The inherent opening delay of each circuit breaker to which a relay sends a trip signal is
modeled by a circuit breaker timer. The action of these timers is similar to that of the zone
timer.
19.2.4 Circuit Breaker

The circuit breaker model (Figure 19-3) has three signals that combine to provide the start
input to the circuit breaker timer: a trip signal, normally the output of the zone timer; a
permissive flag, which must be set for the breaker to respond to the trip signal; and a force
trip signal that always produces a trip when set. The circuit breaker opens (i.e., current
goes to zero), when the circuit breaker timer times out. No action is taken if the line has
already been tripped (or reclosed) when the trip (or reclosure) timer times out.
Force Trip
Trip Self-Trip
Breaker TRIP
Timer
Permissive
Flag AND OR
Reset Self
Reclosure RECLOSE
Breaker
Timer
Figure 19-3. Logic for Breaker
19.2.5 Reclosure Timer

The reclosure timer is started when the breaker timer times out (i.e., when the line is
tripped). When the reclosure timer times out, a reclose signal is sent to the breaker.
19.2.6 Minimum Current Detector

Several relays incorporate a minimum current detector. This unit sets a low-current flag if
the absolute value of the current input to the impedance detection unit is below a thresh-
old. This flag blocks action (i.e., current must be above threshold to permit action) in sev-
eral places in the relays. (See overall relay logic diagrams.) This threshold current should
be entered as per unit on system base.
19-5
Bases for Relay Data Program Application Guide: Volume II
19.2.7 Supervisory Input - Permissive Flag

The relay models have supervisory inputs. These take the form of permissive flags, one
affecting tripping of the relay’s own circuit breaker, and one that has a common effect on
all of the relay’s transfer trip actions. The permissive flag may have the value of 0 (trip-
ping is blocked), 1 (tripping is permitted), or –1 (tripping is forced immediately, regardless
of condition of the supervised relay.)
The permissive flags are initialized by activity DYRE to a value of 1 and are not affected
by STRT. Their values may be changed manually via activity ALTR when the relay to
which they apply is not supervised by another relay such as CIROS1 or SLNOS1. Use of
the supervisory capability of CIROS1 or SLNOS1 automates the setting of the permissive
flags and hence they should not be changed by ALTR when they are used as communica-
tion between relay models. Supervision is achieved by specifying the ICON number of the
supervised relay’s permissive flag in the supervisory ICON number. This can be done via
the relay editors or activity ALTR.
19.2.8 Blinders
The relays are provided with blinders that form a straight-line boundary to restrict the
extent of their otherwise circular zones. The characteristic of each blinder is a straight line
separating the R-X plane into a trip-permitted zone on one side and a trip-prohibited zone
on the other, as shown in Figure 19-4 (top). The blinder’s effective zone is logically and-ed
with all basic zones of the relay in which the blinder is incorporated to produce a restricted
effective zone as shown in Figure 19-4 (bottom).
19.2.9 Relay Time-Gating Unit

PSS®E applies a time-gating logic to some of the detailed relay models. This logic causes
the relay model to be executed every n time steps, instead of every time step, which is the
normal practice. Use of the time-gating logic with a value of n = 5, for example, would be
appropriate in special situations, i.e., a particularly small time step is used to accommo-
date a short time constant in a model like GENDCO or SHAF25. The time-gating logic is
overridden to force execution of all relay models at all pauses or internal switching opera-
tions within the execution of activity RUN (Figure 19-5).
As indicated by Figure 19-1a, zone timers are started and stopped when the apparent
impedance locus enters and leaves the effective trip zone and not when the locus crosses
the basic circular zone boundary. That is, the blinder characteristic line becomes a part of
the zone boundary and the excluded arc of the circle disappears.
19.3 Bases for Relay Data

All impedance relay models in Chapter 11 of the PSS®E Model Library (DISTR1,
CIROS1, SLNOS1, and RXR1) use common bases for data:
1. All angles are stated in degrees.
2. All impedance intercepts, zone center distances, and zone diameters, expressed with
respect to the R-X plane, are stated in terms of impedance per-unit on system base.
19-6
Program Application Guide: Volume II Bases for Relay Data
3. All time delays are stated in cycles.
4. All currents, voltages, and detected impedances are stated per unit with respect to
system MVA base.
19-7
Bases for Relay Data Program Application Guide: Volume II
X X
Type 2 Type 2
Trip Trip
Trip Trip-Prohibited Prohibited Permitted
Permitted Angle
R R
Intercept Intercept Angle
X Trip X
Prohibited
Type 1 Type 1
Trip Angle
Permitted
Intercept
R R
Intercept
Angle Trip
Permitted
Trip
Prohibited
a. Blinder Types
Type 2 Blinder
Basic Circular Zone

Characteristic
Shaded Area Indicates

Effective Trip Zone
b. Type 2 Blinder Application
Figure 19-4. Straight-Line Relay Blinders
19-8
Program Application Guide: Volume II Model DISTR1
Calculation Applied After

Major Disturbances
Calculation Applied
Prior to Major
Disturbance 1
2
3
1
Zone
Flag Set
t t – t
n=3
Relay Model is Called
Zone Boundary Every Three Time Steps
Figure 19-5. Operation Time Gating Logic
19.4 Model DISTR1
19.4.1 General Characteristics

The features of the distance relay model (DISTR1) follow. Its overall logic diagram is
• Mho, impedance, or reactance characteristic with up to three independent (separately
specified diameter and center offset) circles for mho or impedance units.
• Trip of up to three remote lines (transfer trips) as well as monitored line (self trip).
• Single attempt reclosure for self transfer trip for zone 1 faults.
• Supervisory signal input to prevent tripping or to force immediate tripping.
• Up to two straight-line blinders.
• Minimum pickup current logic.
• Model may be used in a monitoring only mode.
19-9
Model DISTR1 Program Application Guide: Volume II
Low-Current Flag
Zone 1 Flag • Start
Zone 1
• Timer Self-Trip
Reset Permissive
Flag Force
• Trip Signal
•
Low-Current Flag
Zone 1 Trip Signal
• Timer
Zone Reset
Detector •
Low-Current Flag
Zone 1
• Timer
Reset
Low-Current Flag
Trip Signal
Self-Trip Self Trip
Reset Breaker Timer
Self-Trip Self Reclose

Reset Reclosure Timer
Transfer Trip Transfer Trips

Reset Breaker Timer
Transfer Trip Transfer Reclosures

• Reset Reclosure
Transfer Trip Timer
Permissive Flag
Figure 19-6. Overall Logic Diagram
19-10
19.4.2 Specific Characteristics

The following should be noted with respect to the DISTR1 model data sheet:
• A trip action caused by DISTR1 opens the circuit breakers at both ends of a line
simultaneously.
• Relay is at the from bus looking toward the to bus.
• If invalid line data is specified, the appropriate message is printed out during
initialization.
• DISTR1 requires certain ICONs and VARs for local storage. Do not reuse these
storage locations in other models or user-written logic in CONEC and CONET
subroutines.
• The first three VARs used by DISTR1 are the per-unit resistance, reactance, and cur-
rent. These values may be put into output channels.
• All angles are specified in degrees from horizontal; anti-clockwise direction is positive.
• All timer settings are specified in cycles and includes all delays.
• Resetting of timers is assumed instantaneous.
• A from bus of 0 for a transfer trip bypasses the logic for that transfer trip. The permissive
flags should be set to 1 in user input if tripping is desired.
• The permissive flags are not initialized by the program. The self-trip permissive flag of
a supervised relay may be changed by a supervising relay during a simulation run.
19.4.3 Data
ICON(I) The relay type is specified by this ICON. The zone shape associ-
ated with each type is shown on the data sheet.
ICON(I+1) Specifying 0 disables all tripping actions of relay model, but
allows detected impedance to be recorded as if model was active.
ICON(I+2,3,4) Specify a line to be tripped by transfer-trip action from this relay.
Setting these ICONs to zero deactivates transfer tripping.
ICON(I+5,6,7) See above.
ICON(I+8,9,10) See above.
ICON(I+11) Specify permissive flag supervising the relay’s tripping of its
(Note: ICON(I+11) own circuit breaker. Its value is initialized to 1 (permit tripping)
through ICON(I+28) by activity DYRE and is not affected by activity STRT. Supervis-
are not included in the ing relays such as CIROS1, or the user activity ALTR, may
data record for change the value of this ICON during a simulation run. Supervis-
DYRE.) ing action by another relay is achieved by setting the supervisory
ICON number of the supervising relay to (I+11). (See
Sections 19.6 and 19.7.)
19-11
ICON(I+12) Specify the permissive flag supervising all transfer tripping

actions of the relay. Specification and/or control is the same as
for ICON(I+11).
CON(J) Set to the relay’s basic operating time: the minimum time for the
relay to see the apparent impedance in zone 1 in order to issue a
trip signal. May be set to 0 to represent instantaneous tripping
action.
CON(J+1,2,3) Specify the size and position of the relays zone 1. Zone 1 must be
the innermost of three zones specified.
CON(J+4) Specify sum of the time delay setting of zone 2 and the relay’s
basic operating time.
CON(J+5,6,7) Specify the size and position of zone 2.
CON(J+8) Specify sum of the time delay setting of zone 3 and the relay’s
basic operating time.
CON(J+9,10,11) Specify the size and position of zone 3.
CON(J+12) Specify the angle of directional element in an impedance dis-
tance relay where three circular zones are centered at the origin.
The angle is defined on the DISTR1 data sheet.
CON(J+13) Specify current threshold of the minimum current detector.
DISTR1 will neither start its zone timers nor issue trip signals if
the current input to the apparent impedance detector is below this
value.
CON(J+14) Specify the operating time of the relay’s own circuit breaker: the
time elapsed between receipt of a trip signal at the circuit breaker
and interruption of current.
CON(J+15) Specify the reclosure delay for the relays own circuit breaker.
The reclosing delay timer is started upon interruption of current.
Reclosure occurs (i.e., contacts touch) when this timer times out.
CON(J+16,17) Specify circuit breakers controlled by relay’s transfer trip out-
puts.
CON(J+18,19,20) Specify first of two blinders. (See Section 19.2.7.) Set to zero if
blinder not active.
CON(J+21,22,23) Specify second of two blinders. Set to zero if blinder not active.
19.4.4 Example Application - DISTR1

The application of DISTR1 is illustrated by example. The example considers the sample
system used in Section 15.4 to illustrate the application of the CSVGN1 static var device
model. Figure 19-7 shows a filled-in data sheet for an initial trial application of DISTR1.
The example checks the effect of mho distance relays on the three 500-kV circuits from
bus 1 to 2. The concern is with relays located at bus 1 looking toward bus 2 when a fault
19-12
occurs on circuit 1 at the bus 1 (sending) Figure 19-10 shows the locus of apparent
end. The data shown in Figure 19-7 and 19-8 impedance, Z, seen by DISTR1 on circuit 2
is a first trial at relay settings. in relation to that relay’s impedance zones.
Figure 19-9 shows the dialog and relay mes- Figure 19-11 shows the simulation result in
sage output from a first simulation run in more conventional time-domain terms and
which a fault is simulated by the applica- emphasizes the collapse of the system after
tion of a 100 per-unit reactive admittance the tripping of the second transmission cir-
(i.e., 0.01 per-unit fault impedance) at the cuit.
line reactor location of circuit 1. The fault
must be applied by changing the value of
the branch shunt admittance, rather than by
changing the bus shunt admittance at bus
1, because the bus shunt would be behind
the relays and would result in a contribu-
tion to line current in the wrong direction.
The second page of Figure 19-9 shows the
use of activity RUN where DISTR1, rather
than the user, handles the clearance of the
fault. Because DISTR1 will handle fault
clearance in this run, the initial run dura-
tion can be any convenient length, rather
than the duration of the fault. An initial run
to 0.2 sec, with the fault applied at t = 0.1,
permits plotting and printing every time
step while faults and switchings are caus-
ing rapid changes. (Output channel print-
ing is deleted from Figure 19-9 after t =
0.133 to conserve space.) From t = 0.2, the
run is extended to t = 5.0 with more eco-
nomical printing and plotting intervals.
The simulation output shows messages
from the two DISTR1 models. These mes-
sages indicate that the relay on circuit 1
sees the fault in its zone 1 and that this
results in clearance by the 1.5 cycle
breaker after 0.033 sec, or 2 cycles, as
expected, given the 0.5 cycle zone 1 relay
time. The messages from DISTR1 also
indicate that the relay on circuit 2 sees the
fault in its third zone from t = 0.117 sec-
onds. This zone 3 timer times out as the
message indicates at t = 0.25 seconds,
which initiates the breaker tripping 1.5
cycles later. The voltage never recovers
and appears to still be dropping.
19-13
DISTR1
mho, Impedance, or Reactance Distance Relay
Relay is located from bus #______ IBUS.

to bus #______ JBUS.
circuit #______ ID.
relay slot (1 or 2) #______ RS.
ICONs Value Description VARs Description

I 1 Type 1 - mho distance L Apparent R
Type 2 - impedance distance
L+1 Apparent X
Type 3 - reactance distance
L+2 Current
I+1 1 0 - Monitor
1 - Monitor and operate L+3
.
I+2 0 From bus VARs required for internal
.
number program logic
First .
I+3 0 To bus number transfer trip L+9
I+4 0 Circuit ID
I+5 0 From bus X
number
Second
I+6 0 To bus number transfer trip
I+7 0 Circuit ID Diameter
I+8 0 From bus Centerline
number Angle
Third
I+9 0 To bus number transfer trip
Center
I+10 0 Circuit ID Distance
I+11 X Permissive flag for self trip* R
I+12 X Permissive flag for transfer trip†
I+13 X
. Zone Reach
ICONs required for
.
.
internal program logic
I+28
*Set to 1 and -1 by supervisory relay to

block trip and force trip, respectively.
†Set to 1 by supervisory relay to block
trip.
IBUS, ’DISTR1’, JBUS, ID, RS, first 11 ICONs, CON list /
Figure 19-7. Application of DISTR1 to Circuits 1 and 2

Between Buses 1 and 2 of Sample System (Sheet 1 of 3)
19-14
CONs Value Description

J .5 Zone 1 operating time (cycles)
J+1 .065 Zone 1 reach (diameter or reac-
tance) (pu)
J+2 86 Zone 1 centerline angle in
degrees (0 for reactance relay)
J+3 .03 Zone 1 center distance (0 for
reactance relay)
J+4 .5 Zone 2 pickup time (cycles)
J+5 .085 Zone 2 reach (diameter or reac-
tance) (pu)
J+6 86 Zone 2 centerline angle (0 for
reactance relay)
J+7 .04 Zone 2 center distance (0 for
reactance relay)
J+8 8 Zone 3 pickup time (cycles)
J+9 .11 Zone 3 reach (diameter)
J+10 86 Zone 3 centerline angle (degrees)
J+11 .055 Zone 3 center distance (pu)
J+12 0 Angle of directional unit (only for
impedance relay)
J+13 1.0 Threshold current (pu)
J+14 1.5 Self trip breaker time (cycles)
J+15 1E6 Self trip reclosure time (cycles)
J+16 1.5 Transfer trip breaker time (cycles)
J+17 1E6 Transfer trip reclosure time
(cycles)
J+18 0 1st blinder type (±1 or ±2)
J+19 0 1st blinder intercept (pu)
J+20 0 1st blinder rotation (degrees)
J+21 0 2nd blinder type (±1 or ±2)
J+22 0 2nd blinder intercept (pu)
J+23 0 2nd blinder rotation (degrees)
19-15
Zone 3
Zone 2
Zone 1
mho Distance
Zone 3
Reach
Zone 3
Center
Distance
Zone 3 Zone 2
Centerline Zone 1 Reach
Angle Reach
Reactance Distance
Zone 3
Zone 2
Zone 1
R
Directional
Element
Impedance Distance
Figure 0-0 (Cont). Application of DISTR1 to Circuits 1 and 2 Between Buses 1 and 2
of Sample System (Sheet 2 of 3)
19-16
Inactive Area X
2
1
X R
Intercept
R
All blinder rotations

measured from
horizontal
X X
-1 -2
R R
Blinder Types
Directional
Element
Angle
R
mho, Impedance, or Reactance Distance Relay
Figure 0-0 (Cont). Application of DISTR1 to Circuits 1 and 2 Between Buses 1 and 2
of Sample System (Sheet 3 of 3)
19-17
*** CALL DISTR( 4, 59, 4) ***
ICON'S C O N ' S V A R ' S

4- 39 59- 82 4- 13
REF # TYPE MON/TRIP FLAG RES MONITORED LINE TRANSFER TRIP
1 1 1 1 1 2 '1 ' 0 0 '0 ' No transfer trip
TRANSFER TRIP TRANSFER TRIP SELF FLAG TRANS FLAG
0 0 '0 ' 0 0 '0 ' 1 1 Supervisory flags
initialized by STRT
CY1 REACH1 ANGLE1 DIST1 CY2 REACH2 ANGLE2 DIST2
0.500 0.065 86.000 0.030 5.000 0.085 86.000 0.040
CY3 REACH3 ANGLE3 DIST3 DIR ANG THRESH SELF TB SELF TR
8.000 0.110 86.000 0.055 0.000 1.000 1.500********* Large number
prevents reclosure
TRAN TB TRAN TR BL1 TYP BL1 INT BL1 ROT BL2 TYP BL2 INT BL2 ROT
1.500********* 0.000 0.000 0.000 0.000 0.000 0.000
*** CALL DISTR( 40, 83, 14) ***

ICON'S C O N ' S V A R ' S
40- 75 83- 106 14- 23
REF # TYPE MON/TRIP FLAG RES MONITORED LINE TRANSFER TRIP
2 1 1 1 1 2 '2 ' 0 0 '0 '
TRANSFER TRIP TRANSFER TRIP SELF FLAG TRANS FLAG

0 0 '0 ' 0 0 '0 ' 1 1

0.500 0.065 86.000 0.030 5.000 0.085 86.000 0.040

8.000 0.110 86.000 0.055 0.000 1.000 1.500*********
1.500********* 0.000 0.000 0.000 0.000 0.000 0.000
Figure 19-8. DOCU Output Corresponding to Figure 19-7
19-18
Executing activity STRT
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E TUE, NOV 30 1999 12:02


1 HYDRO 500 1 1.0600 2.4913 2690.23 1029.58 0.9339 44.43 0.8729 0.4001
3 LOAD 230 1 1.0252 2.4828 450.00 200.00 0.9138 -5.58 0.9334 0.4451
4 TERT1 33.0 1 1.0000 0.0000 0.00 461.58 0.0000 0.00 0.6976-0.6046
5 TERT2 33.0 1 1.0000 0.0000 0.00 461.58 0.0000 0.00 0.6976-0.6046
CHANNEL OUTPUT FILE IS C:\pag\gp.out
-0.0167 0.91666 VOLT 2 [RECEIVE 500.00] 44.435 ANGL 1 [HYDRO 500.00] [1 ]
3 -5.5805 ANGL 3 [LOAD 230.00] [1 ] 26.902 POWR 1 [HYDRO 500.00] [1 ]
5 4.5000 POWR 3 [LOAD 230.00] [1 ] 10.296 VARS 1 [HYDRO 500.00] [1 ]
7 0.10219 R-1-2-1 0.39110E-01 X-1-2-1 R+jX as seen by DISTR1
9 0.10219 R-1-2-2 0.39110E-01 X-1-2-2
for circuit 1
0.1000 0.91666 44.435 -5.5805 26.902 4.5000 10.296
7 0.10219 0.39110E-01 0.10219 0.39110E-01
R+jX as seen by DISTR1
BRANCH DATA FOR CKT 1 FROM 1 [HYDRO 500.00] TO 2 [RECEIVE 500.00]:
STATUS LINE R LINE X CHARGING RATE-A RATE-B RATE-C LENGTH for circuit 2
OLD 1 0.00500 0.07200 5.50000 2000.0 1000.0 2500.0 0.0 CHANGE IT?
LINE SHUNTS: BUS 1 [HYDRO 500.00] BUS 2 [RECEIVE 500.00]
OLD 0.00000 -2.00000 0.00000 0.00000 CHANGE IT?
NEW 0.00000-100.00000 0.00000 0.00000 0+j.01 per unit impedance fault
METERED END IS BUS 1 [HYDRO 500.00]. ENTER 1 TO REVERSE: applied on circuit 1
RELAY DISTR1 # 1 CIRCUIT 1 FROM 1 TO 2 MESSAGES AT TIME = 0.100: DISTR1 on circuit 1

APPARENT IMPEDANCE ENTERED ZONE(S) 1 2 3
ZONE 1 2 3 TIMER(S) STARTED sees the fault
0.1000 0.38679 VOLT 2 [RECEIVE 500.00] 44.435 ANGL 1 [HYDRO 500.00] [1 ]
7 0.89754E-03 R-1-2-1 0.94847E-02 X-1-2-1
9 0.70637E-01 R-1-2-2 0.46409E-01 X-1-2-2
RELAY DISTR1 # 1 CIRCUIT 1 FROM 1 TO 2 MESSAGES AT TIME = 0.108: End of ½ cycle zone 1 operating time of
ZONE 1 TIMER(S) TIMED OUT
SELF TRIP BREAKER TIMER STARTED relay; circuit breaker receives trip signal
0.1083 0.32616 44.435 -5.5805 5.5109 2.4114 22.557
7 0.87426E-03 0.93942E-02 0.64095E-01 0.49172E-01
RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 0.117: DISTR1 on circuit 2 sees fault on
APPARENT IMPEDANCE ENTERED ZONE(S) 3
ZONE 3 TIMER(S) STARTED
circuit 1 as voltage at bus 1 falls
0.1167 0.27259 44.575 -5.5008 4.3274 1.7994 18.634
7 0.83635E-03 0.93107E-02 0.58145E-01 0.51909E-01
0.1250 0.25894 44.922 -5.2395 4.0101 1.6501 17.597

7 0.82162E-03 0.92991E-02 0.57263E-01 0.52866E-01
Figure 19-9. Dialog for Simulation with DISTR1 on Circuits 1 and 2

Between Buses 1 and 2 (Sheet 1 of 2)
19-19
RELAY DISTR1 # 1 CIRCUIT 1 FROM 1 TO 2 MESSAGES AT TIME = 0.133:

SELF TRIP BREAKER TIMER TIMED OUT Breakers interrupt current in
*** CIRCUIT 1 FROM 1 [HYDRO 500] TO 2 [RECEIVE 500]
TRIPPED AT TIME = 0.133 circuit 1
SELF RECLOSURE TIMER STARTED
0.1333 0.24418 45.401 -4.8232 3.7039 1.4822 16.601

7 0.80582E-03 0.92773E-02 0.55627E-01 0.53776E-01
7 0.10000E+11 R-1-2-1 0.10000E+11 X-1-2-1
9 0.49250E-01 R-1-2-2 0.66303E-01 X-1-2-2
0.1417 0.38844 46.022 -4.2640 8.3022 1.9846 11.326

7 0.10000E+11 0.10000E+11 0.49856E-01 0.68015E-01
0.1500 0.40314 46.759 -3.5779 8.8965 1.9467 12.307

7 0.10000E+11 0.10000E+11 0.50676E-01 0.70104E-01
0.1583 0.40628 47.594 -2.7809 9.0824 1.8917 12.774

7 0.10000E+11 0.10000E+11 0.50732E-01 0.71352E-01
0.1667 0.41013 48.540 -1.8533 9.2767 1.8519 13.219

7 0.10000E+11 0.10000E+11 0.50816E-01 0.72411E-01
7 0.10000E+11 R-1-2-1 0.10000E+11 X-1-2-1
9 0.50858E-01 R-1-2-2 0.73278E-01 X-1-2-2
0.1833 0.41616 50.755 0.39483 9.5724 1.7997 13.908

7 0.10000E+11 0.10000E+11 0.50909E-01 0.73971E-01
0.1917 0.41898 52.021 1.7166 9.6958 1.7868 14.174

7 0.10000E+11 0.10000E+11 0.50977E-01 0.74525E-01
0.2000 0.42177 53.393 3.1707 9.8087 1.7814 14.397

7 0.10000E+11 0.10000E+11 0.51071E-01 0.74965E-01
3 3.1707 ANGL 3 [LOAD 230.00] [1 ] 9.8083 POWR 1 [HYDRO 500.00] [1 ]
7 0.10000E+11 R-1-2-1 0.10000E+11 X-1-2-1
9 0.51071E-01 R-1-2-2 0.74965E-01 X-1-2-2
RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 0.250: Circuit 2 breakers receive
SELF TRIP BREAKER TIMER STARTED trip signal
RELAY DISTR1 # 1 CIRCUIT 2 FROM
SELF TRIP BREAKER TIMER TIMED OUT
1 TO 2 MESSAGES AT TIME = 0.275: Breakers are open on
*** CIRCUIT 2 FROM 1 [HYDRO 500] TO 2 [RECEIVE 500] circuit 2
TRIPPED AT TIME = 0.275
0.4000 0.30029 120.10 75.084 5.2397 2.1318 13.190

7 0.10000E+11 0.10000E+11 0.10000E+11 0.10000E+11
0.6083 0.29719 266.31 216.95 5.3446 1.9859 13.817

7 0.10000E+11 0.10000E+11 0.10000E+11 0.10000E+11
Figure 0-0 (Cont). Dialog for Simulation with DISTR1 on Circuits 1 and 2
Between Buses 1 and 2 (Sheet 2 of 2)
19-20
Figure 19-10. Locus of Apparent Z Seen by DISTR1 on Circuit 2
19-21
PELEC Hydro
PELEC Load
Voltage 2
Circuit 2 Tripped
Figure 19-11. Time Plots of Simulation Result
19-22
The undesirable effect of the third zone of the distance relay on circuit 2 can be corrected
in this example by the application of a blinder. The proper location of the blinder can be
found by rerunning the simulation of Figure 19-9 with no changes except for setting the
zone 3 time delay of the DISTR1 relay on circuit 2 to a large value. The result of this
appears in the relay messages in Figure 19-12 and by the locus of apparent impedance seen
by DISTR1 on circuit 2, shown in Figure 19-13. The messages and the plotted locus show
that the apparent impedance seen by DISTR1 on circuit 2 hovers close to the edge of zone
3 and swings into and out of the zone. The time response shown in Figure 19-14 indicates
that this low value of apparent Z is the result of severely depressed voltage and consequent
high reactive power flow into the load area. The timings in Figure 19-12 indicate that the
zone 3 time delay would have to be about 0.3 sec to avoid trips during the successive
incursions of apparent Z into the third zone. This delay is too long and would prevent the
relay from providing its intended backup protection for other relays.
Tripping due to the swings shown in Figure 19-12 can also, though, be prevented by use of
a blinder to clip off the right-hand side of zone 3. Figures 19-15 through 19-17 show the
DISTR1 data, messages, and simulation output involved when a blinder is placed at the
right-hand side of the standard circular relay zones. The placement of the blinder is shown
in Figure 19-16, and the corresponding data are shown in Figure 19-15. This final simulation
run includes a major load shedding action at t = 3.0 sec, after which system voltages
recover strongly and continued satisfactory operation becomes possible.
19-23

ZONE 1 2 3 TIMER(S) STARTED

SELF TRIP BREAKER TIMER STARTED
RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 0.117: DISTR1 on

ZONE 3 TIMER(S) STARTED circuit 2
sees fault
*** CIRCUIT 1 FROM 1 [HYDRO 500] TO 2 [RECEIVE 500]
TRIPPED AT TIME = 0.133
RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 0.308: Loses sight of it

APPARENT IMPEDANCE OUTSIDE ZONES
ZONE 3 TIMER(S) RESET
RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 1.833: Sees it again

RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 1.917: Loses it

RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 2.550: Sees it








Figure 19-12. Relay Messages for Fault at Bus 1 End of Circuit 1

with Normal Clearing by DISTR1 but with Infinite Time Delay in DISTR1 on Circuit 2
19-24
Figure 19-13. Locus of Apparent Z Seen by DISTR1

on Circuit 2 Infinite Zone 3 Time
19-25
Sending End Power
Receiving End Voltage

Receiving End Power
Figure 19-14. Time Domain Response

Corresponding to Figure 19-12 and 19-13
19-26
** DISTR1 ** RELAY SLOT FROM BUS NAME BSKV TO BUS NAME BSKV CKT ID
1 1 HYDRO 500 2 RECEIVE 500 2
I C O N S C O N S V A R S
32-60 83-106 13-22
DISTANCE TYPE MON/TRIP SETTING

1 (MHO) 1
TRANSFER TRIP 1: FROM BUS TO BUS CKTID

0 0 ’0 ’

0 0 ’0 ’

0 0 ’0 ’
FLAG FOR SELF TRIP FLAG FOR TRANSFER TRIP

0 0

0.500 0.065 86.000 0.030 5.000 0.085 86.000 0.040

8.000 0.110 86.000 0.055 0.000 1.000 1.500*********
1.500********* 2.000 0.035 86.000 0.000 0.000 0.000
Blinder added to
DISTR1 on circuit 2
ZONE 1 2 3 TIMER(S) STARTED

SELF TRIP BREAKER TIMER STARTED
RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 0.117: Apparent Z seen by

APPARENT IMPEDANCE ENTERED ZONE(S) 3 BEHIND BLINDER 1
circuit 2 DISTR1 enter
RELAY DISTR1 # 1 CIRCUIT 1 FROM 1 TO 2 MESSAGES AT TIME = 0.133: zone 3 circle, but does
*** CIRCUIT 1 FROM 1 [HYDRO 500] TO 2 [RECEIVE 500] not cross blinder –
TRIPPED AT TIME = 0.133 timer not started
7 DIAGONAL AND 8 OFF-DIAGONAL ELEMENTS Circuit 2 DISTR1

RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 0.308: loses sight of zone
APPARENT IMPEDANCE OUTSIDE ZONES apparent Z
APPARENT IMPEDANCE ENTERED ZONE(S) 3 BEHIND BLINDER 1 Sees it again but
RELAY DISTR1 # 1 CIRCUIT 2 FROM 1 TO 2 MESSAGES AT TIME = 1.917: blinder prevents
APPARENT IMPEDANCE OUTSIDE ZONES timer start


Figure 19-15. Relay Messages

from Same Event as in Figures 19-9 and 19-12,
but with Normal Zone 3 Time and Blinder on DISTR1 of Circuit 2
19-27
Load Shedding
Prefault
Blinder
Figure 19-16. Locus of Apparent Z Seen by DISTR1 on Circuit 2
19-28
Power Receiving End

Load Shedding
Power Sending End
Figure 19-17. Time Domain Response
19-29
Model RXR1 Program Application Guide: Volume II
19.5 Model RXR1

The distance relay model (RXR1) represents a three-zone distance relay where logic char-
acteristics are similar to those of DISTR1, but where zones are shaped as polygons rather
than as blinded circles. The principal difference in logic between RXR1 and DISTR1 is
that RXR1 starts all of its zone timers when the apparent impedance locus crosses into its
outermost, or starting, zone. After timers are started, they continue to run until either one
of them times out with the apparent impedance in its zone to send a trip signal, or the
apparent impedance crosses back outside the starting zone. The timer for zone 1, 2, or 3 is
not reset if the apparent impedance goes outside that zone but remains within the starting
zone. The logic flow of RXR1 is shown in Figure 19-18.
19.6 Model CIROS1
19.6.1 Operation
The double circle or lens out-of-step relay model (CIROS1) detection is based on circular
characteristics. CIROS1 may be used to trip its own line, transfer-trip up to three other
lines, or control the supervisory flag of another relay. The purpose of the out-of-step relay
is to detect the passage of the apparent impedance locus through zone 2. The arrangements
of characteristics is shown in Figure 19-19. Setting the type code, ICON(I+1), to +1 or -1
gives a pair of full-circle boundaries. Setting the type code to +2 or -2 gives a pair of lens-
shaped boundaries. In both cases, the important thing is that the boundaries produce an
inner zone, zone 1, surrounded by a second zone, zone 2. A rapid passage through zone 2
into zone 1 is interpreted as evidence of a fault. A passage taking more than a defined
time, however, indicates a system swing or out-of-step situation. The detection logic of the
relay is shown in Figures 19-20 and 19-21.
A positive value of the type code, ICON(I+1), puts CIROS1 in tripping mode. In this
mode it has no effect on the supervisory flag, if any, specified by its IFL argument; rather
it establishes trip signals that are passed to its own and its transfer-trip circuit breakers, or
are not passed, depending upon its own supervisory permissive flags, ICON(I+16) and
ICON(I+17).
A negative value of type code, ICON(I+1), puts CIROS1 in its blocking mode, and it
sends no trip signals to its own circuit breaker. Instead it sets the permissive flag of its
supervised relay to zero on detection of an out-of-step condition. Transfer tripping actions
of CIROS1 are the same in its tripping and blocking modes. CIROS1 is made to supervise
another relay, a DISTR1 for example, by setting the calling argument, IFL, of CIROS1 to
the ICON address of the supervised relay’s permissive flag. This cross referencing must
be done via the relay editors or activity ALTR.
The supervisory, or permissive, flags of the supervised relay are set to 1 (permit tripping)
by activity DYRE, and are not affected by activity STRT. CIROS1 affects the self-trip
supervisory permissive flag of its supervised relay only on detection of the out-of-step
condition, where upon CIROS1 sets this flag to zero. CIROS1 does not ever reset the per-
missive flag to 1.
19-30
Program Application Guide: Volume II Model CIROS1
Self-Trip
Permissive Force
Flag Trip Signal
Zone 1 Flag •
Zone 1 •
Timer
Zone 2 Flag
Zone 2
Timer
Trip
Zone 3 Flag Signal
Zone 3
Pickup Timer •
Timer
Start Time
Memory
Low Current Flag
• Start
Zone 4 Signal Pickup
Flag (Zone 4)
• Reset Timer
Signal
Low Current Flag
Trip
Signal Self-Trip Self Trip
Reset Breaker Timer
Self-Trip Self Reclose

Reset Reclosure Timer
Transfer Trip Transfer

Reset Breaker Timer Trips
Transfer Trip Transfer

• Reset Reclosure Timer
Transfer Trip Reclosures
Permissive Flag
Figure 19-18. Relay Model RXR1 Logic Diagram
19-31
Model CIROS1 Program Application Guide: Volume II
CON(J+4)
Inner Zone
CON(J+1)
CON(J+6)
CON(J+3) Outer Zone
CON(J+5)
R
CON(J+2) CIROS1 Type 1
Outer Zone
Inner Zone
CON(J+4
)/2
C ON
(J+1
CON(J+3) ) /2
CON(J+1)/2
CON(J+4)/2
CON(J+6)
R
CON(J+5)
CON(J+2)
Figure 19-19. Definition of Circular Zones in CIROS1
19-32
Program Application Guide: Volume II Model CIROS1
Timer Started
Blinder
Center Timer Started

Timer #2
Timer Reset
Started Zone 1 Center
Out of Step: Zone 2 #1 Timer
Interzone Timer Reset R
Timed Out and
in Zone 1 Timer Started
Timer Started: Blinder

Enter Zone 2
Timer Reset:
Leave Zone 2
Timer Reset
a. Double Circle Geometry
Zone 2 R
Zone 1 R
R
Center Circles with Diameter
#2 1 and 2 Drawn With
Center Center #1
S R #1
R S
Blinder
b. Lens Geometry
Figure 19-20. Conditions for Setting/Resetting Timers in CIROS1
19-33
Model CIROS1 Program Application Guide: Volume II
Permissive Flag of
Supervised Relay
•
Zone 2 Flag
Start
Low Current Flag I
• Z O/S
Zone Signal • Blocking Force
T • Trip
Detector Zone 2 Flag Reset
• T • Tripping Signal
•
Zone 1 Flag Self-Trip Self Trip
Breaker
Timer
Reset
•
Self-Trip
Permissive Flag
Transfer
Trip Transfer Trips
Breaker
• Reset Timer
Transfer Trip
Permissive Flag
Figure 19-21. CIROS1 Logic Diagram
The following points should be noted in filling out the data sheet:
• The relay is at the from bus looking toward the to bus.
initialization.
• A from bus of 0 for a transfer trip bypasses the logic for that transfer trip.
• CIROS1’s own permissive flags should be set to 1 by the user if tripping on out-of-step
is required. The model does not initialize its own or any other permissive flags. Upon
detection of an out-of-step, the permissive flag of a supervised relay contained in
ICON(IFL) will be set to zero if CIROS1 is in the blocking mode.
• CIROS1 requires certain ICON and VAR locations for local storage of flags, etc. Care
should be exercised not to overwrite these storage locations.
• The first three VARs used by CIROS1 are the per-unit resistance, reactance, and cur-
rent. These values may be put into output channels. If the resolution of the relay is other
than 1, the values will be constant between sample points.
• All angles are specified in degrees from the horizontal with anti-clockwise direction pos-
itive. Figure 19-19 specifies the geometrids.
• All timer settings are specified in cycles and include any delays.
• If CIROS1 is used only for tripping on an out-of-step, the supervisory ICON can have
a value of 0. If both tripping or blocking may be considered, it is recommended that the
19-34
Program Application Guide: Volume II Model SLNOS1
supervisory ICON be set to the address of the supervised relay’s permissive flag. The
permissive flag of a supervised relay will be set to 0 only during the blocking mode upon
detection of an out-of- step.
• The self-trip permissive flag of CIROS1 prohibits only self trip. All other operation is
independent of the self-trip permissive flag.
19.7 Model SLNOS1

SLNOS1 is a generalized straight-line blinder out-of-step relay. SLNOS1 may be used to
trip its own line, transfer-trip up to three other lines, or control the supervisory flag of
another relay. A positive value of ICON(I+1) indicates that the model is used for tripping
while a negative value indicates it is used for blocking. A magnitude of 1 for ICON(I+1)
indicates the relay is the single blinder type while a magnitude of 2 indicates a double
blinder type.
The purpose of the relay is to note an out-of-step situation through a system swing by
measuring the time taken for the apparent impedance locus to pass through defined zones.
The detection logic for the single blinder type relay is shown in Figure 19-22; the zones are
defined in Figure 19-23. An out-of-step signal will be sent when after spending a defined
time in zone 1, the apparent impedance locus goes to zone 5 if it originated in zone 2 or to
zone 2 if it originated in zone 5. Figure 19-24 diagrams the remaining relay logic after an
out-of-step signal. The detection logic for the double blinder characteristic is shown in
Figure 19-25; zones are defined in Figure 19-26. This relay sends an out-of-step signal when
in zone 2 only if the apparent impedance loci takes a minimum predefined time, the inter-
zone travel time, to go from the outer zones (4 or 5) through the inner zones (3 and 1,
respectively). The remaining logic upon receipt of an out-of-step signal is as for the single
blinder shown in Figure 19-24. SLNOS1 has additional blinders to further restrict its sensi-
tive zones. The blinder type can be set to zero to ignore the extra blinders.
In the tripping mode the supervisory ICON for this model remains unchanged. In the
blocking mode the model can still be used for up to three transfer trips.
SLNOS1 is made to supervise another relay, a DISTR1 for example, by setting the super-
visory ICON, of SLNOS1 to the ICON address of the supervised relay’s permissive flag.
This cross-referencing can be done via the relay editors or activity ALTR.
The supervisory, or permissive, flags of the supervised relay are set to 1 (permit tripping)
by activity DYRE, and are not affected by activity STRT. SLNOS1 affects the self trip
supervisory permissive flag of its supervised relay only on detection of the out-of-step
condition, whereupon SLNOS1 sets this flag to zero. SLNOS1 does not ever reset the per-
missive flag to 1.
19-35
Model SLNOS1 Program Application Guide: Volume II
Zone 2 Flag
Initiating Zone 5 Flag
Zone 5 Flag
Previous Zone 5 Flag

Previous Zone 2 Flag Out-of-Step
Memory Memory
Signal
Previous Initiating
Zone 1
• Zone 2 Flag
Flag
Start
Zone 1 Flag Interzone
Signal Travel
Timer
Low Current
Flag
• Reset Signal
Zone
Detector
Zone 1 Flag
•
Figure 19-22. SLNOS1 Single-Line Blinders
Zone 2 Zone 1 Zone 5
Figure 19-23. Zone Definition for Single Blinder of SLNOS1
19-36
Program Application Guide: Volume II Model SLNOS1
Permissive Flag of
Supervised Relay
• Blocking Force
Out-of-Step • Trip Signal
Signal •
• Tripping
Self Trip Self Trip

Reset Breaker Timer
•
Self-Trip
Permissive Flag
Transfer Trip Transfer Trips

Reset Breaker Timer
•
Figure 19-24. SLNOS1 Relay Logic After Receiving Out-of-Step Signal
Zone 1 Flag
Low Current Flag
•
M Previous Zone 5 Flag
E
M Start
O Signal
R I
Previous Zone 4 Flag Z
Y Out-of-Step
Zone Low Current Flag T
Detector • Reset T Signal
Zone 3 Flag Signal
Zone 1 Flag
•
Zone 3 Flag
•
Zone 2 Flag
Figure 19-25. SLNOS1 Double Line Blinders
19-37
Model SLNOS1 Program Application Guide: Volume II
e4
e3
Zon
Zon
Zone 2
e1
e5
Zon
Zon
Figure 19-26. Zone Definition for Double Blinder SLNOS1
The following points should be noted in filling out the data sheet:
initialization.
• SLNOS1’s own permissive flags should be set to 1 by the user if tripping on out-of-step
is required. The model does not initialize its own or any other permissive flags. Upon
detection of an out-of-step, the permissive flag of a supervised relay contained in the
supervisory ICON will be set to zero if SLNOS1 is in the blocking mode.
• SLNOS1 requires certain ICON and VAR locations for local storage of flags, etc. Care
• The first three VARs used by SLNOS1 are the per-unit resistance, reactance, and cur-
rent. These values may be put into output channels. If the resolution of the relay is other
than 1, the values will be constant between sample points.
• Angles are specified in degrees from the horizontal with anti-clockwise direction
positive.
• All timer settings are specified in cycles and include any delays.
19-38
Program Application Guide: Volume II Model DPDTR1

• If SLNOS1 is used only for tripping on an out-of-step, the supervisory ICON can have
a value of 0. However, if both tripping or blocking is considered, this ICON should be
set to the address of the supervised relay’s permissive flag. This permissive flag will be
set to 0 only during the blocking mode after detecting an out-of-step.
• The self-trip permissive flag of SLNOS1 prohibits only self trip. All other operation is
independent of the self-trip permissive flag.
19.8 Model DPDTR1

Model DPDTR1 represents a power relay rate of change. The logic diagram for this relay is
shown in Figure 19-27. For this relay, both the derivative of the power and the power itself
must be above their individual thresholds to start the timer. The timer is reset, after the
delay time, if either of these thresholds is not met. If both thresholds are met after the
delay time, the line to which it is applied will trip after the breaker time.
•
Power
Flow Reset
Detector
• Timer
Start
Rate of
Change Trip
Detector Signal
Figure 19-27. DPDTR1 Logic Diagram
The following should be noted about this model:

initialization.
• A from bus of 0 for the transfer trip bypasses the logic for it.
• Timers are specified in sec.
19.9 Model TIOCR1

Model TIOCR1 is an overcurrent relay with an inverse-time characteristic as defined in
Figure 19-28. The four current points and the saturation or largest current point are all
entered as CONs and should be entered as a multiple of the pickup or threshold current
that is entered on system base, usually 100 MVA. The time-to-close relay for each point is
the time that the relay would take to close its trip contact if presented with a constant input
19-39
Model TIOCR1 Program Application Guide: Volume II
current. The actual time to close the trip contact in response to a varying input current is
given by the integral of the rate of motion of the sensitive element, as defined in Figure 19-
28. The movement occurs for any value of current exceeding the threshold current.
1 hr
1
Threshold
Time to Close Relay (sec)
 =  ---------------------------------
Time to Close
- dt
First Current Point
Trip when  = 1
2
– 1 – I 
 =  ---------------------- dt
T reset
where I = per unit current on system base
Second Current Point
Third Current Point
Fourth Current Point

Saturation Current
Threshold Current (pu)

Current
Figure 19-28. Time-Inverse Operating Characteristic of TIOCR1
The model assumes that at constant threshold current it would take one hour for the con-
tact to close. The relay does not reset immediately if current goes below the threshold; it
resets in a linear fashion at a rate proportional to the square of the deviation of current
from the threshold. With zero current, it would take Treset sec to reset, while with a current
just at the threshold it would never reset. If the current goes above the pickup threshold
current during the resetting process, the relay commences integrating toward its trip con-
tact closure from its partially reset condition.
This model allows for the remote (transfer) trip of up to three lines and allows for the dis-
connection (shedding) of load at any bus. The model can self-trip if specified as one of the
transfer-trip lines. A from bus of 0 for a transfer-trip bypasses logic for that transfer-trip.
19-40
Program Application Guide: Volume II Advanced Uses of Branch Relay Models
Also, specifying a bus number of 0 for the load shedding bypasses that logic. When load
on a bus is shed, the relay model operates identically on constant MVA load, constant cur-
rent load, and constant admittance load.
19.10 Advanced Uses of Branch Relay Models

The user may want to add information to subroutine CONET to run a more exotic coordi-
nation of relays than the simple supervisory logic built into the models. This section
describes how specific ICONs are set by the relays. An assumed breaker time greater than
zero for all the models achieved the ICON summarizations that follow.
19.10.1 DPDTR1 Flags

When the timer has timed out and the relay is still in the active zone of relay DPDTR1,
ICON I+6 is set to 2. ICON I+6 is set to 3 when the breaker has timed out.
19.10.2 TIOCR1 Flags

ICON I+12 is set to –1 when the timer has timed out for relay TIOCR1. When it is time for
the breaker to trip, ICON I+14 is set to 1.
19.10.3 DISTR1 Flags

The ICONs, which are set to 1 when zone 1, 2, and 3 time out for relay DISTR1, are I+17,
I+19, and I+21, respectively. ICON I+24 is set to 1 when the breaker opens.
19.10.4 RXR1 Flags

If any timer times out for relay RXR1, ICON I+14 is set to 1. Breaker opening occurs when
ICON I+16 is set to 1.
19.10.5 CIROS1 Flags

Both ICON I+17 and I+18 are set to 1 when an out-of-step condition is detected by
CIROS1. If it is tripping, ICON I+21 is set to 1 at breaker opening.
19.10.6 SLNOS1 Flags

SLNOS1 sets ICON I+17 to 1 when it detects an out-of-step condition. If a breaker is to
open, ICON I+21 is set to 1.
19.10.7 SLLP1 Flags

SLLP1 sets ICON I+19 to 2 when it detects an out-of-step condition. To open a breaker,
ICON I+16 is set to 1.
19-41
Model SLLP1 Program Application Guide: Volume II
19.10.8 LOEXR1 Flags

LOEXR1 sets ICONs M+2, M+4, and M+6 to 1 when it detects loss of excitation on zones
1, 2, and 3, respectively, and the corresponding zone timer has timed out. At breaker open-
ing, ICON M+8 is set to 1.
19.11 Model SLLP1

The out-of-step relay model (SLLP1) detection is based on three lenses. There are two phi-
losophies that are used for out-of-step tripping with SLLP1: (1) tripping on the way in and
(2) tripping on the way out. For tripping on the way in, a passage of time greater than T1
through zone 1 and a passage time greater than T2 through zone 2 must occur consecu-
tively before the apparent impedance enters zone 3. Figure 19-29 shows the tripping
sequence in this mode. For tripping on the way out, in addition to the passage times on the
way in, the apparent impedance must stay in zone 3 for T3 cycles and exiting passage time
T4 through zones 2 and 1 must also be satisfied. The relay will then send an out-of-step
signal upon exiting all lenses. Figure 19-30 shows the tripping sequence in this mode. Both
T3 and T4 must be greater than zero to force SLLP1 to assume tripping on the way out.
X
Timer Started.
Timer Reset
Zone 1 Zone 2 Zone 3 Zone 2 Zone 1
First Timer Started

Second Timer Started
T4 = 0 T3 = 0 T2 T1
Out-of-Step Detected,
Trip Signal Sent
Figure 19-29. Tripping on Way In
19-42
Program Application Guide: Volume II Model LOEXR1
Zone 1 Zone 2 Zone 3 Zone 2 Zone 1

Out-of-Step
Detected First Timer Started
Trip Signal
Sent Second Timer Started
T4 T3 T2 T1
Third Timer Started
R
Fourth
Timer
Started
Figure 19-30. Tripping on Way Out
Rapid passage through any zone signifies that an out-of-step has not occurred. The appar-
ent loci must subsequently leave all zones and reenter to restart the timers. The model will
print messages based on zone changes even though timers are not active. SLLP1 may be
used to trip its own line and to transfer-trip up to three other lines. The following points
should be noted in filling out the data sheet:
initialization.
• SLLP1’s own permissive flags should be set to 1 if tripping an out-of-step is required.
• SLLP1 requires certain ICON and VAR locations for local storage of flags, etc. Care
• The first three VARs used by SLLP1 are per-unit resistance, reactance, and current.
These values may be put in output channels.
• All time settings are specified in cycles.
• Resetting of timers is assumed to be instantaneous.
19.12 Model LOEXR1

Model LOEXR1 detects loss of excitation that can lead to armature core-end heating (end-
turn heating) or rotor overheating due to operation as an induction generator, and may be
19-43
Model LOEXR1 Program Application Guide: Volume II
detrimental to system performance due to the var drain imposed on the system. End-turn
heating is more often associated with under-excitation operation and forms a part of the
generator capability curve. Rotor heating due to induced rotor currents may cause danger-
ous over-temperatures in just several minutes. Stator over-heating may also occur due to
combination of the power being delivered by the generator and the large var flow into the
generator, but the rate of heat buildup will generally be considerably less than that occur-
ring in the rotor. In many cases, the effect of the loss of var supply and the var drain on the
system may be paramount to other effects and result in instability and severe low-voltage
conditions in the vicinity of the failed generator.
The loss of excitation condition is generally detected by under-current or under-voltage
relays in the field circuit, or by directional distance relays looking into the generator from
the generator terminals and modeled by LOEXR1. Operation of the current and voltage
relays in the field circuit is fairly obvious, and operation of the distance relay can be most
readily visualized by use of the RX diagram shown in Figure 19-31.
As the generator field is lost, rotor direct axis flux will decay, and the machine will speed
up and go out-of-step with the rest of the system. The shock to the system as the machine
goes out-of-step will not be severe because the generator rotor flux level (or machine
internal voltage) will be low.
80 MW, 80 Mvar
X
80 MW, 40 Mvar
80 MW, 100 Mvar
Assumed Initial
Operating Point
80 MW, 0 Mvar
R
80 MW, –200 Mvar

Distance
Relay Loss of Field
Trajectory
80 MW, –40 Mvar
80 MW, -80 Mvar

80 MW, – 100 Mvar
Figure 19-31. Loss of Field Trajectory and Protective Relay
Neglecting any power swings, due to flux in the rotor, the machine will make the transi-
tion from synchronous generator to induction generator including a period of increased
rotor speed with approximately constant power output. During the transition, however, var
output will decrease to zero and go negative, to a value two to four times machine rated
var output. Figure 19-31 indicates the trace or locus that would be scribed by a 100-MVA
19-44
Program Application Guide: Volume II Model SLYPN1
generator (0.8 pf) initially operating at rated output and for which excitation is lost. Hold-
ing the power flow at the relay CT and PT locations constant and varying var flow from
plus infinity to minus infinity allows a circle to be scribed on the RX diagram as indicated
in Figure 19-31.
The LOEXR1 model has been written to accommodate up to three zones, though usually
only one or two are used. The zone reach (diameter) should be set to zero to disable a
zone. The model has also be supplied with a voltage pickup point, VOLPIC. On some
relay units, the voltage must drop below this value for the relay to trip. VOLPIC should be
set to a large number if voltage is to be ignored.
Distances and reaches (diameters) should be entered on the machine base, MBASE of the
unit being protected. The center distances are normally negative values based on the
assumption that relays are set looking into the generator from its terminals, typically at a
90° angle.
19.13 Model SLYPN1

SLYPN1 models a positive-sequence distance relay, such as the GE solid-state types
SLYP51B and SLYP51D used for directional comparison relaying. This model also repre-
sents, as an option, overcurrent relays such as the GE SLCN51D, which is often provided
with the above mentioned units.
Directional comparison relaying often requires an additional distance relay at each termi-
nal; this model was written so that identical units with different setpoints are assumed to
be located at both line ends. These units are first made up with underreaching positive-
sequence directional mho tripping functions for zone 1 protection. Either unit can send a
trip signal to the breaker. This first zone protection is not sensitive (i.e., it ignores) to the
relay unit looking away from the line or any optional blocking schemes provided with this
relay. Figure 19-32 indicates the method of directional comparison relaying assumed in this
model. This model has included in it two directional units at each terminal, one unit look-
ing through the protected line, like a normal second zone unit, and one unit looking away
from the line. These two units are indicated as Z2 and Zb in Figure 19-32. At each terminal
the Z2 unit is connected to trip through a normally closed pair of contacts operated by a
receiver. The Zb unit is connected only to operate the transmitter, which in turn picks up
the receiver at each terminal. The Zb unit is arranged to have faster operating time (instan-
taneous) then the Z2 unit so that when an external fault occurs the Zb unit can successfully
block tripping by the Z2 unit.
Operation then occurs as follows. Fault F1 (anywhere on the protected line) will cause
both the Z2 units to pickup and trip with high-speed. Faults F3 and F4 will cause the Zb
unit of relay R1 and the Z2 unit of relay R2 to pickup. However, the Zb unit will pick up
first and transmit a block signal that will prevent tripping by the Z2 unit. Fault F5 will
pickup only the relay R1 Zb unit with no consequence. Fault F6 will prompt no action
from the directional comparison relaying.
19-45
Model SLYPN1 Program Application Guide: Volume II
Zb
Z2
Z2
F3
F6 F5 F4 F1 R2
Z2 Z2
R1
Zb
RCVR RCVR
XMTR XMTR
Figure 19-32. Directional Comparison Scheme Assumed
The zone 2 unit has a lens shape; the blocking relay has a standard circular mho shape.
The model reports that the apparent impedance has entered zone 4 when the reverse reach-
ing relay looking away from the terminal has picked up. When the reverse looking relay
blocks, this model assumes the zone 2 timer is reset.
As an option, the SLYP51X relays may have a positive-sequence out-of-step distance
function with a characteristic concentric with the zone 2 lens. The model flags the appar-
ent impedance as being in zone 3 when it is in the area between the lens and its concentric
circle. A slow passage through zone 3 thus indicates a swing or out-of-step situation. If the
loci passages takes longer than the interzone travel time specified by the user, this optional
function will block subsequent zone 2 tripping by this model. When blocked, the zone 2
timer will reset. The zone 3 timer only resets when the apparent impedance goes outside
the outer circle.
The model also has an optional overcurrent supervisory function provided with it. If the
supervisory function is set for blocking, current above the user specified supervisory level
will cause zone 2 tripping to be blocked and the zone 2 timer to be reset. When the super-
visory function is set for tripping, current above the user specified supervisory level will
cause a trip signal to be sent instantaneously to the breaker regardless of any blocking sig-
nals or if the apparent impedance falls in the R-X plane. The model is also assumed to
have single attempt reclosure for zone 1 trips only. The model assumes all distances and
reaches are entered as positive values, and angles are between 0 and 90°. As for all PSS®E
relays, all times should be specified in cycles.
This relay has one additional feature that other PSS®E relays do not have: the capability to
disconnect one phase. As part of the input data for the model, the user can place equivalent
positive-sequence data to model the protected line while one phase is opened. This equiv-
19-46
Program Application Guide: Volume II Model SCGAP2
alent may be obtained from activity SPCB or by other means. If the user inputs a Xequiv
greater than 1000, the model assumes that all three phases are to be tripped.
The user should be careful, when using this model to trip a single phase, to ensure that no
other relays are present on the line because the apparent impedance values and currents
the other relays will see while one phase is opened will be calculated on the basis of the
equivalent.
19.14 Model SCGAP2

The series capacitor gap relay model (SCGAP2) represents a current-induced relay that
shorts out a series capacitor (flashes the gap) and can also be used to protect series induc-
tors. SCGAP2 allows the user to specify first a monitored element and then the branch to
be bypassed; they may be the same. When current in the monitored branch exceeds the
gap firing current, a signal is sent to bypass the series capacitor branch. It bypasses after a
communication delay, which can be specified as zero.
When the capacitor has flashed, a shorting switch timer starts. If the current does not go
below the reinsertion current and satisfy the reinsertion duration, this switch permanently
closes and does not allow capacitor reinsertion. The user specifies the number of allow-
able reinsertion attempts.
The model will transfer-trip up to three lines. Transfer-tripping can be initiated either the
first time the capacitor bypasses or when the shorting switch closes.
19-47
Model SCGAP2 Program Application Guide: Volume II
19-48
Chapter 20
Load Modeling
Chapter 20 - Load Modeling
20.1 Introduction – Application of Load Models

Load characteristic and load relay models may be introduced by subsystem. Subsystems
can be designated as bus, owner, zone, area, or all. Each load model is available as a fam-
ily of models; one for each subsystem type. For example, the IEEE load characteristic
model is available in five forms:
Subsystem Type Model Name
Bus IEELBL
Owner IEELOW
Zone IEELZN
Area IEELAR
All IEELAL
These models are collectively referred to as the IEEL type models or as the IEELBL fam-
ily of models. The last two characters of the model name refer to the subsystem type. This
convention is applied throughout the load model library.
The presentation of a single data record for a subsystem model can generate multiple
instances of the model’s application. For example, presenting a bus-specific model with a
wildcard element identifier (e.g., load-related model with IL=’*’) will generate an
instance of the model for every load connected to the bus. Also, presenting an area spe-
cific subsystem model with a specific identifier (e.g., load-related model with IL=’1’) will
generate as many instances of the model as there are loads within the area that are
assigned that identifier.
Precedence is an issue unique to subsystem models. The data presentation for subsystem
models may involve intersection among the subsystems. The data may contain a presenta-
tion for both a bus-specific model and an area-specific model, and the bus associated with
the first model may belong to the area associated with the second. Therefore, a precedence
order is provided to accommodate subsystem models. The decreasing order of precedence
for resolving conflicts is from the most specific to most general subsystem; bus, owner,
zone, area, and all. A model applied to a bus will take precedence over all conflicting
models applied to any other subsystem in which the bus exists. A model applied to an
20-1
Load Modeling PSS®E 33.4
Load Characteristic Models Program Application Guide: Volume II
owner will take precedence over any conflicting model applied to a zone, area, or to all
buses.
A precedence order is provided to accommodate conflicts in element identifier as well. A
data record that introduces a subsystem model with a specific element identifier will take
precedence over a subsystem model with a wildcard (’*’) identifier.
Load-type models employ data sharing among the loads for which a model is applied.
That is, one set of CON and ICON data is used for all applications of the model. However,
if required, VAR, STATE, and reserved ICON space is allocated for each load to which
the model is applied. In other words, VAR, STATE, and reserved ICON data is unique for
each instance of the model’s application. For example, if the user enters one CIM5BL
model data record (induction motor model) with a wildcard load identifier, specified for a
bus with three loads, then the model will be applied to all three loads. Space will be
reserved for only one set of CON and ICON data, as that data is shared among all applica-
tions of the model. However, space will be reserved for three sets of VARs, STATEs, and
reserved ICONs (one for each application of the model).
For each load, there is one available slot for a load characteristic model and one available
slot for a load relay model. That is to say, it is not permissible to apply more than one load
characteristics model or more than one load relay model to the same load.
When DYRE is being run and a load record is encountered in which the subsystem is not
found (for example, load on a bus), this record is ignored and no message is printed. This
is consistent with how machine models are handled if the machine does not exist.
Further discussion of subsystem models can be found in the PSS®E Program Operation
Manual, Section 14.1.5.
20.2 Load Characteristic Models
20.2.1 Basic Considerations

The PSS®E package recognizes three basic algebraic load versus voltage characteristics in
its power flow database (see Section 7.2). These characteristics: constant MVA, constant
current, and constant admittance, are carried through to dynamic simulation.
Activity CONL may be used either in power flow work prior to initialization of a dynamic
simulation, or during dynamic simulation work, to change the mix of the three basic com-
ponents in the loads. These components, while useful in the absence of better information,
do not always give an adequate characterization of a system’s load versus voltage charac-
teristic. They cannot recognize the dependence of load components, other than constant
admittance, on bus frequency.
More detailed load versus voltage and bus frequency characteristics are handled by a set
of load models that participate directly in the network solution at each time step. All of
these models determine values of load current on the basis of local bus frequency, voltage,
and the algebraic or dynamic characteristics of the model.
All models, with the exception of the LDFR type models, will replace the characteristics
of ALL components of load (constant MVA, current, and admittance), at loads for which
the models are applied. The model’s initial load level is determined by first reconverting
all load components to constant MVA.
20-2
PSS®E 33.4 Load Modeling
Program Application Guide: Volume II Load Characteristic Models
The effective value of load, as the models influence it, may be monitored during the simu-
lation by assigning the desired loads to output channels in activities CHSB or CHAN. The
user should be aware that the models convert the effective loads to current injections dur-
ing the simulation. Therefore, while assigned output channels will show the correct effec-
tive loads, the power flow listing and reporting activities available through activity LOFL
will not recognize the effective load.
20.2.2 Algebraic Characteristic Models
20.2.2.1 LDFR Type Models (LDFRBL, LDFROW, LDFRZN, LDFRAR, LDFRAL)

The LDFR family of models (LDFRBL, LDFROW, LDFRZN, LDFRAR, LDFRAL) make the
constant current and MVA components of all loads to which the model is applied depen-
dent on bus frequency. These models make load dependent on frequency in accordance
with:
where the constants m, n, r, and s are given by CON(J) through CON(J+3). These indices
are not necessarily integers, may be negative (if meaningful), and may be zero if the corre-
sponding load components are independent of frequency.
The models have no effect on constant admittance load or shunt devices as specified by
line or bus shunt data. Frequency dependence of all shunts is always represented in accor-
dance with the inductive or capacitive sign of the shunt admittance.
20.2.2.2 IEEL Type Models (IEELBL, IEELOW, IEELZN, IEELAR, IEELAL)

All constant MVA, current and admittance load for which the models are applied is
replaced by a new load component defined by:
where Pload and Qload are the real and reactive power loads that would be drawn at nominal
frequency and one per unit voltage.
20-3
The a and n constants in the model need not be an integer, maybe negative, and may be
zero.
The effective value of load as it varies during the simulation may be tracked by assigning
the load to an output channel in activities CHSB or CHAN. The effective load is treated as
a current injection from a fixed portion of load and a variable portion of load. Therefore,
while the output channel will show the correct effective load, the power flow listing and
reporting activities available through activity LOFL, will only show the fixed portion.
Compatibility note: The obsolete models IEELCB, IEELCZ, and IEELCA only
recognized, and operated on, the constant MVA component at loads for which these
models were applied.
20.2.3 Complex Models
20.2.3.1 CLOD Type Models (CLODBL, CLODOW, CLODZN, CLODAR, CLODAL)

The CLOD type models (CLODBL, CLODOW, CLODZN, CLODAR, CLODAL) replace all con-
stant MVA, current, and admittance load with a composite load consisting of induction
motors, lighting, and other types of equipment such as would be fed from many typical
substations. The models may represent a composite load of induction motors, lighting, and
other types of equipment such as would be fed from many typical substations. It is
intended for use in situations where it is desirable to represent loads at the dynamic level,
as distinct from the algebraic characteristic level used in power flow, but where detailed
dynamics data is not available. The models allow the user to specify a minimum amount
of data stating the general character of the composite load. It uses this data internally to
establish the relative sizes of motors modeled in dynamic detail and to establish typical
values for the detailed parameter lists required in the detailed modeling.
The models assume that all load components are connected at 0.98 per-unit voltage. At
initialization, a tap is calculated to obtain that voltage based on the load and voltage shown
on the bus at which it is connected. By entering R and X values in the appropriate CONs,
a user may add some distribution transformer impedance or distribution line impedance.
The load on the bus is then split according to the percentages the user inputs. The perfor-
mance curves and data used for typical large motors is shown in Figure 20-1. The small
motor curves are shown in Figure 20-2.
The distribution network between the supply point and the load is not modeled explicitly
in the power flow. The CLOD type models permit representation of transformer current, if
the user wants to represent the saturation effects of distribution transformers. The curve
shown in Figure 20-3 (SAT2 model) handles the transformer saturation. The discharge
lighting part of the load is handled as follows:
1. Real part of load as constant current.
2. Imaginary part of load as the voltage raised to the 4.5 power.
3. For voltage between 0.65 and 0.75 per unit, as 1) and 2) times a linear reduction.
4. For voltage below 0.65 per unit, the lighting is assumed to be extinguished (i.e., no
load).
20-4
During initialization the discharge lighting is assumed to have approximately a 0.9 power
factor.
Figure 20-1. Large Motor Performance Curve

(H = 1.0, Load Damping Factor = 1.0, Initial Slip = -0.00837)
20-5
Figure 20-2. Small Motor Performance Curve

(H = 0.6, Load Damping Factor = 1.0, Initial Slip = -0.02149)
20-6
V
Air Gap Line
Im
Figure 20-3. Voltage versus Magnetizing Current
The make-up of the composite load is specified by CON(J) through CON(J+4) of

CLOAD. These CONs specify the percentage of the total load (as shown in the power
flow case) that is due to large motors, small motors, magnetizing current, discharge light-
ing, constant MVA devices, and voltage sensitive load. The percentages may be specified
arbitrarily on the basis of the user’s knowledge of the nature of the particular plant, feeder,
or substation.
The value of CON(J+2) stating the percentage of load current due to magnetizing should
not exceed about 10%. All real load remaining on the bus after applying the specified per-
centages varies as the voltage is raised to the KP power. The reactive power left to balance
the case varies as the voltage is raised to the second power.
Compatibility note: The obsolete model, CLOAD, only recognized, and operated
on, the constant MVA component at loads for which this model was applied.
20.2.4 Induction Motor Models
20.2.4.1 Level of Detail

Induction motors and their driven loads, while not affected by rotor angle dynamics, are
sensitive to changing frequency and voltage. The driven loads usually consume increased
power at increased speed and hence contribute to system damping. Induction motor loads
may be modeled in three levels of detail:
1. By the standard single-valued voltage/load characteristics provided by activity CONL
and by the voltage/frequency/load characteristics provided by models such as
LDFRBL.
2. With the dynamics of the rotating load represented by the inertial differential equation,
and with the motor’s steady-state electrical characteristic represented in detail, but with
the electromagnetic dynamics of the motor neglected. This level of detail is handled by
model CMOTOR.1
20-7
3. With both rotating load dynamics and motor electromagnetic dynamics represented in
detail. This model is handled by the family of CIM5BL, CIM6BL and CIMWBL models,
as well as CIMTR2 and CIMTR4.
Approach 1 is reasonable for many of the loads in a large-scale system-wide study because
the details of individual loads are often not known and the provision of detailed
motor/load data for thousands of loads may be a difficult task. Approach one is not ade-
quate for studies of events where the transient behavior of motor loads has a critical effect
on bus voltages and hence on their own, and other, loads in the system. It is also, of
course, unusable where the induction motors themselves are of specific interest.
Approach 2 only recognizes flux linkages for which time variation consists of an equilib-
rium component corresponding to operation at fixed slip and voltage influence the rotor of
the induction motor. This component is a unidirectional when the reference axes are syn-
chronized with supply frequency. This approach neglects the synchronizing action pro-
vided by the induction machine during transients and as well as flux and voltage decay
following tripping.
Approach 3 models a transient component that is zero in the steady state but takes the
form required to change rotor flux linkages when the machine is subjected to a sudden
change of supply frequency or voltage. The magnitude of this component is determined by
the disturbance applied to the machine, and its decay is governed by the transient and sub-
transient time constants of the rotor winding.
20.2.4.2 Principal Time Constants

Unlike synchronous generators where principal time constants, both inertial and electro-
magnetic, are relatively long and comfortably compatible with the 0 to 30 rad/sec band-
width of PSS®E; induction motor dynamics usually have time constants that correspond to
the upper end of the PSS®E bandwidth.
The transfer function relating motor speed to electrical torque is:
where:
Dm = Te T1
n
+ n
1 At PSS®E-26, CMOTOR is obsolete and has been superseded by the model CIM5BL. Discus-
sions concerning CMOTOR have been provided for backward compatibility with pre-PSS®E-
26 dynamic simulation setups. The user should refer to the PSS®E Compatibility Reference for
more information.
20-8
Te
= Slope of motor torque-speed characteristic.
n
T1
= Slope of load torque-speed characteristic.
n
Hm = Rotational inertia of motor and load.
20-9
A typical large motor might develop full-load torque at a slip of about 0.005 pu, which
corresponds to:
Te 1.0
 = 200
n 0.005
A typical centrifugal pump or fan load might have a characteristic corresponding to:
T1
 2
n
A typical motor plus load inertia for a centrifugal pump or similar load might be 1.0 sec.
These data give the time constant governing speed changes as
2Hm 2×1
T = = = 0.01 sec
Dm 200
This time constant corresponds to the outer edge of the PSS®E bandwidth and may require
the integration step to be reduced from the standard value of 0.00833 sec. The short induc-
tion motor inertial time constant has a direct effect on the electrical dynamics of the motor
because the electrical slip, and hence shaft speed, has a direct effect on the electrical
equivalent circuit. As a result, system simulations including induction motors should be
checked carefully and repeatedly for evidence of numerical instability.
20.2.4.3 Motor Equivalent Circuit and Data (IMD)

Induction motor models require data either as the parameters of the equivalent circuit
shown in Figure 20-4 or as the standard parameters Ls, Ls, Ls, Tso, Tso, and Ra, the
apparent stator reactances, time constants, and resistance of the motor. Induction motor
data is seldom given in these forms, however. The more usual set of motor data gives:
• Rated torque and slip.
• Rated power and power factor.
• Starting torque and current.
• Starting power factor.
• Peak torque.
• •
Ra La L1
L2
Lm R1
-------
S
R
------2-
S
• •
Figure 20-4. Circuit for Type 2 Specification in Program IMD
20-10
The required equivalent circuit data must be estimated from the nameplate and other avail-
able data. The estimation of equivalent circuit data can be handled quite effectively in
many cases by the use of the support program, IMD, which allows the user to propose trial
equivalent circuit data and calculates the corresponding motor performance curves. The
operating procedure for IMD is provided below, and a sample dialog from a run of IMD is
shown in Figures 20-5 and 20-6.
1. Start the program by typing the command IMD.
2. Enter INITIAL SELECTIONS for program setup.
Terminal voltage should match power flow voltage at motor bus. Use Type 2 for equiv-
alent circuit in Figure 20-4, Type 1 for equivalent circuit of Figure 20-14.
3. Enter EQUIVALENT CIRCUIT PARAMETERS for motor definition.
Values are entered per-unit on motor base. Specify R2 = L2 = 999 for single-cage
motor.
4. Enter INDIVIDUAL CURVE POINTS for checking performance at specified motor
speeds.
5. Enter SCALES for setting speed, current, torque and power factor scales prior to plot-
ting and printing.
6. Select OUTPUTTING DYRE if desired.
7. Enter PRINTING DEVICE to redirect tabular output, e.g., hard copy output.
20-11
$ imd
PSS®E Version 20.0

Copyright (c) 1992 Power Technologies, Inc.
This program is a confidential unpublished work created and first
licensed in 1976. It is a trade secret which is the property of
Power Technologies Inc. All use, disclosure, and/or reproduction
not specifically authorized by Power Technologies Inc. is prohibited.
This program is protected under the copyright laws of non-U.S.
countries and by application of international treaties. All Rights
Reserved Under The Copyright Laws.

ENTER DESIRED CATEGORY:
0 FOR EXIT,
1 FOR INITIAL SELECTIONS,
2 FOR EQUIVALENT CIRCUIT PARAMETERS,
3 FOR PRINTING DEVICE,
4 FOR SCALES,
5 FOR TABULAR OUTPUT,
6 FOR GRAPHICAL OUTPUT,
7 FOR FLUX MODEL PARAMETERS,
8 FOR INDIVIDUAL CURVE POINTS: 1
CURRENT INDEPENDENT VARIABLE IS 'SPEED',

ENTER 0 FOR 'SPEED', 1 FOR 'SLIP':
CURRENT VALUE OF ETERM = 1.000, FREQ = 60.00,

ENTER NEW TERMINAL VOLTAGE, FREQUENCY :
SYSTEM BASE = 100.0, MOTOR BASE = 100.0,

ENTER NEW SYSTEM BASE AND MOTOR BASE : ,3
CURRENT MOTOR TYPE = 2

ENTER 1 FOR TYPE 1, 2 FOR TYPE 2 MOTOR (USE TYPE 2 FOR CMOTOR) :
0 FOR EXIT,
4 FOR SCALES,
CURRENT MACHINE PARAMETERS ARE:

RA LA LM R1 L1 R2 L2 Default motor parameters
0.0380 0.0830 3.0000 0.0550 0.0280 0.0110 0.0550
ENTER NEW RA, LA, LM, R1, L1, R2, L2: .01 .09 4.2 .035 .09 .012 .04
NEW MACHINE PARAMETERS ARE:

RA LA LM R1 L1 R2 L2 Initial estimate of motor parameters
0.0100 0.0900 4.2000 0.0350 0.0900 0.0120 0.0400
Figure 20-5. Use of IMD to Estimate Induction Motor Equivalent Circuit Parameters
to Match Given Nameplate Data (Sheet 1 of 3)
20-12
0 FOR EXIT,
4 FOR SCALES,
ENTER SPEED : .9915 Rated speed
MOTOR BASE MVA = 3.00000
SYSTEM BASE MVA = 100.00000
AT PU SPEED OF 0.99150
AND PU VOLTAGE OF 1.00000 Motor at rated speed with initial
TORQUE = 0.86435
CURRENT (MAG.) = 0.95815 estimate: current too low,
POWER FACTOR
P + JQ (P.U.)
=
=
0.91168
0.87353 +J 0.39370
power factor too high
YMOTOR (MOT BASE) = 0.87353 +J -0.39370
YMOTOR (SYS BASE) = 0.02621 +J -0.01181
ENTER SPEED : 0

AND PU VOLTAGE OF 1.00000 Motor at zero speed with initial
TORQUE = 0.51029
CURRENT (MAG.) = 5.18997 estimate: torque too low,
POWER FACTOR = 0.15022 current too low
P + JQ (P.U.) = 0.77965 +J 5.13108
YMOTOR (MOT BASE) = 0.77965 +J -5.13108
YMOTOR (SYS BASE) = 0.02339 +J -0.15393
ENTER SPEED :
Several iterations are needed to refine motor parameters
0 FOR EXIT,
4 FOR SCALES,
CURRENT MACHINE PARAMETERS ARE:

RA LA LM R1 L1 R2 L2
0.0100 0.0900 4.2000 0.0350 0.0900 0.0120 0.0400
ENTER NEW RA, LA, LM, R1, L1, R2, L2: .004 .083 4.0 .04 .08 .011 .05
NEW MACHINE PARAMETERS ARE:

RA LA LM R1 L1 R2 L2 Final selection of
0.0040 0.0830 4.0000 0.0400 0.0800 0.0110 0.0500
motor parameters
0 FOR EXIT,
4 FOR SCALES,
Figure 0-0 (Cont). Use of IMD to Estimate Induction Motor Equivalent Circuit Parameters to Match
Given Nameplate Data (Sheet 2 of 3)
20-13
ENTER SPEED : .9915

AND PU VOLTAGE OF 1.00000
TORQUE
CURRENT (MAG.)
=
=
0.90604
1.00001
Motor at rated speed with final parameters:
POWER FACTOR = 0.91003 current OK, power factor OK
P + JQ (P.U.) = 0.91004 +J 0.41455
YMOTOR (MOT BASE) = 0.91004 +J -0.41455
YMOTOR (SYS BASE) = 0.02730 +J -0.01244
ENTER SPEED : 0

AND PU VOLTAGE OF 1.00000 Motor at zero speed with final parameters:
TORQUE = 0.71758
CURRENT (MAG.) = 5.59853 current OK, power factor OK,
POWER FACTOR
P + JQ (P.U.)
=
=
0.15057
0.84295 +J 5.53471
torque OK (0.718/0.906= 0.793 pu)
YMOTOR (MOT BASE) = 0.84295 +J -5.53471
YMOTOR (SYS BASE) = 0.02529 +J -0.16604
ENTER SPEED :
0 FOR EXIT,
4 FOR SCALES,
ETERM TYPE RA LA LM R1 L1 R2 L2
1.000 2 0.0040 0.0830 4.0000 0.0400 0.0800 0.0110 0.0500
SPEED SLIP CURRENT TORQUE PWR FACT Y - MOTOR

0.0000 1.0000 5.5985 0.7176 0.1506 0.8430 -5.5347
0.0200 0.9800 5.5886 0.7199 0.1512 0.8448 -5.5244
0.0450 0.9550 5.5759 0.7227 0.1519 0.8471 -5.5112
Abbreviated
0.9700 0.0300 2.8856 2.2632 0.7959 2.2965 -1.7471 tabular output
0.9950 0.0050 0.6275 0.5470 0.8742 0.5485 -0.3047
1.0000 0.0000 0.2449 0.0000 0.0010 0.0002 -0.2449
0 FOR EXIT,
4 FOR SCALES,
$
Figure 0-0 (Cont). Use of IMD to Estimate Induction Motor Equivalent Circuit Parameters to Match
Given Nameplate Data (Sheet 3 of 3)
20-14
Full 0.91
Full Load Power Factor = 0.91
PSS®E 33.4
Starting
Starting Current
Current5.6
= 5.6
p.u.pu
Torque =
Pullout
PulloutTorque
2.45/0.906
2.45/0.906 == 2.70 pu
2.70 p.u.
20-15
Starting Power Factor = 0.15
Starting Power Factor 0.15
Full
FullLoad Torque
Load Torque =
0.906/0.906
0.906/0.906= =1.0
1.0p.u.
pu
Figure 20-6. Plotted Induction Motor Performance from IMD

Starting Torque =
Starting Torque Full
FullLoad
LoadCurrent
Current=1.0
1.0p.u.
pu
0.718/0.906
0.718/0.906 = 0.793 pu
= 0.793
Load Characteristic Models
Load Modeling
Establishing equivalent circuit parameters within IMD requires an iterative process of

adjusting the values and checking how closely they match the given standstill and related
conditions of the motor. The following guidelines relate to the equivalent circuit shown in
Figure 20-4.
1. When executing IMD for rated conditions, current and voltage have to be 1.0 pu. The
torque at rated conditions is less than 1.0 pu (see Section 20.2.6.1.1). If, for example,
torque under rated conditions is 0.9 and the motor data set specifies 135% starting
torque, starting torque from IMD should be 0.9  1.35 = 1.215.
2. Estimate La and L1
La + L1  1/(per-unit locked rotor current)
assume initially that La  L1.

3. Adjust Lm to set full-load power factor.
4. Set Ra initially to 0.01-0.02, then adjust to set starting power factor. The starting power
factor is given roughly by (Ra + R1)/(La + L1).
5. Adjust R1 to set starting torque. For a single-cage motor the starting torque is given
roughly by Is2R1, where Is is the starting current. While increasing rotor resistances
increases starting torque, it also increases slip at full load.
6. Set L2 initially to 0.04-0.05. L2 may have to be increased to get starting load current to
flow into R1 to set starting torque.
7. Adjust R2 to set slip at full load. Initial value should be close to desired slip. L2 may have
to be adjusted in conjunction with R2. Increasing rotor resistances decreases per unit
pullout (peak) torque in relation to rated and starting torque.
As an example, consider an induction motor with the following data:

Voltage: 3.3 kV
Current: 525 A
Rating: 3500 hp
Speed: 3569 rpm
Inertia: 0.9 sec
Power Factor (running): 0.91
Locked rotor pf: 0.15
Locked rotor current: 5.6 pu
Locked rotor torque: 0.8 pu
Pullout torque: 2.7 pu
Using IMD and the procedure outlined in this section, it is possible to determine the equiv-
alent motor circuit of the above data.
20-16
20.2.4.4 Representation of Rotor Flux Linkages in Induction Motors

The CIM5 type, CIM6 type and CIMW type family of induction motor models, as well as
CIMTR2 and CIMTR4, fully recognize the rotor dynamic electromagnetic behavior. Their
block diagrams are shown in Figure 20-7.
20.2.4.4.1 CIM5 Type Induction Motor Models (CIM5BL, CIM5OW, CIM5ZN, CIM5AR,
CIM5AL)
The CIM5BL family of models (CIM5BL, CIM5OW, CIM5ZN, CIM5AR, CIM5AL) can be used
to model either single-cage or double-cage induction motors including rotor flux dynam-
ics. The motor is modeled in the power flow as a bus load where all of the load at a spe-
cific load id is taken as the steady-state motor load. These models may be applied to an
individual load or a subsystem of loads. For example, the CIM5BL model can be applied
to a specific load in order to model a specific induction motor. The CIM5AR model can be
applied to all loads in a specific area in order to model generic motor load using typical
data. The load composition can be any percentage of constant MVA, constant current or
constant admittance.
The data input for the model are the equivalent circuit impedances for either a Type 1 or a
Type 2 equivalent circuit model (Figure 20-8). The Model Type is specified in
ICON(M+2). The CIM5 models translate the equivalent circuit parameters into transient
parameters (flux linkage components) for use in the actual model calculations, according
to the equations in Section 20.2.6.0.1.
The equivalent circuit impedances are specified in per unit on motor MVA base. The user
has two choices for the specification of motor MVA base:
1. When CON(J+11) > 0., the motor MVA base is specified as CON(J+11).
2. When CON(J+11) = 0., the motor MVA base is specified as CON(J+12)*MW load.
Option 2 is incorporated to allow the motor size to be scaled along with the load at the bus.
The CIM5 type models use the following equation as its representation for mechanical
load torque:
Tload = Tnom (1 + )D
where  is the motor speed deviation from nominal (per unit slip), D is the load damping
factor, CON(J+17), and Tnom is the motor load torque at synchronous speed.
At initialization, the CIM5 type models pick up the total power for the specified load id at
the bus and together with the equivalent circuit data and the bus voltage, calculates the ini-
tial slip and reactive power consumption of the motor. The Mvar difference between the
load and the actual motor reactive consumption, VAR(L+1), is accounted for automati-
cally by the assignment of a hidden shunt at the bus, VAR(L). If the model is unable to
resolve the load with the specified motor data, the user will be alarmed and the model will
be disabled. Based on the motor load and the initial slip, the model then calculates the syn-
chronous load torque (Tnom) in order to balance electrical and load torque in the steady
state. The value for Tnom is stored in VAR(L+4). For online initialization, any value spec-
ified for Tnom in CON(J+18) will be ignored.
20-17
E KR
Tp 
X-X l
--------------
X-X l +
E' I + – +
1 + 1 +
-------- --------- X-X EI 
TS TS E --------------
– KI X-X l
X-X
--------------------- X-X l
2
 X-X l 
– +
– X – X +
IR
+ –
ER 
E' R  ----------
E
T' p 
E
EI 
--------
-
E' I  E
+
– + +
X – X II
– –
X-X
---------------------
2 X-X l
 X-X l 
1 + E KR X-X
-------- + + 1-
-------- -------------- + ER 
TS E R – TS X-X l
- +
X-X l
--------------
X-X l
Tp 
E KI
Figure 20-7. Induction Model With Rotor Flux Linkages

Represents Both Type 1 and Type 2 Models in IMD
20-18
Type 1 Type 2
RA + jXA RA + jXA jX1
jX1 jX2 jX2

R1
jXm jXm ------
R1 R S R2
------ -----2- ------
S S S
Figure 20-8. CIM5 Type Models
The models include a relay that can be used to trip the motor for an undervoltage condi-
tion. CON(J+14) is specified as the per unit voltage level (VI) for which the relay will
begin timing. CON(J+15) is the time in cycles (TI) for which the voltage must remain
below the threshold in order for the relay to trip. The breaker time delay (TB) is specified
in CON(J+16). The user may disable the relay by specifying CON(J+14) as zero.
Motor Starting Analysis

The CIM5 type models can be used for analyzing the start-up of a motor. In most cases,
the CIM5BL model will be the most appropriate to use because of the need to specify a
specific MBASE for motor starting. If the load status for the motor load in the power flow
is specified as zero, the motor will be initialized as off-line by activity STRT. At any point
in the simulation, the user may start the motor by toggling the load status to 1 (in-service)
and supplying a positive per unit value for Tnom (synchronous load torque) in CON(J+18)
to simulate the load on the motor. Any load for this bus/load ID that may be present at the
time the status is toggled to one (in-service) will be ignored.
The value specified for Tnom will vary depending on the level of mechanical load it is
intended for the motor to accelerate. In PSS®E, full mechanical load corresponds to the
rated electrical torque for the motor, which is the torque developed at one per unit voltage
and current. This value is typically in the range of 0.8 to 0.9 pu. The rated electrical torque
(which must balance the load torque, Tload, in the steady state), together with the rated
speed, enables the user to determine the value for Tnom corresponding to full mechanical
load. Program IMD can be used to find the rated torque at rated speed, which can then be
applied to the equation for load torque to find Tnom.
Tload
Tnom =
(1 + )D
The value for Tnom can also be found through an online initialization of the motor model,
where Tnom is determined (and stored in VAR(L+4)) from the power flow conditions and
motor data.
For motor starting analysis, the user must ensure that CON(J+11), MBASE is used to
specify the motor MVA base rather than CON(J+12), PMULT. Also, the simulation time
20-19
step may need to be reduced to accommodate the very fast motor starting transients. It is
recommended that the data for the motor always be tested first by running a simulation
with the motor online.
20.2.4.4.2 CIMW Type Induction Motor Models (CIMWBL, CIMWOW, CIMWZN,

CIMWAR, CIMWAL)
The CIMW type models (CIMWBL, CIMWOW, CIMWZN, CIMWAR, CIMWAL) are a special-
purpose induction motor models created for compatibility with the WSCC Stability Pro-
gram induction machine model.
The CIMW type models can be used to model either a single-cage or double-cage induc-
tion motors including rotor flux dynamics. The motor is modeled in the power flow as a
positive load where all of the load at a specific load ID is taken as the steady-state motor
load. The load composition can be any percentage of constant MVA, constant current or
constant admittance. The CIMW type models cannot be used for induction generators
(i.e., negative bus load). The PSS®E model CIMTR3, which is more appropriately mod-
eled as a power flow generator, should be used.
The data input for the model are the equivalent circuit impedances for either a Type 1 or a
Type 2 equivalent circuit model (Figure 20-9). The model type is specified in ICON(M).
The CIMW type models translate the equivalent circuit parameters into transient parame-
ters (flux linkage components) for use in the actual model calculations, according to the
equations in Section 20.2.6.0.1.
Type 1 Type 2
RA + jXA RA + jXA jX1
jX1 jX2 jX2

R
jXm jXm -----1-
R R2 S R2
-----1- ------ ------
S S S
Figure 20-9. CIMW Type Models
The equivalent circuit impedances are specified in per unit on motor MVA base. The user
has two choices for the specification of motor MVA base:
1. When CON(J+11) > 0., the motor MVA base is specified as CON(J+11).
2. When CON(J+11) = 0. the motor MVA base is specified as CON(J+12)*MW load.
Option 2 is incorporated to allow the motor size to be scaled along with the load at the bus.
The CIMW type models use the following equation as its representation for mechanical
load torque:
Tload = To(A2 + B + Co + D)
where:
20-20
Co = 1 - Ao2 - Bo - Do

o = 1 + o
To is the initial load torque, w is current motor speed, o is the initial motor speed and
o is the speed deviation from synchronous (per unit slip).
At initialization, the CIMW type models pick up the total power at each load ID for which
the model is applied, and together with the equivalent circuit data and the bus voltage, cal-
culates the initial slip and reactive power consumption of the motor. The Mvar difference
between the bus load and the actual motor reactive consumption, VAR(L+1), is accounted
for automatically by the assignment of a hidden shunt at the bus, VAR(L). If the model is
unable to resolve the bus load with the specified motor data, the user will be alarmed and
the model will be disabled. Based on the motor load and the initial slip, the model then
calculates the initial load torque (To) in order to balance electrical and load torque in the
steady state. The value for To is stored in VAR(L+4).
The models include a relay that can be used to trip the motor for an undervoltage condi-
tion. CON(J+14) is specified as the per unit voltage level (VI) for which the relay will
begin timing. CON(J+15) is the time in cycles (TI) for which the voltage must remain
below the threshold in order for the relay to trip. The breaker time delay (TB) is specified
in CON(J+16). The user may disable the relay by specifying CON(J+14) as zero.
The CIMW type models are not valid for use in motor starting analysis.
20.2.4.4.3 CIM6 Type Induction Motor Models (CIM6BL, CIM6OW, CIM6ZN, CIM6AR,
CIM6AL)
The CIM6 type models (CIM6BL, CIM6OW, CIM6ZN, CIM6AR, CIM6AL) have the detailed
load torque representation of CIMW type models, and also have the motor starting capa-
bility of CIM5 type models. The data requirement for CIM6 type models are similar to
those of CIMW type models, except that one extra CON has been added for entering the
nominal torque for motor starting analysis.
20.2.4.4.4 Models CIMTR2 and CIMTR4 – Induction Motor Models

CIMTR2 and CIMTR4 model either a single-cage or double-cage induction motor including
rotor flux dynamics. In the power flow, a generator with a negative electrical power
should be used. To model a single cage machine, either T or X should be set to zero and
ZSORCE in the power flow should be set equal to X’.
The last CON, D, for this model handles the effect of load on the motor according to the
following:
TL = Tnom(1 + n)D
where n is the per-unit deviation of shaft speed from nominal and Tnom is synchronous
load torque.
CIMTR4 uses one extra state variable but is more accurate than CIMTR2. It is recom-
mended for large frequency deviations. CIMTR4 is similar to CIMTR3 for starting except
that the user specifies the load torque at synchronous speed, a negative value. This model
cannot be the only machine in an island when using the constant island frequency mode in
extended term dynamics.
20-21
20.2.5 Single-Phase AC Motor Model

The ACMTBL model is meant to represent an aggregation of a large set of single-phase
air-conditioners. It is a grid-level model suitable for studying the impact of residential air-
conditioning loads on the dynamic performance of an interconnected power grid. The
model is able to:
- determine the air-conditioner's real and reactive power and its power factor varia-
tion in response to voltage and frequency variations
- represent the air-conditioner's response to low frequency inter-area oscillations
- reasonably represent the stalling phenomenon, as well as accurately represent
motor current, real and reactive power in the stalled state
- provide reasonable indication of air-conditioning load tripping by thermal protection
The characteristics of the ACMTBL model are consistent with the specifications for Per-
formance Model for Representing Single-Phase Air-Conditioner Compressor Motors in
Power System Studies, prepared by Western Electricity Coordination Council (WECC)
Load Modeling Task Force (LMTF) [1]. The model structure and characteristics given in
this document have been taken from [1].
This model includes the representation of:
• Compressor motor
• Compressor motor thermal relay
• Under-voltage relays
• Contactors
20.2.5.1 Compressor Motor Model

The compressor model represents the motors' behavior as a function of voltage and fre-
quency in the "run" and "stall" states. The model uses an exponential function to represent
the relationship between motor real and reactive powers versus motor voltage and fre-
quency in the "run" state. The real and reactive power curves of the compressor motors, as
obtained from voltage ramp test of the simulation model, are shown in . The model transi-
20-22
tions from the "run" state to the "stall" state at a pre-defined voltage, Vstall. The motors in
the "stall" state are represented by an equivalent impedance (Rstall + j Xstall).
Figure 20-10. Performance Model Characteristics of Compressor Motors2
20.2.5.2 Compressor Motor Thermal Relay Model

The single-phase air-conditioner motors are protected against overheating by thermal
relays. These over-current relays have an inverse time characteristic.
Figure 2 shows the model of the thermal relay, where:
Ic - compressor motor current
Tth - compressor motor heating time constant
Rstall - stall resistance
2 "AC unit model specifications", WECC Load Modeling Task Force, April 2008.
20-23
Based on the computed winding temperature and the characteristics defined by the trip-
ping temperatures, Th1t and Th2t, the model determines the fraction of compressor motors
that are not tripped by thermal protection, Kth.
Figure 20-11. Compressor Motor Thermal Relay Model
20.2.5.3 Under-Voltage Relay Model

ACMTBL models the built-in protection of the compressor motors as a definite-time
under-voltage relay with two thresholds. If the supply voltage stays below a threshold
value for a specified period of time, the motors are tripped. Once the motors are tripped by
the under-voltage relay, they are not reconnected during the simulation.
The following parameters are used to model the under-voltage relay:
Fuvr, fraction of compressor motors with under-voltage relays
UVtr1, 1st voltage pick-up (pu)
Ttr1, 1st definite time voltage pick-up (sec)
UVtr2, 2nd voltage pick-up (pu)
Ttr2, 2nd definite time voltage pick-up (sec)
20-24
The model determines the fraction of motors that are not tripped by under-voltage relay,
Kuv.
20.2.5.4 Contactor Model

The compressor motors are energized through power contactors. The ON-OFF status of
these contactors is affected by the fluctuations in supply voltage. The control circuit is typ-
ically supplied from the utility voltage through a transformer. Thus, a sag in supply volt-
age will be translated into a sag in control voltage. The contactors typically open when the
control voltage drops to 40-45%, and re-close when the control voltage restores above 50-
55% range. The model uses the following parameters to simulate the dropping and reclos-
ing of contactors:
Vc1off, Control voltage 1 at which contactors start dropping out (pu)
Vc2off, Control voltage 2 at which all contactors drop out (pu)
Vc1on, Control voltage 1 at which all contactors reclose (pu)
Vc2on, Control voltage 2 at which contactors start reclosing (pu)
The model determines the fraction of motors that do not drop out, Kcon.
20.2.5.5 Compressor Unit Model Structure

The overall model structure of the single-phase air-conditioner is shown in Figure 20-12.
The set of motors that are aggregated at a bus are categorized into two broad categories,
namely Motor A and Motor B. Both Motor A and Motor B follow the same P vs V, Q vs V,
P vs f, and Q vs f characteristics in the "run" and "stall" modes. Both Motor A and Motor
B will stall when the voltage falls below Vstall. The main difference between these motors
is in their capability to re-start. When the voltage recovers above the Vrst level, Motor B
can restart, while Motor A will remain stalled. Frst is the fraction of motors that are capa-
ble of restart. Thus, Motor B are represented by fraction Frst and Motor A are represented
by the balance fraction, i.e., 1-Frst. KthA and KthB represent the fractions of the Motor A
and Motor B loads that are not tripped by thermal relay. Kuv and Kcon are the fractions
of the motor load that are not tripped by under-voltage relay and contactors respectively.
The compressor model calculates the reactive power component of the motor load based
on the voltage of the bus. The difference between the motor reactive power and the reac-
20-25
tive component of the load connected at the bus during initialization is represented as a
shunt at the bus.
Figure 20-12. Model Structure of ACMTBL
20.2.5.6 Modeling of Time Delays

The ACMTBL models the following time delays from the instant of various threshold
voltage levels:
- Stall delay: This is the time delay from the instant the voltage drops below the stall
voltage to the instant the motor stalls. If the voltage recovers above the stall voltage
before the completion of stall delay, the stall timer is reset.
- Restart delay: This is the time delay from the instant the voltage exceeds the restart
voltage. This is applicable only for the set of Motor B. During the period of restart
delay, the model assumes that the compressor motor thermal relay will not operate
to trip the motors that are expected to restart.
20.2.5.7 ACMTBL Model Initialization

The single phase air-conditioner compressor load is represented in the power flow at a bus
as a constant MVA, or constant current or constant admittance load.
During initialization of a dynamic simulation run, the ACMTBL model replaces all con-
stant MVA, current, and admittance load at the defined bus with an aggregate single phase
air-conditioner compressor load.
Similar models are available for defining the air-conditioning loads by owner
(ACMTOW), zone (ACMTZN), area (ACMTAR) and the entire system (ACMTAL).
20-26
20.2.6 Composite Load Models

A new composite load, CMLDBL, has been developed for use in PSS®E to simulate the
dynamic behavior of an aggregate of three-phase motors, a single-phase air conditioner
motor, electronic loads and static loads connected to a low-voltage load bus. The dynamic
response is reflected at the high-voltage system bus. In addition to representing the mix of
loads at the low-voltage bus, this model includes also an equivalent circuit of distribution
transformer, substation compensation, distribution feeder equivalent, and feeder compen-
sation. The structure of the new composite load model CMLDBL is shown in Figure 20-13.
Figure 20-13. Composite Load Model Structure3
The main components of the composite load model CMLDBL are the following:
• Substation transformer with load tap changer (LTC) control
• Substation shunt
• Distribution feeder equivalent
• Feeder compensation
• Motor loads (up to four)
- Up to three different types of three-phase induction motors with built-in protection
- Single-phase air conditioner compressor motor with built in protection
• Electronic loads
• Static load
• Load shedding
Fractions of load components at Load Bus
- Motor A fraction, FmA
3 "Load Modeling",MMWG Report to TSS, August 2009.
20-27
- Motor B fraction, FmB

- Motor C fraction, FmC
- Motor D fraction, FmD
- Electronic Load Fraction, Fel
- Static load fraction is computed as Fs = 1 - FmA - FmB - FmC - FmD - Fel
Static Load Component Model
The static load component model is defined by the following parameters
- Power Factor - PFs
- Voltage exponential parameters - P1e, P1c, P2e, P2c, etc.
- Frequency sensitivity parameters - Pfrq, Qfrq
The static load model uses the following equations;
P=P0* (P1c* V/V0P1e + P2c* V/V0P2e+P3)*(1+Pfrq * Df)
Q=Q0* (Q1c* V/V0Q1e + Q2c* V/V0Q2e + Q3)*(1+Qfrq * Df)
P0=Pload* (1.-FmA-FmB-FmC-FmD-Fel)
Q0=P0* tan(arc cos (PFs))
P3=1.-P1c-P2c
Q3=1-Q1c-Q2c
Electronic Load Component Model

The electronic load component model is defined by following parameters
- Power Factor, PFel
- Voltage drop out values, Vd1 and Vd2
The electronic load model varies the load as a function of voltage
Constant P, Q down to V=Vd1
P, Q reduces to zero linearly between Vd1 and Vd2
Three Phase Induction Motor Load Model

The composite load model has the capability to represent three different types of 3-phase
induction motor loads. The various motor types are the following:
Three-phase motors driving constant torque loads, such as commercial/industrial air con-
ditioner compressors and refrigerators
Three-phase motors driving torque speed-squared loads with high inertia, such as fans
Three-phase motors driving torque speed-squared loads with low inertia, such as pumps
Motor A, Motor B, and Motor C are 3-phase induction motors. These motors use the
induction motor model CIM6BL from PSS®E standard model library. However, the input
parameters for the 3-phase motor model are required in terms of impedances and time
constants. The mechanical load torque is modeled as:
TL=T0*Etrq
Where Etrq is torque speed exponent.
20-28
The 3-phase motor model has a built-in protection based on undervoltage trip settings.
Single Phase Air Conditioner Load Model
Motor D is the single phase air conditioner load. These motors stall when voltage drops
below a set value and a portion of these motors restart when voltage recovers. It is repre-
sented by user-defined model ACMTBL. Following are some of the main features of the
single-phase air conditioner load model:
Stall characteristics
- if V<Vstall for Tstall sec
Restart characteristics
- if V>Vrst for Trst sec, Frst fraction of motors restart
Undervoltage trip
- if V<Vtr1 for Tr1 sec, Fuvr fraction trips
- if V<Vtr2 for Tr2 sec, Fuvr fraction trips
Contactor
- trips linearly between Vc1off and Vc2off
- reconnects linearly between Vc2on and Vc1on
Thermal protection
- Trips linearly between Th1t and Th2t
Substation Transformer Model
The substation transformer is modeled by the following parameters:
- Transformer short circuit reactance
- Fixed taps on both sides
- Variable tap on the low side
There is an option to model the load tap change control in dynamic simulation. The model
uses the online tap changer model OLTC1 from PSS®E standard model library to model
the dynamics of load tap change function. The online tap changing model allows the mod-
eling of transformer tap adjustments to help control low side voltage: it has two main com-
ponents. The first is the voltage sensor, which compares the input voltage to the low side
voltage control band defined by Vmin and Vmax. If the voltage input to the sensor is out
of the control band, the control will operate after the time delay has been exceeded. Thus,
the output of the regulator will be either raised or lowered until the voltage feedback into
the sensor is again within the control band.
Load Shedding
The composite load model is responsive to load shedding signal from undervoltage and
underfrequency relay connected at the system high voltage bus.
The undervoltage and underfrequency relay models disconnect a fraction of each of the
component of the composite load namely static load, electronic load, three-phase induc-
tion motor load, single-phase air conditioner load based on the applicable undervoltage or
underfrequency relay model setting. The relay model increases the feeder impedance in
20-29
reverse proportion to the load shedding fraction and reduces the feeder compensation as a
direction proportion of the load shedding fraction to simulate the tripping of an equivalent
fraction of the feeder from the substation. The substation transformer and substation shunt
values do not change due to load shedding.
CMLDBL Model Initialization
The composite load could be represented in the power flow at the system high voltage bus
as a constant MVA, or constant current or constant admittance load.
During initialization of a dynamic simulation run, the CMLDBL model replaces all con-
stant MVA, current, and admittance load at the defined bus with an aggregate single phase
air conditioner compressor load.
Similar models are available for defining the composite loads by owner (CMLDOW),
zone (CMLDZN), area CMLDAR and the entire system (CMLDAL).
During initialization, if the transformer reactance and / or feeder reactance are less than
threshold values, the respective component is not represented in the composite load
model. Furthermore, if the calculated far end load bus voltage is less than 0.95 pu, the
feeder impedance, Rfdr and Xfdr are reduced to bring the voltage above 0.95 pu.
20-30
20.2.6.0.1 Data Conversion From One

Model Representation to Another
R L
Data describing induction machines will a a
L
often come in the form of the steady-state 1
equivalent. Program IMD also uses the L
m
data in this form. Figure 20-14 shows the R
1
model noted as type 1 in Program IMD. -------
s
Figure 20-14. Induction Motor Equivalent Circuit for Type 1 Specification in Program IMD
The following equations can be used to

convert data from this form to the flux
linkage model given in Figure 20-7:
Type 1
Double Cage Single Cage
L = LA + LM L = LA + LM
L1 = LA L1 = LA
1 1
L’ = LA + L’ = LA +
1 1 1 1
+ +
LM L1 LM L1
1 L” = 0
L” = LA +
1 1 1
+ +
LM L1 L2
L1 + LM L M + L1
To’ = To’ =
oR1 oR1
L 1L M T”o = 0
L2 +
L1 + LM
To” =
oR2
where o is rated supply frequency in following equations should be used to con-

rad/sec or 377 rad/sec for a 60-Hz system; vert data into the flux linkage components:
X and L are equal in the per-unit system.
If the equivalent circuit describing the
induction motor is that noted as type 2
motor in IMD and shown in Figure 20-4, the
20-31
Type 2
Double Cage Single Cage
L = LA + LM L = LA + LM
Ll = LA Ll = LA
1 1
L’ = LA + L’ = LA +
1 1 1 1
+ +
LM L1 LM L1
1 L" = 0
L” = LA +
1 1
+
LM L1 + L2
LM + L1 + L2 L M + L1
To’ = To’ =
oR2 oR1
1 T"o = 0
1 1
+
”
Tdo L + L 1 L2
= M
oR1
20.2.6.1 Model CMOTOR4
20.2.6.1.1 Definition and Setup

Model CMOTOR represents a double-cage induction motor driving a load such as a fan or
pump where load torque is a function of its rotational speed. The motor is modeled by its
steady-state equivalent circuit as shown in Figure 20-4. A single-cage motor may be mod-
eled by setting R2 and L2 to 999. The instantaneous electrical slip and torque are given by:
s = BSFREQ – n
I12 R1 + I22 R2
Te = s
where:
BSFREQ = The per-unit frequency deviation at the bus to which the motor is connected.
n = The per-unit deviation of shaft speed from nominal value.
4 Obsolete; this discussion is provided for backward compatibility with pre-PSS®E-26 dynamics
setups.
20-32
The load on the motor is given by:

TL = Tnom(1 + n)D
where:
Tnom = The load torque when the load is turning at synchronous speed.
D = The exponent relating load torque to speed.
The value of D must be specified in CON(M+2). The specification of Tnom is described

below in the initialization discussion.
All motor equivalent-circuit parameters and driven load parameters are stated per unit rel-
ative to electrical motor base quantities. This implies that the per-unit electrical torque
developed when the motor operates at 1 per-unit voltage and 1 per-unit current is:
power factor × efficiency
Te =
1 - slip
A typical value for Te under such conditions is 0.9.

Because the induction motor can present the network with an apparent admittance varying
with a very small time constant, CMOTOR uses an implicit numerical integration scheme
and requires coordinated calls of the form:
CALL CMOTOR (N,I,M,K,J,L) in CONEC
CALL TMOTOR (N,I,M,K,J,L) in CONET
The argument values must be identical in the two calls.

The implicit integration process affects the convergence of the network solution at each
time step and usually requires deceleration from iteration to iteration. Accordingly, each
call to TMOTOR references a CON, (CON(L)), containing a deceleration factor value.
The value of this factor may be set initially at about 0.5, though trials will be needed to
determine an optimum value.
The CMOTOR model handles the electrical connection of the motor to the network by
overriding the values of the total load at the bus to which the motor is connected. The val-
ues of load are not changed in the working case even though the motor model has the
effect of changing their values for the purpose of network solutions during the dynamic
simulation run. The actual real and reactive power flow into the motor at its terminals are
available for output plotting in VAR(J+5) and VAR(J+6), respectively, and may be
selected for plotting by assigning these VARs to channels with activity CHAN. CMOTOR
may be initialized either online (running) or off-line (standstill).
20-33
20.2.6.1.2 Initialization Online

When CMOTOR is initialized online, the real power consumed by the motor is considered
to have been included in the real power load specified in the power flow case. Similarly
the reactive power consumed by the motor is considered to have been included in the con-
stant Q reactive load at the bus in the power flow case.
For example, consider a bus having a load consisting of the following:
• Controlled rectifiers drawing constant real and reactive power of 2.0 MW and 1.0 Mvar.
• A 3000-hp induction motor consuming 2.25 MW at 0.85 power factor.
• Static capacitors to deliver 1.5 Mvar at unity voltage.
• Electric heating load of 0.5 MW.
The motor reactive load is:
(2.25 / 0.85) × (1 - 0.85 × 0.85) = 1.39 Mvar
The bus load should, therefore, be specified for the initial condition power flow solution
as:
MWLOAD = (2.0 + 2.25) = 4.25
MVARLOAD = (1.0 + 1.39) = 2.39
GLOAD = 0.5
BLOAD = 1.5
ILOAD = (0 + J0)
Model CMOTOR initializes itself in accordance with the initial condition slip specified by
the user. The value of slip for a given bus voltage and load may be determined by
activity IMD. Usually a small imprecision in the estimation of initial condition slip exists
because it is often desirable to specify one value based on an assumed bus voltage and
leave this unchanged on the simulation data from one case to the next. This imprecision
will cause the motor to initialize its real and reactive consumption at values slightly differ-
ent from those used in determining the power flow data.
Figure 20-15 summarizes the action of CMOTOR on initialization. The initial condition
slip specified for the motor considered above might, for example, at the voltage existing in
the initial condition power flow solution, give a motor consumption of:
P = 2.21 MW
Q = 1.30 Mvar
This implies that the ensuing dynamic simulation run will treat the load at the bus as the
motor plus:
(4.25 – 2.21) = 2.04 MW of constant power load
(2.39 – 1.30) = 1.09 Mvar of constant Mvar load
0.5 MW of constant admittance load
1.5 MW of shunt capacitors
20-34
Initial
Data as Specified in Effective Values of Load Parameters in
Condition
Condition Converted Power Flow Initial Condition and Subsequent
Motor
Case Simulation
Consumption
Total Bus Load = 4.25+j1.74 Total Bus Load = 2.04+j1.04

MVARLOAD
ILOAD = const
> ILOAD = const
MOTOR 2.21
GLOAD+jBLOAD = 0.5+j1.5
CONSUMP- +j0.70 GLOAD+jBLOAD = 0.5+j1.5
TION 2.21
v = 0.992 +j0.70
Total Bus Load = 0.01+j0

Total Bus Load =2.25+j0.74
MVARLOAD ILOAD = const
ILOAD = const
< GLOAD+jBLOAD = 0.5+j1.5
MOTOR 2.24
GLOAD+jBLOAD = 0.5+j1.5 2.24
CONSUMP- +j0.78
TION +j0.78
0.04 = 0.0385 (Mvar)
B=
v = 1.020
1.022
v=1.02
Figure 20-15. Adjustment of Bus Loads when Model CMOTOR is Initialized as Online
20-35
If the constant Mvar load at the bus is less than the reactive power consumption of the
motor at initial voltage and slip, CMOTOR automatically adds static capacitors at the
motor terminals to make up the difference. For example, if the power flow case has:
MWLOAD = 2.25 MW
MVARLOAD = 1.39 Mvar
Bus voltage = 1.02 pu
The motor’s initial consumption is:

P = 2.24 MW
Q = 1.42 Mvar
The dynamic simulation run will represent the motor plus:

• (2.25 – 2.24) = 0.01 MW of constant power load.
• 0 constant Mvar load.
• (1.42 – 1.39)/1.02  1.02 = 0.0288 Mvar of shunt capacitors at the motor terminals
(unity voltage value).
plus constant current (ILOAD) and constant admittance load (GLOAD + jBLOAD) as
contained in the power flow setup.
The initialization and subsequent simulation of the motor by CMOTOR overrides the
specified values of bus load, BLOAD, to replace the appropriate part of each with the
motor’s instantaneous consumption, but does not modify the value of any of these data
items in the power flow working case.
CMOTOR requires the presence of constant power load equal to or greater than its con-
sumption in order to be initialized online. This requirement must be recognized when
using activity CONL to create the converted initial condition power flow. It is recom-
mended that all buses to be connected to CMOTOR be placed in a separate zone so that
they can easily be excluded from a general system-wide conversion of load characteristics
by CONL.
CMOTOR cannot initialize itself if the initial condition slip (specified by the user) and its
initial condition voltage (from the power flow solution) give an initial consumption
greater than the value of the MW load at its bus. The user can provide a small safety mar-
gin for initialization by specifying a MW load slightly larger than the motor’s expected
consumption or by reducing the value of initial condition slip slightly. Both approaches
result in a small residual constant power load remaining on the bus throughout the
dynamic simulation.
When it is initialized online, the data entered for CMOTOR should follow these rules:
1. CON(M), the motor base MVA, must be nonzero.
2. CON(M+4), the nominal load torque, may be given a nonzero value, but is ignored.
When it is initialized online, CMOTOR calculates the exact value of synchronous-speed

load torque corresponding to the initial voltage and slip, and places this in VAR(J+4),
allowing for a convenient check of the initialization. The value of VAR(J+4) should be
between 0.8 and 0.9 if the initial condition requires the motor to be at full load.
20-36
20.2.6.1.3 Example — Online Initialization

As an example, consider the small system of Figure 13-1. For this case, assume the load on
bus 151 consists of four 2.8 MW induction motors with parameters as determined by the
IMD program (Figures 20-5 and 20-6), and a controlled SCR device accounting for the
remainder of the 15 MW. Figures 20-5 and 20-6 show that the four motors will consume
approximately 10.92 + j4.97 MVA. The reactive power load of 7.9 Mvar is assumed to
consist of the 4.97 Mvar consumed by the motors plus 2.93 Mvar of reactive power con-
sumed by the SCR device.
Assume that the SCR control scheme is such that the voltage variation of the 2.93 Mvar
reactive load corresponds to constant inductive susceptance. Accordingly the (15 + j7.9)
constant MVA load at bus 151 must be changed to the following:
15 MW constant P: 10.92 for motors and 4.08 for SCR devices
4.97 Mvar constant Q: for motors
2.93 Mvar constant B: for SCR device
This does not imply that there will be (15 + j4.97) MVA of constant MVA load during the
dynamic simulation; the CMOTOR model will override (10.92 + j4.97) MVA leaving
only 4.08 MW of constant power load. The load at all other buses (bus 150 in this case)
will be assumed to be 50% constant current and 50% constant admittance for dynamics
purposes.
Table 1 summarizes the files used in setting up this example simulation.
Table 1: Files Used in Simulation of Induction Motors Initialized Online
Original system power flow saved case SME1

Power flow saved case as converted and pre-
SMECM
pared for use with CMOTOR model
Dynamic Simulation Raw Data File SMEDM
Dynamics CONEC subroutine CONEC
Dynamics CONET subroutine CONET
Initial condition snapshot SN1M
Figure 20-16 shows the dialog for establishing the converted initial condition power flow
case. Figure 20-17 shows the CMOTOR data sheet filled in to prepare a record for the data
file, SMEDM.
Note the following from the data sheet:
1. The initial motor load is stated only by implication through the initial condition slip,
CON(M+3). This value is arrived at by the use of program IMD. In this case, Figures 20-
5 and 20-6 show that the slip of 0.0085 pu will give a power consumption of 4  3 
(0.910 + j0.415) = 10.92 + j4.97 at unity voltage. This value of slip is accurate enough
because the discrepancy between 11.2 and 10.92 can be taken from the 4.08 MW of
constant power load at bus 151 with no significant effect on the accuracy of the simu-
lation. The exact initial condition power will be placed in VAR(J+5) and VAR(J+6)
during activity STRT.
20-37
2. The load torque at synchronous speed is entered as zero; it will be calculated by CMOTOR
in activity STRT to balance electrical and load torque in the initial condition.
The procedure for setting up the dynamic simulation is straightforward5. Activity DYRE
automatically places the required coordinated calls to CMOTOR and TMOTOR in
CONEC and CONET, respectively. Activity CHAN is used to place these VARs in output
channels with appropriate labels. For convenience VAR(15), which is VAR(J+7), is
placed in channel 4 to allow examination of the trimming capacitance (if any) assigned by
CMOTOR. Motor terminal voltage is monitored by using CHAN to place the voltage at
bus 151 into a channel.
After compiling and linking CONEC and CONET, activity STRT may be used to initialize
the CMOTOR model along with all other models as shown in Figure 20-18. The output
from DOCU shows that the motor’s VARs, J to J+7, have been assigned as VAR(8) through
VAR(15) and hence that slip, P, and Q are available in VARs 9, 13, and 14, respectively.
The output from activity DLST at the end of Figure 20-18 shows the initialized values of
VARs 13, 14, and 15. Because the bus Q load is 4.97 Mvar and the motor reactive power
consumption is 4.923 Mvar (see VAR(14)) at a terminal voltage of 0.99484 (see channel
18), the trimming capacitance simulated by CMOTOR would be zero because no deficit of
vars exists (i.e., the trimming capacitance cannot be negative). Because the actual initial
motor power consumption is 10.81 MW, the constant power consumption of the SCR
device will be simulated as 4.19 MW instead of 4.80 MW, an acceptable deviation.
5 Obsolete; the calls to CMOTOR/TMOTOR must be existing in a pre-PSS®E-26 dynamics

setup.
20-38
ACTIVITY? CASE SME1

Recover original base case
CASE SME1.SAV WAS SAVED ON TUE, MAY 23 1989 13:44
ACTIVITY? CHNG Change load on bus 151 to

ENTER CHANGE CODE: model motor load
10 = IMPEDANCE CORRECTION TABLES 11 = MULTI-TERMINAL DC
DATA: 1
BUS DATA FOR BUS 151 [LOAD 3.30] :

CODE PLOAD QLOAD S H U N T
OLD 1 15.00 7.90 0.00 0.00 CHANGE IT? 1
ENTER CODE, PLOAD, QLOAD, G, B Reassign nonmotor part of reactive
,,4.97,,-2.93
NEW 1 15.00 4.97 0.00 -2.93 load to shunt; new Mvar load is motor
OLD
AREA VOLT
2 0.9947
ANGLE NAME
-3.10 LOAD
BASVLT LOSZON
3.3 1 CHANGE IT? t
Q consumption at initial conditions
.
ENTER BUS NUMBER (0 FOR NEW CHANGE CODE, -1 TO EXIT): -1
ACTIVITY? FNSL Make sure case is solved

.
ENTER ITERATION NUMBER FOR VAR LIMITS
0 FOR IMMEDIATELY, -1 TO IGNORE COMPLETELY: 0
ITER DELTAP BUS DELTAQ BUS DELTA/V/ BUS

DELTAANG BUS
0 0.0000( 200) 0.0003( 151) 0.00015( 201)
0.00003( 201)
1 0.0000( 150) 0.0002( 201)
LARGEST MISMATCH: 0.00 MW -0.02 MVAR 0.02 MVA-BUS

200 [STEAM 33.0]
ACTIVITY? CONL Convert loads at other buses

150
ENTER % CONSTANT I, % CONSTANT G FOR REAL POWER: 50 50

ENTER % CONSTANT I, % CONSTANT B FOR REACTIVE POWER: 50 50 Model assumption for nonmotor
|
LOAD TO BE REPRESENTED AS: loads in this study
REAL REACTIVE
0.00% 0.00% CONSTANT POWER
50.00% 50.00% CONSTANT CURRENT
50.00% 50.00% CONSTANT ADMITTANCE
ENTER 1 IF O.K., 0 OTHERWISE: 1

0
ENTER 1 TO CONVERT LOADS AT REMAINING 1 BUSES: 0
ACTIVITY? CONG Convert generators

GENERATORS CONVERTED
ACTIVITY? SAVE SMECM Save new converted case

CASE SAVED IN FILE SMECM.SAV ON FRI, MAY 26 1989 15:34
as dynamics initial condition
Figure 20-16. Setup of Power Flow Case to Accommodate CMOTOR Load

of 10.92 + j4.97 MVA at Bus 151
20-39
CMOTOR
Double Cage Induction Motor
CALL CMOTOR(I,J,M,K,L,M) from CONEC

CALL TMOTOR(I,J,M,K,L,M) from CONET
This motor is connected at the bus for #_____ I,
which the number is in ICON RA + jLA jL1
using motor equivalent circuit data #_____ J,
stored at CONs
and motor rating data stored at CONs #_____ M,
R1 R2
and with motor speed deviation at #_____ K, jLM ------- ------- + jL2
S S
STATE
and VARs starting with #_____ L,
and using local iteration acceleration #_____ N.
factor at CON Resistance, Impedance on Motor Base MVA
The power flow admittance for this motor is ______ + j

on motor base and ______ + j on system base.
ICON # Description
I 151 Bus number

J 0.005 RA
J+1 0.083 LA
J+2 4.0 LM
J+3 0.04 R1
J+4 0.08 L1
J+5 0.011 R2
J+6 0.05 L2
M 12 Motor base MVA
M+1 0.9 Motor base, H
M+2 2.0 Load damping factor, D
M+3 0.000085 Initial condition slip
M+4 0 Load torque at 1 pu speed;
on motor base
N 0.9 Local iteration acceleration
STATE # Description
K Motor speed deviation
20-40
VARs # Description
L Old speed deviation
L+1 Present speed deviation
L+2 Internal use
L+3 Internal use
L+4 Load torque at 1 pu speed; on
motor base
L+5 Motor P
L+6 Motor Q
L+7 Admittance of initial condition Mvar
shortage
IBUS, ’CMOTOR’, RA, LA, LM, R1, L1, R2, L2, MBASE, H, D, SLIP, TORQUE, ACC/
100
100
'GENSAL' 1 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
'GENSAL' 2 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
SMEDM raw data file
100 'GENSAL' 3 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/ including motor data record
201 'GENROU' 1 6 .05 1 .05 3 0 1.4 1.35 .3 .6 .2 .1 .03 .4/
100 'SCRX' 1 .1 10 200 .05 0 5 0 0/
100 'SCRX' 2 .1 10 200 .05 0 5 0 0/ Data from
previous run
100 'SCRX' 3 .1 10 200 .05 0 5 0 0/
201 'SEXS' 1 .1 .1 100 .1 0 3/
100 'HYGOV' 1 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
100 'HYGOV' 2 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
100
201
'HYGOV' 3 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
'TGOV1' 1 .05 .5 1. .3 1. 1. 0./ Obsolete; CMOTOR will be
151 'CMOTOR' 0.004 0.083 4.0 0.04 0.08 0.011 0.05 12 0.9 2.0 0.0085 0 0.9/
converted to CIM5BL
Figure 20-17. CMOTOR Model Data Sheet as Completed From Figure 20-15
20-41
INITIATED AT DYNAMICS ENTRY POINT ON THU, JUN 01 1989 08:41
ACTIVITY? RSTR SNIM. Recover dynamics setup

SNAPSHOT SNIM.SNP WAS SAVED ON THU, JUN 01 1989 08:38

136 44 17 3 6
ACTIVITY? LOFL
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY? CASE SMECM Initial condition power flow
CASE SMECM.SAV WAS SAVED ON FRI, MAY 26 1989 15:34
ACTIVITY? STRT. Initialize simulation

CMOTOR AT BUS 151--MVA/CURRENT LOAD ARRAYS MODIFIED
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E THU, JUN 01 1989 08:42


100 HYDRO 33.0 1 1.0352 1.7804 4.01 0.48 0.9929 41.39 0.5875 0.5251
100 HYDRO 33.0 2 1.0352 1.7804 4.01 0.48 0.9929 41.39 0.5875 0.5251
100 HYDRO 33.0 3 1.0352 1.7804 4.01 0.48 0.9929 41.39 0.5875 0.5251
201 STEAM 3.30 1 1.0167 1.8531 8.00 2.36 0.9591 40.05 0.6551 0.4938
ENTER CHANNEL OUTPUT FILENAME: GPM Output file of channels used

ENTER SNAPSHOT FILENAME: SNIM.
later for plotting results
CONS STATES VARS ICONS CHANNELS Save the initialized setup
136 44 17 3 6
SNAPSHOT STORED IN FILE SNIM.SNP AT TIME = -0.017
ACTIVITY? DOCU,ALL

6 FOR ALTERNATE SPOOL DEVICE:
CONEC MODELS
*** CALL CMOTOR( 1, 125, 132, 44, 8, 124) ***
BUS NAME BASKV ICON CIRCUIT CON’S RATING CON’S STATE V A R ’ S ACC CON
151 LOAD 3.30 1 125- 131 132- 136 44 8- 15 124
Figure 20-18. Initialization of Simulation Using Model CMOTOR

(Sheet 1 of 2)
20-42
0.004 0.083 4.000 0.040 0.080 0.011 0.050
MOTOR BASE H DAMP INIT SLIP TORQUE ACCEL

12.00 0.900 2.000 0.00850 0.000 0.9000
CONET MODELS
*** CALL TMOTOR( 1, 125, 132, 44, 8, 124) ***
151 LOAD 3.30 1 125- 131 132- 136 44 8- 15 124
ACTIVITY? DLST List CONs and VAR

ENTER OUTPUT DEVICE CODE: initialization by CMOTOR

ENTER OUTPUT CHANNEL RANGE:1 6
CONS:
1: 5.000 2: 0.5000E-01 3: 0.6000E-01 4: 5.084
5: 1.000 6: 1.500 7: 1.200 8: 0.4000
9: 0.2500 10: 0.1200 11: 0.3000E-01 12: 0.2500
125: 0.4000E-02 126: 0.8300E-01 127: 4.000 128: 0.4000E-01

129: 0.8000E-01 130: 0.1100E-01 131: 0.5000E-01 132: 12.00
133: 0.9000 134: 2.000 135: 0.8500E-02 136: 0.0000E+00
VARS:
1: 0.4043E-01 2: 1.000 3: 0.4043E-01 4: 1.000
5: 0.4043E-01 6: 1.000 7: 0.4000E-01 8:-0.8500E-02
9:-0.8500E-02 10: 0.0000E+00 11: 0.0000E+00 12: 0.9121
13: 10.81 14: 4.923 15: 0.0000E+00 16: 1.045
17: 0.9948 18: 0.0000E+00 19: 0.0000E+00 20: 0.0000E+00
STATES:
1: 1.056 2: 0.8919 3: 0.4988 4: 0.0000E+00
5: 0.7224 6: 1.056 7: 0.8919 8: 0.4988
9: 0.0000E+00 10: 0.7224 11: 1.056 12: 0.8919
13: 0.4988 14: 0.0000E+00 15: 0.7224 16: 1.013
17: 0.3092 18: 0.8823 19: 0.5561 20: 0.0000E+00
21: 0.6990 22: 0.7739E-02
ICONS:
1: 151 2: 150 VAR(J+6) contains
motor initial Qelec
OUTPUT CHANNELS: VAR(J+5) contains
motor initial Pelec
#:ADDR 1: 44009 2: 44013 3: 44014 4: 44015 5: 44016
VAR VAR VAR VAR BUS VOLTAGE
9 13 14 15 150
IDENT MOT-SLIP MOT-P MOT-Q TRIM-CAP V-150
VALUE -0.85000E-02 10.807 4.9229 0.00000E+00 1.0450
#:ADDR 6: 44017
BUS VOLTAGE
151 Initial condition voltage VAR(J+7) contains value of capacitor
IDENT V-151 at motor terminals needed to make up excess of motor Q
VALUE 0.99479 load over Q assigned to bus
ACTIVITY? RUN Run simulation
Figure 0-0 (Cont). Initialization of Simulation Using Model CMOTOR (Sheet 2 of 2)
20-43
After CMOTOR has been set up and initialized, simulation runs may be made in the con-
ventional manner. Figure 20-19 shows the effect on the motor of a three-cycle single-phase
fault with effective positive sequence impedance of j0.02 pu at bus 150. The fault reduces
the positive sequence voltage at bus 151 to 0.43 pu, which causes the motor slip to
increase from 0.0085 pu to 0.024 pu. On clearance of the fault, the positive sequence volt-
age at bus 151 returns immediately to 0.70 pu and then returns fairly slowly to nominal
voltage as the motor accelerates back to its initial speed.
The small overshoot in the motor slip curve when the fault is cleared is an aberration intro-
duced by the numerical algorithm of the CMOTOR model; it has no significant effect on
the true result, which may be seen by smoothing through the overshoot.
20.2.6.1.4 Initialization Off-Line

If the motor MVA base is specified as zero in CON(M), the motor is initialized as being
off-line and at standstill. In this case, the value of Tnom, the load torque at synchronous
speed should be specified in CON(M+4) and the initial condition slip need not be speci-
fied because it is set to unity by the model in activity STRT. Initialization of motors off-
line is useful primarily for simulations of motor starting where the voltage dip produced
by the starting of one motor may be sufficient to knock other motors off the line.
The setup of an off-line motor requires no special treatment in the initial condition power
flow case, except for recognition that only one CMOTOR model may be placed at any one
bus, and hence that an initially off-line motor should not be assigned to the same bus as
any other motor whether online or off-line.
Set-up of CMOTOR for off-line initialization is handled by activity DYRE in the same way
as for online motors. Switching the motor onto the bus at which it is assigned is handled
simply by using activity ALTR to change the value of motor base MVA in CON(M) from
zero to the proper value.
20.2.6.1.5 Example — Off-Line Initialization

Consider a case in which the four 2.8 MW motor of the previous example are initially off-
line, the majority of the other load on the system is supplied by the steam turbine unit, and
the four motors are to be started by switching them directly onto the line.
The first step is to expand bus 151 in the power flow case as shown in Figure 20-20 to rep-
resent the three load components individually. The data records for the two off-line motors
can then be prepared as shown in Figure 20-21. The synchronous load torque, CON(M+4),
is specified as 0.9216 pu indicating that each motor must accelerate its full mechanical
load when switched on, as would be the case in starting a large ventilation fan without
shut-off dampers. The synchronous load torque is determined from the rated torque and
the rated speed through the equation for the load torque:
TL = Tnom(1 + n)D
where the load torque, TL, is equal to the rated electrical torque in the steady-state. The value
for rated torque was found from program IMD to be 0.906 at a rated speed of 0.9915, see
Figure 20-5. The synchronous torque is then:
Tnom = TL / (1 + n)D = 0.906/(0.9915)2 = 0.9216
20-44
12:55
POWER
TECHNOLOGIES,
SMALL EXAMPLE
INC.R MOTOR BEHAVIOR FOLLOWING A FAULT AT BUS 150
THU APR 21, 1994

FILE: gpm
FAULT AT BUS 150

CHNL# 6: [V-151]
2.0000 0.0
CHNL# 3: [MOT-Q]
15.000 0.0
CHNL# 2: [MOT-P]
15.000 0.0
CHNL# 1: [MOT-SLIP]
0.0 -0.0500
3.0000
2.7000
2.4000
2.1000
1.8000
TIME (SECONDS)
1.5000
1.2000
0.90000
0.60000
0.30000
0.0
Figure 20-19. Motor Behavior Following Fault at Bus 150
20-45
150
151
15 + j7.9
151 152 153 154 155
3.8 -j2.93
M M M M
2.0 MW 2.0 MW 2.0 MW 2.0 MW
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E FRI APR 29, 1994 11:26
SMALL EXAMPLE SYSTEM BUS DATA
BUS# NAME BSKV COD VOLT ANGLE MVA-LOAD CUR-LOAD Y-LOAD S H U N T AREA ZONE
100 HYDRO 33.0 2 1.0500 0.0 0 0 0 0 0 0 0 0 1 1
150 LOAD 33.0 1 1.0488 -0.1 0 0 2 1 2 -1 0 0 2 1
151 LOAD 3.30 1 0.9860 -2.8 4 0 0 0 0 0 0 -3 2 1
152 MOTOR-1 3.30 1 1.0232 -0.1 0 0 0 0 0 0 0 0 1 1
153 MOTOR-2 3.30 1 1.0232 -0.1 0 0 0 0 0 0 0 0 1 1
154 MOTOR-1 3.30 1 1.0232 -0.1 0 0 0 0 0 0 0 0 1 1
155 MOTOR-1 3.30 1 1.0232 -0.1 0 0 0 0 0 0 0 0 1 1
200 STEAM 33.0 1 1.0511 0.2 0 0 0 0 0 0 0 0 1 1
201 STEAM 3.30 2 0.9990 3.9 0 0 0 0 0 0 0 0 1 1
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E FRI APR 29, 1994 11:26
SMALL EXAMPLE SYSTEM BRANCH DATA
100* 150 1 HYDRO LOAD 0.0100 0.1000 0.0180 1 15 18 20
100* 200 1 HYDRO STEAM 0.0200 0.2000 0.0300 1 15 18 20
150* 151 1 LOAD LOAD 0.0000 1.2500 0.0000 F 1 4 10 15
150* 152 1 LOAD MOTOR-1 0.0000 1.2500 0.0000 F 1 4 10 15
150* 153 1 LOAD MOTOR-2 0.0000 1.2500 0.0000 F 1 4 10 15
150* 154 1 LOAD MOTOR-1 0.0000 1.2500 0.0000 F 1 4 10 15
150* 155 1 LOAD MOTOR-1 0.0000 1.2500 0.0000 F 1 4 10 15
150* 200 1 LOAD STEAM 0.0100 0.1000 0.0180 1 15 18 20
200* 201 1 STEAM STEAM 0.0000 0.8000 0.0000 F 1 10 12 15
Figure 20-20. Expansion of Load Bus 151 into Separate Buses for SCR Load
and Each 2.0-MW Motor for Motor Starting Simulation Using CMOTOR
20-46
CMOTOR
Double Cage Induction Motor
CALL CMOTOR(I,J,M,K,L,M) from CONEC

CALL TMOTOR(I,J,M,K,L,M) from CONET
This motor is connected at the bus for #____ I,
which the number is in ICON RA + jLA jL1
using motor equivalent circuit data stored #____ J,

at CONs
and motor rating data stored at CONs #____ M, R1 R2
jLM ------- ------- + jL2
and with motor speed deviation at STATE #____ K, S S
and VARs starting with #____ L,
and using local iteration acceleration factor #____ N.
at CON Resistance, Impedance on Motor Base MVA
The power flow admittance for this motor is _____ + j on
motor base and ______ + j on system base.
ICON # Description
I Bus number

J 0.004 RA
J+1 0.083 LA
J+2 4.0 LM
J+3 0.04 R1
J+4 0.08 L1
J+5 0.011 R2
J+6 0.05 L2
M 0 Motor base MVA
M+1 0.9 Motor base, H
M+2 2.0 Load damping factor, D
M+3 1.0 Initial condition slip
M+4 0.9216 Load torque at 1 pu speed;
on motor base
N 0.5 Local iteration acceleration
STATE # Description
K Motor speed deviation
20-47
VARs # Description
L Old speed deviation
L+1 Present speed deviation
L+2 Internal use
L+3 Internal use
L+4 Load torque at 1 pu speed; on
motor base
L+5 Motor P
L+6 Motor Q
L+7 Admittance of initial condition Mvar
shortage
IBUS, ’CMOTOR’, RA, LA, LM, R1, L1, R2, L2, MBASE, H, D, SLIP, TORQUE, ACC/
100 'GENSAL' 1 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
100 'GENSAL' 2 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
100 'GENSAL' 3 5 .05 .06 5.084 1 1.5 1.2 .4 .25 .12 .03 .25/
201 'GENROU' 1 6 .05 1 .05 3 0 1.4 1.35 .3 .6 .2 .1 .03 .4/
100 'SCRX' 1 .1 10 200 .05 0 5 0 0/
100 'SCRX' 2 .1 10 200 .05 0 5 0 0/
100 'SCRX' 3 .1 10 200 .05 0 5 0 0/
201 'SEXS' 1 .1 .1 100 .1 0 3/
100 'HYGOV' 1 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/ Motor initially off line
100 'HYGOV' 2 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
100 'HYGOV' 3 .05 .75 8 .05 .5 .2 1 0 1.3 1.1 0 .08/
201 'TGOV1' 1 .05 .5 1. .3 1. 1. 0./
152 'CMOTOR' 0.004 0.083 4.0 0.04 0.08 0.011 0.05 0 0.9 2.0 1.0 0.9216 0.5/ Obsolete; CMOTOR
153 'CMOTOR' 0.004 0.083 4.0 0.04 0.08 0.011 0.05 0 0.9 2.0 1.0 0.9216 0.5/
154 'CMOTOR' 0.004 0.083 4.0 0.04 0.08 0.011 0.05 0 0.9 2.0 1.0 0.9216 0.5/ records will be converted
155 'CMOTOR' 0.004 0.083 4.0 0.04 0.08 0.011 0.05 0 0.9 2.0 1.0 0.9216 0.5/ to CIM5BL records
Figure 20-21. CMOTOR Data for Off-Line Initialization
Figure 20-22 shows the CONEC/CONET subroutines, partial DOCU report, and channel
assignments established in setting up the simulation with activity DYRE and CHAN.
CONEC and CONET have no special logic, and the initial values of motor real and reac-
tive power in VARs 13, 14, 21, and 22 are zero. This case requires the following for
acceptable convergence of the network solutions in dynamic simulations:
1. The induction motor acceleration factor, CON(L), is set to a value between 0.3 and 0.5.
(DYRE used CON(124) as CON(L) for both calls to CMOTOR, hence forcing the same
acceleration for each motor.)
2. The dynamics network solution accelerating factor is set to approximately 0.7 via the
dynamic solution parameters option of activity ALTR.
With the setup established, the dynamic simulation of motor starting is handled simply by
changing CON(132) to switch the motor onto bus 152 and changing CON(144) to switch
on the motor at bus 153. The PSAS input file (PSAS is the Simulation Run Assembler sec-
20-48
tion of PSS®E) to simulate the sequential starting of the motors at a time interval of 10 sec
is shown at the end of Figure 20-22.
SUBROUTINE CONEC
C
$INSERT COMON4
C
CALL CMOTOR( 1, 125, 132, 44, 8, 124)
CALL CMOTOR( 2, 137, 144, 45, 16, 124)
CALL CMOTOR( 3, 149, 156, 46, 24, 124)
CALL CMOTOR( 4, 161, 168, 47, 32, 124)
C
RETURN
END
INTEGER MC,SLOT,IT
SELECT (IT)
FIN
RETURN
END
SUBROUTINE CONET
C
$INSERT COMON4
C
CALL TMOTOR( 1, 125, 132, 44, 8, 124)
CALL TMOTOR( 2, 137, 144, 45, 16, 124)
CALL TMOTOR( 3, 149, 156, 46, 24, 124)
CALL TMOTOR( 4, 161, 168, 47, 32, 124)
C
C
C
C
9000 CONTINUE
C
RETURN
END
Figure 20-22. Setup of Dynamic Simulation for Motor Starting using Model CMOTOR
(Sheet 1 of 3)
20-49
CONEC MODELS
*** CALL CMOTOR( 1, 125, 132, 44, 8, 124) ***
152 MOTOR-1 3.30 1 125- 131 132- 136 44 8- 15 124
0.004 0.083 4.000 0.040 0.080 0.011 0.050

0.00 0.900 2.000 1.00000 0.9216 0.5000
*** CALL CMOTOR( 2, 137, 144, 45, 16, 124) ***
153 MOTOR-2 3.30 2 137- 143 144- 148 45 16- 23 124
0.004 0.083 4.000 0.040 0.080 0.011 0.050

0.00 0.900 2.000 1.00000 0.9216 0.5000
*** CALL CMOTOR( 3, 149, 156, 46, 24, 124) ***
154 MOTOR-3 3.30 3 149- 155 156- 160 46 24- 31 124
0.004 0.083 4.000 0.040 0.080 0.011 0.050

0.00 0.900 2.000 1.00000 0.9216 0.5000
*** CALL CMOTOR( 4, 161, 168, 47, 32, 124) ***
155 MOTOR-4 3.30 4 161- 167 168- 172 47 32- 39 124
0.004 0.083 4.000 0.040 0.080 0.011 0.050

0.00 0.900 2.000 1.00000 0.9216 0.5000
Motors off line CONET MODELS

*** CALL TMOTOR( 1, 125, 132, 44, 8, 124) ***
152 MOTOR-1 3.30 1 125- 131 132- 136 44 8- 15 124
*** CALL TMOTOR( 2, 137, 144, 45, 16, 124) ***
153 MOTOR-2 3.30 2 137- 143 144- 148 45 16- 23 124
Figure 0-0 (Cont). Setup of Dynamic Simulation for Motor Starting Using Model CMOTOR (Sheet 2 of 3)
20-50
*** CALL TMOTOR( 3, 149, 156, 46, 24, 124) ***
154 MOTOR-3 3.30 3 149- 155 156- 160 46 24- 31 124
*** CALL TMOTOR( 4, 161, 168, 47, 32, 124) ***
155 MOTOR-4 3.30 4 161- 167 168- 172 47 32- 39 124
ACTIVITY? DLST

ENTER CON RANGE:

ENTER VAR RANGE:
ENTER STATE RANGE:
ENTER ICON RANGE:
OUTPUT CHANNELS:
#:ADDR 1: 44009 2: 44010 3: 44011 4: 44017 5: 44018

VAR VAR VAR VAR VAR
9 10 11 17 18
IDENT SLIP-1 MOT-1-P MOT-1-Q SLIP-2 MOT-2-P
VALUE -1.0000 0.00000E+00 0.00000E+00 -1.0000 0.00000E+00
#:ADDR 6: 44019 7: 44025 8: 44026 9: 44027 10: 44033

VAR VAR VAR VAR VAR
19 25 26 27 33
IDENT MOT-2-Q SLIP-3 MOT-3-P MOT-3-Q SLIP-4
VALUE 0.00000E+00 -1.0000 0.00000E+00 0.00000E+00 -1.0000
#:ADDR 11: 44034 12: 44035 13: 44040 14: 44041 15: 44042
VAR VAR BUS VOLTAGE BUS VOLTAGE BUS VOLTAGE
34 35 150 151 152
IDENT MOT-4-P MOT-4-Q V-150 V-151 V-152
VALUE 0.00000E+00 0.00000E+00 1.1507 1.0740 1.1227
#:ADDR 16: 44043 17: 44044 18: 44045 19: 16001 20: 16004
BUS VOLTAGE BUS VOLTAGE BUS VOLTAGE FIELD VLTAGE FIELD VLTAGE
153 154 155 BUS 100 MC 1 BUS 201 MC 1
IDENT V-153 V-154 V-155 EFD100-1 EFD201-1
VALUE 1.1227 1.1227 1.1227 1.1522 1.1904
#:ADDR 21: 20001 22: 20004 23: 24001 24: 24004

P MECHANICAL P MECHANICAL SPD DEVIAT’N SPD DEVIAT’N
BUS 100 MC 1 BUS 201 MC 1 BUS 100 MC 1 BUS 201 MC 1
IDENT PMECH100-1 PMECH201-1 SPEED100-1 SPEED201-1
VALUE 0.42843 0.34421 0.00000E+00 0.00000E+00
RECOVER initial conditions from SNIM2 AND SMECM1

INITIALIZE, OUTPUT=GPM2, SNAPSHOT=SNIM2
RUN to 0.1 SECONDS PRINT=25 PLOT=3
SET CON 132 TO 3.0
SET CON 144 TO 3.0 Switch motors 3 and 4 online
SET CON 156 TO 3.0
SET CON 168 TO 3.0 Switch motors 1 and 2 online
Figure 0-0 (Cont). Setup of Dynamic Simulation for Motor-Starting Using Model CMOTOR
(Sheet 3 of 3)
Figures 20-23 and 20-24 show the simulation result. Switching on motors 1 and 2 drops the
voltage at its terminals instantaneously to 0.65 pu; it then droops to 0.58 pu before the
generator voltage regulator action starts to return it to schedule. The motor electrical
power rises first as a function of recovering terminal voltage and then less rapidly as the
voltage hovers around 0.825 pu. As the motor approaches rated speed, the reactive power
20-51
demand drops very rapidly, voltage rises rapidly to a peak of 1.4 pu, and the motor snaps
up to rated speed. Motors 1 and 2 reach rated speed at 5.7 sec. At 10.0 sec, motors 3 and 4
are switched online. Figure 20-24 shows the response of the generators to the motor start-
ing. The bus fed excitation systems on the hydro generators do not reach their 5.0 per-unit
ceiling until t = 2.0 because of depressed bus voltage. The steam-turbine governor
responds to raise turbine power to the maximum of 1.0 pu while the response of the hydro
governors is very sluggish. Both excitation systems drop to zero field voltage briefly as
motor 1 snaps-over the hump of its characteristic and its heavy reactive power demand
disappears.
Figure 20-25 and 20-26 show the behavior of the system when the same two motors are
started in the same way as before, but against 0.1 per-unit load, as would occur in starting
pumps with closed discharge valves. The PSAS input file to produce this changed simula-
tion is:
RECOVER initial conditions FROM SNIM2 AND SMECM1
SET CON 136 TO 0.0906
SET CON 148 TO 0.0906
SET CON 160 TO 0.0906
SET CON 172 TO 0.0906
INITIALIZE, OUTPUT=GPM3, SNAPSHOT=SNIM3
SET CON 132 TO 3.0
SET CON 144 TO 3.0
RUN TO 10.0 SECONDS PRINT=25 PLOT=1
SET CON 156 TO 3.0
SET CON 168 TO 3.0
RUN TO 20.0 SECONDS PRINT=25 PLOT=1
END
The file differs from Figure 20-22 only in setting the CON(M+4) for each motor (CONs
136 and 148) to 0.09216 pu instead of 0.9216 pu. As expected, Figures 20-25 and 20-26
show more rapid acceleration of the motor and a reduced dip in system frequency.
20.2.6.1.6 Simulation of Switching

The CMOTOR model simulates the run-down of the motor if it is tripped off the line.
Tripping is handled by changing the value of CON(M), the motor MVA base, to zero. The
motor may be reconnected at any time after tripping by returning CON(M) to the proper
value. Changes of mechanical load may be handled at any time by appropriately changing
the value of VAR(J+4).
20.2.6.2 Summary of PSS®E Induction Motor Models

Table 2 contains a new summary of the important features for the different induction motor
models available within PSS®E.
20-52
10:59
POWER
TECHNOLOGIES,
INC.R
FRI APR 29, 1994

STARTING MOTOR
FILE: GPM2
CHNL# 3: [MOT-1-Q]
20.000 0.0
CHNL# 2: [MOT-1-P]
10.000 0.0
CHNL# 1: [SLIP-1]
0.0 -2.000
CHNL# 15: [V-152]
2.0000 0.0
21.000
18.900
16.800
14.700
12.600
TIME (SECONDS)
10.500
8.4000
6.3000
4.2000
2.1000
0.0
Figure 20-23. Behavior of Motor 1 – Direct Online Starting Against Full Load
20-53
11:01
POWER
TECHNOLOGIES,
INC.R
FILE: GPM2
FRI APR 29, 1994

GENERATOR BEHAVIOR
CHNL# 22: [PMECH201-1]
2.0000 0.0
CHNL# 21: [PMECH100-1]
1.0000 0.0
CHNL# 20: [EFD201-1]
5.0000 0.0
CHNL# 19: [EFD100-1]
10.000 0.0
CHNL# 23: [SPEED100-1]
0.10000 -0.1000
21.000
18.900
16.800
14.700
12.600
TIME (SECONDS)
10.500
8.4000
6.3000
4.2000
2.1000
0.0
Figure 20-24. Generator Behavior Direct Online Starting Against Full Load
20-54
15:16
POWER
TECHNOLOGIES,
INC.R
FRI APR 29, 1994

FILE: GPM3
MOTOR W/0.1 PU LOAD

CHNL# 3: [MOT-1-Q]
20.000 0.0
CHNL# 2: [MOT-1-P]
10.000 0.0
CHNL# 1: [SLIP-1]
0.0 -2.000
CHNL# 15: [V-152]
2.0000 0.0
21.000
18.900
16.800
14.700
12.600
TIME (SECONDS)
10.500
8.4000
6.3000
4.2000
2.1000
0.0
Figure 20-25. Motor Starting Behavior

as Figure 20-23 Except Motor Starting Against 0.1-pu Load
Figure 20-26. Motor Starting Behavior

as Figure 20-24 Except Motor Starting Against 0.1-pu Load
20-55
Load Characteristic Models

Load Modeling
Table 2: Summary of PSS®E Induction Motor Models
Model
Level of Representation Power Flow Representation Input Data Motor Starting Special Features
Type
Rotating load plus motor electro- Bus load consisting of con- Equivalent circuit Model calculates initial slip.
stant MVA, constant current or data. User sets load
magnetic dynamics. ID in-service Motor MVA base can be
constant admittance.
CIM5 Single or double cage motor: Type 1 or type 2. and specifies assigned as a multiplier of
All load for an individual load positive value MW load at bus.
ID is treated as the motor For single cage:
TLOAD = TNOM(1 + n)D R2 = X2 = 0. for Tnom. Undervoltage relay.
load.
Rotating load plus motor electro- Load consisting of constant Equivalent circuit Model calculates initial slip.
magnetic dynamics. MVA, constant current or con- data.
stant admittance. Not valid for Motor MVA base can be
CIMW Single or double cage motor: Type 1 or type 2. motor starting assigned as a multiplier of
All load for an individual load analysis. MW load at bus.
ID is treated as the motor For single cage:
TLOAD = To(A2 + B + Co + De) R2 = X2 = 0. Undervoltage relay.
20-56
load.
CIMTR2 See CIMTR4. CIMTR4 is recommended for use over CIMTR2.
Operational
Rotating load plus motor electro- impedance and User sets
Machine with negative power
time constants as

magnetic dynamics. output. ZSOURCE must machine in-
derived from type service and
CIMTR4 Single or double cage motor: match X or X’ (single cage). Model calculates initial slip.
1 or 2 equivalent specifies nega-
All load for the machine is circuit. tive value for
TLOAD = TNOM(1 + n)D treated as the motor load.
For single cage: Tnom.
X = T = 0.
Rotating load dynamics with steady- Load consisting of constant Equivalent circuit User sets User must specify initial slip
state electrical characteristics. MVA or constant current. data. MBASE to from which the model deter-
CMOTOR All load IDs attached to this motor MVA mines the required MW.
PSS®E 33.4
Single- or double-cage motor: Type 2.
(Obsolete) bus are considered available base and speci-
For single cage: fies a positive Not valid for use in
to be represented as motor
TLOAD = TNOM(1 + n)D load. R2 = L2 = 999 value for Tnom. extended term simulations.
Program Application Guide: Volume II Load Relay Models
20.3 Load Relay Models
20.3.1 Basic Considerations

The PSS®E model library includes two general groups of load relay models: underfre-
quency and undervoltage load shedding relays. The models provide for the setting of mul-
tiple load shedding stages. For each stage, the user may specify a frequency or voltage
threshold, pickup time, and the fraction of load to shed.
All models have the capability of shedding load for either a CONL-type representation of
load (constant MVA, current and admittance components), or if the load is modeled using
one of the load characteristic models. In the case where the traditional CONL load compo-
nents are present, all models operate identically on the real and reactive parts of the con-
stant MVA, current and admittance components.
The effective value of load as it varies during the simulation may be tracked by assigning
the load to an output channel in activities CHSB or CHAN. Likewise, a total subsystem
load may be monitored by assigning subsystem totals output channels in activity CHSB.
The load shedding relays also set flags (ICONs) that can be then utilized by user-supplied
logic in subroutine CONET. The user-supplied logic can be developed in order to run a
more complex coordination of relays or for additional supervisory action not provided
within a specific model.
20.3.2 Underfrequency Load Shedding Models
20.3.2.1 LDSH Type Models (LDSHBL, LDSHOW, LDSHZN, LDSHAR, LDSHAL)

The LDSH type models (LDSHBL, LDSHOW, LDSHZN, LDSHAR, LDSHAL) represent solid-
state type load-shedding relays. The models disconnect either a fraction of the load at
which the model is applied or sets flags to switch lines, capacitors, etc. with user-supplied
logic when frequency falls below each of its pickup points. The load to be shed at each of
three steps is specified in CON(J+2), CON(J+5), and CON(J+8) as a fraction of the origi-
nal load.
Each of the above loads is reduced to a fraction of its original value, with the fraction
being equal to the value specified for the stage, as corrected to account for load already
shed in the prior stage. For example, if the three load-shedding stages are set to shed 0.3,
0.3, and 0.2 per-unit of the original load, the above load elements are all multiplied by:
1 - 0.3 on the first stage
0.7 - 0.3 on the second stage
0.7
0.4 - 0.2 on the third stage
0.4
When used to set flags, CON(J+2), CON(J+5), and CON(J+8) must be set to zero.
ICON(N+2), ICON(N+5), and ICON(N+8) are set to –1 when the breaker timer times out
for that state. Each stage is triggered only once in a simulation; when set to –1 it stays
there. The time characteristic of the load-shedding relay is illustrated by Figure 20-27.
20-57
Load Relay Models Program Application Guide: Volume II
Frequency
Underfrequency Detector
Resets Instantaneously
Rated Frequency
Load Shedding Threshold
Breaker Opens Shedding Load

Breaker Trip Signal Picks Up
Load Shed
Relay Timer
Starts
Pickup Time for Underfrequency Indication
Time
Figure 20-27. Underfrequency Detection and Load Shedding in LDSH Type Models
The timer for each stage is started when frequency falls below its threshold value and is
reset instantaneously if frequency rises above its threshold. A trip signal is sent to the cir-
cuit breaker if the timer reaches its setting; frequency must have remained below the
threshold for the entire time delay for shedding to occur. Actual shedding of load or setting
of the flag is delayed by the circuit breaker time.
The three shedding stages are independent of one another, and the second stage could
operate before the first if the threshold and time delay settings called for it. The models do
not affect bus connected shunt devices as characterized by the bus shunt data entries
GSHUNT and BSHUNT.
20.3.2.2 LDST Type Models (LDSTBL, LDSTOW, LDSTZN, LDSTAR, LDSTAL)

The LDST type models (LDSTBL, LDSTOW, LDSTZN, LDSTAR, LDSTAL) represent load-
shedding as controlled by a relay with an inverse time characteristic. The models imple-
ment single-stage load-shedding which, when the circuit breaker(s) opens, disconnects a
fraction of the load components. The models utilize an underfrequency detector with an
inverse-time characteristic, defined by Figure 20-28, as distinct from the simple delay char-
acteristic modeled by the LDSH type models.
20-58
1
 =  --Z- dt
Trip when  = 1
Z1 1
d –1 d d 1
------ = ------------- ------ = 0 ------ = ---
dt T reset dt dt Z
Normal Operating Time (Z)
2
Z2
3
Z3 4
Z4
Nominal f f1 f2 f3 f4 Zero
Value reset
Frequency (Hz)
Figure 20-28. Time Inverse Operating Characteristic of LDST Type Models
Points 1, 2, 3, and 4 are specified by CONs J through J+7. The values of the nominal oper-
ating time, Z, are the times that the relay would take to close its trip contact if presented
with a constant input frequency. The actual time taken to close the trip contact in response
to a varying input frequency is given by the integral of the rate of motion of the sensitive
element. The frequency detector does not reset immediately if frequency rises above its
threshold value; it resets in a linear fashion at such a rate that it takes Treset seconds to
reset after just reaching its trip point. If frequency falls below the pickup threshold
(CON(J)) during the resetting process, the relay commences integrating from its partially
reset condition.
The models can also be used just to set a flag that can be used by user supplied logic for
functions such as tripping a line with subroutine LINTRP. CON(J+9) should be set to zero
to disable load tripping. ICON(M) is set to –1 when the timer times out. ICON(M+2) is set
to 1 when the breaker timer times out and stays there.
20.3.2.3 DLSH Type Models (DLSHBL, DLSHOW, DLSHZN, DLSHAR, DLSHAL)

The DLSH type models (DLSHBL, DLSHOW, DLSHZN, DLSHAR, DLSHAL) are an extension
to the LDSH type models to include frequency of decay. Timers will only be active when
20-59
frequency is below the shedding point values and the magnitude of the rate of change of
frequency is above its threshold. Violation of either of the above causes timers to reset.
These models can either shed load or set flags for user-supplied logic. The flags are set
identically to the LDSH type models.
20.3.2.4 LDS3 Type Models (LDS3BL, LDS3OW, LDS3ZN, LDS3AR, LDS3AL)

The LDS3 type models (LDS3BL, LDS3OW, LDS3ZN, LDS3AR, LDS3AL) represent solid-
state load shedding relays with generator transfer trip capability. These models are similar
in function and performance to the LDSH type models. These models add the following
enhancements to the LDSH type models:
1. It allows up to five stages of load to be tripped.
2. This model allows separate circuit breaker times for each stage.
3. This model allows the user to trip the bus connected shunt devices as characterized by
GSHUNT and BSHUNT in same fractions as load.
4. This model allows the user to trip a plant or specific machine if any breakers become
activated. The transfer trip breaker has its own timer.
20.3.3 Undervoltage Load Shedding Models
20.3.3.1 LVSH Type Models (LVSHBL, LVSHOW, LVSHZN, LVSHAR, LVSHAL)

The LVSH type undervoltage load shedding relay models (LVSHBL, LVSHOW, LVSHZN,
LVSHAR, LVSHAL) represent a solid-state type load-shedding relays that disconnect load
based on low voltage. The model has three stages. The fractions of load to be shed are
specified in CON(J+2), CON(J+5) and CON(J+8). The models can also be used to set
flags. For this case, the ICONs are set identically to the LDSH type models.
20.3.3.2 LVS3 Type Models (LVS3BL, LVS3OW, LVS3ZN, LVS3AR, LVS3AL)

The LVS3 type models (LVS3BL, LVS3OW, LVS3ZN, LVS3AR, LVS3AL) are an extension of
the LVSH models:
1. It allows five stages of load, each with its own individual circuit breaker timer.
2. It optionally allows tripping of GSHUNT and BSHUNT in the same fractions as load.
3. It allows the user to specify two branches that may also be tripped. Each branch has
its own circuit breaker timer.
20.3.4 Under voltage and underfrequency load shedding models
20.3.4.1 UVUF Type Models (UVUFBL, UVUFOW, UVUFZN, UVUFAR, UVUFAL)

The UVUF type models (UVUFBLU1, UVUFOWU1, UVUFZNU1, UVUFARU1, UVUFALU1)
represent solid-state type load-shedding relays. The models disconnect a fraction of each
of the component of the composite load namely static load, electronic load, 3-phase induc-
tion motor load, single-phase air conditioner load based on the applicable undervoltage or
underfrequency setting. The model increases the feeder impedance in reverse proportion
20-60
to the load shedding fraction and reduces the feeder compensation as a direct proportion of
the load shedding fraction to simulate the tripping of an equivalent fraction of the feeder
from the substation. The substation transformer and substation shunt values do not change
due to load shedding. The model has 3 voltage-based trip settings and 3 frequency-based
trip settings.
20-61
20-62
Chapter 21
Generic Wind Models
All varieties of available wind turbines can be conventionally split into several types. Cur-
Chapter 21 - Generic Wind Models
rently, four types have been suggested as follows:

• Type 1. Direct connected Conventional Induction Generator
• Type 2. Wound rotor Induction Generator with Variable Rotor Resistance
• Type 3. Doubly-Fed Induction Generator
• Type 4. Full Size Converter Unit
The description of generic models in this chapter correspond to the above classification.
Generic Wind Models are designed to be used in studies related to the integration of Wind
Turbine Generators (WTG) in an Electrical Power System.
The generic WTG models (WT1, WT2, WT3, and WT4) were not developed with the
intent of being accurate for the study of frequency excursions. Furthermore, generic mod-
els were not designed to reproduce the behavior of advanced power management features
such as programmed intertia and "spinning reserve" by spilling wind.
Standard PSS®E CONET models VTGDCA and FRQDCA can be used to simulate char-
acteristics of the voltage and frequency protection systems.
Note: The representation of implicit transformers is not allowed for any of these four
types of wind generators.
21.1 Generic WT1 (Type 1) Model
21.1.1 Introduction
The WT1 PSS®E wind turbine stability model was developed to simulate the performance
of a wind turbine employing a conventional induction generator directly connected to the
grid.
21-1
Generic Wind Models PSS®E 33.4
Generic WT1 (Type 1) Model Program Application Guide: Volume II
Figure 21-1. Conventional Directly-Connected Induction Generator
21.1.2 Power Flow Setup

A wind turbine generator in power flow is treated as related to a wind machine category of
the existing generator record of the Power Flow Raw Data File. The wind control mode
should be set as 3, i.e., a wind machine that operates at a fixed power, with the machine’s
reactive power output and reactive power upper and lower limits all equal and set based on
the specified power factor and on the machine’s active power setting (PG on the data
record). Negative value of the power factor must be specified and is interpreted as a lead-
ing power factor (i.e., the wind machine produces active power and absorbs reactive
power).
All issues of making an equivalent of a wind farm to be modeled is the users’ responsibil-
ity. They should make a decision on how many original units will be lumped into an
equivalent machine presented in the power flow case. For N lumped machine they need to
multiply MBASE of the original machine by N.
They also have to take care of the adequate equivalent of the wind farm feeders, collec-
tors, and step-up transformers.
A vast majority of wind turbines with this type of a machine includes a set of compensat-
ing capacitors to keep the power factor in steady state within the given range. These
capacitors should be added in a fixed shunt data record.
21-2
PSS®E 33.4 Generic Wind Models
Program Application Guide: Volume II Generic WT1 (Type 1) Model
21.1.3 Dynamic Setup

PSS®E WT1 Generic Wind Model comprises models as follows
• WT1G: generator/converter model
• WT12T: wind turbine model
• WT12A: pseudo governor model
Figure 21-2 shows the interaction between these models.
Figure 21-2. WT1 Connectivity Diagram
The generator model WT1G is based on the standard PSS®E model of the induction gen-
erator CIMTR3. This model takes into account the rotor flux dynamics and can be used for
single cage or double cage machines. At initialization this model calculates the reactive
power consumption of the machine Qact at given terminal voltage and MW-dispatch. It
places a “hidden” shunt on the machine terminal bus, with the size equal to a difference
between Qgen from the power flow and Qact.
The turbine model WT12T uses the two-mass representation of the wind turbine shaft
drive train. It calculates the speed deviations of the rotor on the machine and on the blade
sides. By setting the turbine inertia fraction Htfrac = 0 the model can be switched to a con-
ventional single mass representation.
21-3
Figure 21-3. WT12T Two-Mass Shaft model
The pseudo governor model WT12A is an attempt to simplify and generalize calculation
of the aerodynamic torque. This model was designed and developed after thorough inves-
tigation of aerodynamic characteristics and pitch control of several vendor specific wind
turbines. Finally the arrangement shown below was suggested.
21-4
Figure 21-4. Pseudo-Governor Model
The model uses two inputs, one in terms of the blade rotor speed deviation and another in
terms of the real power at the machine terminals. These two inputs combined together are
processed by a PI controller with non-wind-up limits. The filtered output is the mechanical
torque on the rotor blade side that is used by the WT12T model.
The example dynamic data input file is provided below, along with the example documen-
tation of the dynamic parameters.
5 'WT1G1' 1
0.84600 0.0000 3.9270 0.17730 0.0000
0.10000 1.0000 0.30000E-01 1.2000 0.17900 /
5 'WT12T1' 1
5.3000 0.0000 0.91800 5.0000 1.0000 /
5 'WT12A1' 1
0.1500E-01 0.1000 0.1500E-01 0.1000 0.1000
0.1000 0.9000 0.2500 /
Figure 21-5. Example Dynamic Data Input File, WT1
To mitigate sustained increase in power output for a sustained drop in frequency, Limax
should be set to the rated power on the machine MVA base. Typically, Limax = 0.9 on the
machine MVA base at rated power output. For instance, if the generator is rated at
111MVA and it is dispatched at 100 MW, Limax should be 0.9 (100/111). This baseline
behavior is consistent with the assumption that the wind turbine is operating at rated
power output, and wind speed remains constant.
21-5
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS(R)E THU, AUG 14 2008 11:14

WECC WGMG TEST SYSTEM
SCR=20 FULL POWER
WIND MODELS
REPORT FOR ALL MODELS BUS 5 [CLR_1 0.6000] MODELS
** WT1G1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S

5 CLR_1 0.6000 1 3-12 3-7 1-3
MBASE Z S O R C E
111.1 0.00000+J 0.17729
T' T'' X X' X'' XL

0.846 0.000 3.93 0.1773 0.0000 0.1000
E1 S(E1) E2 S(E2)
1.0000 0.0300 1.2000 0.1790
** WT12T1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S

5 CLR_1 0.6000 1 13-17 8-11 4-6
H DAMP Htfrac Freq1 DSHAFT

5.3000 0.0000 0.9180 5.0000 1.0000
** WT12A1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S

5 CLR_1 0.6000 1 18-25 12-15 7-8
DROOP KP TI T1 T2 TPE LIMMAX LIMMIN

0.015 0.100 0.050 0.100 0.100 0.100 0.900 0.250
Figure 21-6. Example of Documentation, WT1
21.2.1 Introduction
The WT2 PSS®E wind turbine stability model was developed to simulate performance of
a wind turbine employing a wound rotor induction generator with the variable rotor resis-
tance control.
21-6
Figure 21-7. Wound Rotor Induction Generator with Variable Rotor Resistance Control

should be set as 3, i.e. a wind machine that operates at a fixed power, with the machine’s
reactive power output and reactive power upper and lower limits all equal and set based on
the specified power factor and on the machine’s active power setting (PG on the data
record). Negative value of the power factor must be specified and is interpreted as a lead-
ing power factor (i.e., the wind machine produces active power and absorbs reactive
power).
A vast majority of wind turbines with this type of a machine includes a set of compensat-
ing capacitors to keep the power factor in steady state within the given range. These
capacitors should be added in a fixed shunt data record.
21-7

PSS®E WT2 Generic Wind Model comprises models as follows
• WT2E: electrical control model
• WT12T: wind turbine model
• WT12A: pseudo governor model
Pseudo
Governor
Model
The generator model WT2G is based on the standard PSS®E model of the induction gen-
erator CIMTR3. This model takes into account the rotor flux dynamics and can be used for
single cage or double cage machines. At initialization, this model calculates the reactive
power consumption of the machine Qact at given terminal voltage and MW-dispatch. It
places a “hidden” shunt on the machine terminal bus, with the size equal to a difference
between Qgen from the power flow and Qact. It also determines what portion of available
external rotor resistance should be added to fit the steady-state operating point.
During the simulation, the value of the external rotor resistance is calculated by the electri-
cal control model WT2E (Figure 21-9). This model uses the machine rotor speed and elec-
trical power as inputs and calculates the portion of the available rotor external resistance
to be added to the internal rotor resistance.
21-8
Figure 21-9. WT2E Electrical Control model
The turbine model WT12T uses the two-mass representation of the wind turbine shaft
drive train. It calculates the speed deviations of the rotor on the machine and on the blade
sides. By setting the turbine inertia fraction Htfrac = 0 the model can be switched to a con-
ventional single mass representation. A diagram of this model is shown in Figure 21-3.
The pseudo-governor model WT12A is an attempt to simplify and generalize calculation
of the aerodynamic torque. A diagram of this model is shown in Figure 21-4.
The model uses two inputs, one in terms of the blade rotor speed deviation and another in
terms of the real power at the machine terminals. These two inputs combined together are
processed by a PI controller with non-wind-up limits. The filtered output is the mechanical
torque on the rotor blade side that is used by the WT12T model.
21-9
5 ' WT2G1' 1
0.12602 6.8399 0.18084 0.44190E-02 0.10994 1.0000 0.0000
1.2000 0.0000 0.0000 0.21700E-01 0.89880 0.90000 0.90500
0.0000 0.54000E-02 0.20000E-01 0.40000E-01 0.10000 /
5 ' WT2E1' 1
0.5000E-01 0.5000E-01 1.000 1.000 0.9900 0.5000E-01 /
5 ' WT12T1' 1
3.4600 0.0000 0.81000 1.5000 0.30000 /
5 ' WT12A1' 1
0.1500E-01 20.00 1.000 .1000 .1000 .1000 1.000 .2500 /
Figure 21-10. Example Dynamic Data Input File, WT2
21-10
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS(R)E FRI, AUG 22 2008 13:38

TEST WT2
WIND MODELS
REPORT FOR WIND MECH. MODELS AT ALL BUSES
** WT2G1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S ICON

5 CLR_1 0.6000 1 3-21 3-5 1-3 1
MBASE Z S O R C E
112.0 0.00000+J 0.30220
XA XM X1 R_ROT_MACH R_ROT_MAX
0.126 6.840 0.18 0.0044 0.1099
E1 S(E1) E2 S(E2)
1.0000 0.0000 1.2000 0.0000
POWER_REF_1 POWER_REF_2 POWER_REF_3 POWER_REF_4 POWER_REF_5

0.0000 0.0217 0.8988 0.9000 0.9050
SLIP_1 SLIP_2 SLIP_3 SLIP_4 SLIP_5

0.0000 0.0054 0.0200 0.0400 0.1000
** WT2E1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S ICON

5 CLR_1 0.6000 1 22-27 6-8 2
Tsp Tpe Ti Kp ROTRV_MAX ROTRV_MIN

0.050 0.050 1.000 1.000 0.990 0.050
** WT12T1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S

5 CLR_1 0.6000 1 28-32 9-12 4-6
H DAMP Htfrac Freq1 DSHAFT

3.4600 0.0000 0.8100 1.5000 0.3000
** WT12A1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S

5 CLR_1 0.6000 1 33-40 13-16 7-8
DROOP KP TI T1 T2 TPE LIMMAX LIMMIN

0.015 20.000 1.000 0.100 0.100 0.100 1.000 0.250
Figure 21-11. Example of Documentation, WT2
21-11
21.3.1 Introduction
The WT3 PSS®E wind turbine stability model was developed to simulate performance of
a wind turbine employing a doubly fed induction generator (DFIG) with the active control
by a power converter connected to the rotor terminals.
Figure 21-12. Doubly Fed Induction Generator
with the Active Control by a Power Converter Connected to the Rotor Terminals
The model is to be used in studies related to the integration of Wind Turbine Generators
(WTG) in an Electrical Power System.

should be set as 2, i.e., a wind machine that controls a remote bus voltage within the given
range of reactive power [Qmin, Qmax].
21-12

PSS®E WT3 Generic Wind Model comprises models as follows:
• WT3E: electrical control model
• WT3T: mechanical control (wind turbine) model
• WT3P: pitch control model.
There are two different generator/converter models available, namely WT3G1 and
WT3G2.
The WT3G2 model, which is recommended for new dynamic setups, includes improve-
ments in the original WT3G1. The original WT3G1 model is being retained for reasons of
backward compatibility.
The electrical control model WT3E1 can be used with WT3G1 as well as with the
improved WT3G2 models. When WT3E1 is used with the WT3G1 model, it is recom-
mended that ICON(M+2) be set to 1, and when used with WT3G2 model, the ICON be
set to 2.
Figure 21-13. Interaction among Generic Wind Models
21-13
The example dynamic data input files are provided below, along with the example docu-
mentation of the dynamic parameters.
3. Xeq = Imaginary (ZSORCE)
Figure 21-14. WT3G1 Model
21-14
Figure 21-15. WT3G2 Model
21-15
Figure 21-16. WT3E1 Model
21-16
Figure 21-17. WT3T1 Model
21-17
Figure 21-18. WT3P1 Model
5 ’WT3G1’ 1
67 0.80000 30.000 0.0000 0.10000 1.5000 /
5 ’WT3E1’ 1 5 1 1 3 2 ’1 ’
0.15000 18.000 5.0000 0.0000 0.50000E-01
3.0000 0.60000 1.1200 0.10000 0.29600
-0.43600 1.1000 0.50000E-01 0.45000 -0.45000
5.0000 0.50000E-01 0.90000 1.2000 40.000
-0.50000 0.40000 0.50000E-01 0.50000E-01 1.0000
0.69000 0.78000 0.98000 1.1200 0.74000
1.2000 /
5 ’WT3T1’ 1
1.2500 4.9500 0.0000 0.70000E-02 21.980
0.0000 1.8000 1.5000 /
5 ’WT3P1’ 1
0.30000 150.00 25.000 3.0000 30.000 0.0000 27.000 10.000 1.0000 /
Figure 21-19. Example Dynamic Data Input File (Set 1) with Reference to
GE 1.5 MW Wind Turbine
21-18
5 'WT3G2' 1 30
0.02000 0.02000 0.0000 0.0000 0.10000
1.5000 0.50000 0.9000 1.1100 1.2000
2.0000 5.0000 0.20000E-01 /
5 'WT3E1' 1 5 1 2 3 2 '1'
0.15000 18.000 5.0000 0.0000 0.50000E-01
3.0000 0.60000 1.1200 0.40000E-01 0.43600
-0.43600 1.1000 0.20000E-01 0.45000 -0.45000
5.0000 0.10000 0.90000 1.1000 40.000
0.50000 1.45000 0.50000E-01 0.50000E-01 1.0000
0.30000 0.69000 0.78000 0.98000 0.74000
1.2000 /
5 'WT3T1' 1
1.2500 4.9500 0.0000 0.70000E-02 21.980
0.0000 1.8000 1.5000 /
5 'WT3P1' 1
0.30000 150.00 25.000 3.0000 30.000
0.0000 27.000 10.000 1.0000 /
Figure 21-20. Example Dynamic Data Input File (Set 2) with Reference to
GE 1.5 MW Wind Turbine
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS(R)E MON, DEC 29 2008 8:14

GE BENCHMARK SYSTEM
TEST OF GE-1.5/3.6 MODEL
WIND MODELS
REPORT FOR WIND MODELS AT ALL BUSES
** WT3G1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S ICON

5 WT 0.5750 1 1-5 1-4 1-4 1
XEQ Kpll Kipll PLLMX Prated
0.8000 30.0000 0.0000 0.1000 1.5000
** WT3E1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S I C O N S

5 WT 0.5750 1 6-36 5-14 5-11 2-7
TFV KPV KIV XC TFP KPP KIP

0.1500 18.0000 5.0000 0.0000 0.0500 3.0000 0.6000
PMX PMN QMX QMN IPMAX TRV RPMX

1.1200 0.1000 0.2960 -0.4360 1.1000 0.0500 0.4500
RPMN T_POWER KQi

-0.4500 5.0000 0.0500
Figure 21-21. Example of Documentation (Set 1) with Reference to GE 1.5 MW Wind Turbine
21-19
VMINCL VMAXCL Kqv XIQmin XIQmax Tv

0.9000 1.2000 40.0000 -0.5000 0.4000 0.0500
Tp Fn Wpmin Wp20
0.0500 1.0000 0.6900 0.7800
Wp40 Wp60 Pwp Wp100

0.9800 1.1200 0.7400 1.2000
REMOTE CONTROLLED BUS # 5
VARFLG = 1 VLTFLG = 1
FROM BUS = 3, TO BUS = 2, ID = ’1 ’
** WT3T1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S ICON

5 WT 0.5750 1 37-44 15-18 12-16 8
Vw H DAMP Kaero Theta2

1.2500 4.9500 0.0000 0.0070 21.9800
Htfrac Freq1 DSHAFT

0.0000 1.8000 1.5000
** WT3P1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S ICON

5 WT 0.5750 1 45-53 19-21 9
Tp Kpp Kip Kpc Kic

0.3000 150.0000 25.0000 3.0000 30.0000
TetaMin TetaMax RTetaMax PMX

0.0000 27.0000 10.0000 1.0000
Figure 21-21. (cont.) Example of Documentation (Set 1) with Reference to GE

1.5 MW Wind Turbine
21-20
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS(R)E MON, DEC 29 2008 8:14

GE BENCHMARK SYSTEM
TEST OF GE-1.5/3.6 MODEL
WIND MODELS
REPORT FOR ALL MODELS AT ALL BUSES BUS 5 [WT 0.5750] MODELS
** WT3G2 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S VAR ICON

5 WT 0.5750 1 3-15 3-7 1 1
Tiqcmd Tipcmd Kpll Kipll PLLMX

0.2000 0.0000 0.0000 0.0000 0.1000
Prated VLVPL1 VLVPL2 GLVPL

1.5000 0.5000 0.9000 1.0000
VHVRCR CURHVRCR Rip_LVPL T_LVPL

1.2000 2.0000 5.0000 0.0200
Number of Lumped Wind Machines: 67

5 WT 0.5750 1 16-46 8-17 2-8 2-7
TFV KPV KIV XC TFP KPP KIP

0.1500 18.0000 5.0000 0.0000 0.0500 3.0000 0.6000
PMX PMN QMX QMN IPMAX TRV RPMX

1.1200 0.0400 0.4360 -0.4360 1.1000 0.0200 0.4500
RPMN T_POWER KQi

-0.4500 5.0000 0.1000
VMINCL VMAXCL Kqv XIQmin XIQmax Tv

0.9000 1.1000 40.0000 0.5000 1.5000 0.0500
Tp Fn Wpmin Wp20
0.0500 1.0000 0.3000 0.6900
Wp40 Wp60 Pwp Wp100

0.7800 0.9800 0.7400 1.2000
REMOTE CONTROLLED BUS # 5
VARFLG = 1 VLTFLG = 2
FROM BUS = 3, TO BUS = 2, ID = ’1 ’
Figure 21-22. Example of Documentation (Set 2) with Reference to GE 1.5 MW Wind Turbine
21-21
** WT3T1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S ICON

5 WT 0.5750 1 47-54 18-21 9-13 8
Vw H DAMP Kaero Theta2

1.2500 4.9500 0.0000 0.0070 21.9800
Htfrac Freq1 DSHAFT

0.0000 1.8000 1.5000
** WT3P1 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S ICON

5 WT 0.5750 1 55-63 22-24 9
Tp Kpp Kip Kpc Kic

0.3000 150.0000 25.0000 3.0000 30.0000
TetaMin TetaMax RTetaMax PMX

0.0000 27.0000 10.0000 1.0000
Figure 0-0 (Cont). Example of Documentation (Set 2) with Reference to GE 1.5 MW Wind Turbine
21.4.1 Introduction
The WT4 PSS®E wind turbine dynamic stability model was developed to simulate perfor-
mance of a wind turbine employing a generator connected to the grid via the power con-
verter.
21-22
Figure 21-23. A Generator Connected to the Grid via the Power Converter

A wind turbine generator is decoupled from the grid by a power converter that is actually
connected to the grid. In power flow a power converter is treated as a generator related to
a wind machine category of the existing generator record of the Power Flow Raw Data
File. The wind control mode should be set as 2, i.e., a wind machine that controls a remote
bus voltage within the given range [Qmin; Qmax] of reactive power.
All issues of making an equivalent of a wind farm to be modeled are the user’s responsi-
bility. A decision must be made for the number of original units to be lumped into an
equivalent machine presented in the power flow case. For N lumped machines, multiply
MBASE of the original machine by N.
The user must also define an adequate equivalent of the wind farm feeders, collectors, and
step-up transformers.

PSS®E WT4 Generic Wind Model comprises two modules as follows
• WT4G: power converter/generator module
• WT4E: electrical control module
21-23
Figure 21-24 shows the interaction between these modules.
The power converter/generator module (Figure 21-25) calculates the current injection to
the grid based on filtered active and reactive power commands from the electrical control
module. Both components of the injected current are processed under the high/low voltage
conditions by means of a special logic.
21-24
Figure 21-25. WT4 Generator/Converter Module
21-25
The converter control module includes reactive and active power controls.
The reactive control calculates the reactive current command for the various control
options, which could be any of the following:
• Remote bus voltage control
• Power factor control
• Reactive power control
The active power control is based on the concept that a machine does not need to be simu-
lated. No matter how the active power control is implemented and what criteria it uses,
this control is responsible for keeping the power balance between the machine and the grid
injection.
In the suggested model (Figure 21-26, lower part), the active power control compares the
active power injected to the grid against the power reference and changes the active com-
ponent of the injected current accordingly. The Converter Current Limiter is an essential
part of the model. Depending on whether active or reactive priority was selected, it uses
different algorithms to update limits of active and reactive components of the converter
current. The example dynamic data input file for WT4G1 and WT4E1 models, with refer-
ence to the GE 2.5 MW wind turbine, are provided in Figure 21-27 and Figure 21-28.
21-26
Figure 21-26. WT4E Electrical Control Module
21-27
5 'WT4G1' 1 0.20000E-01 0.20000E-01 0.40000 0.90000

1.1100 1.2000 2.0000 2.0000 0.20000E-01 /
5 'WT4E1' 1 5 0 1 0
0.15000 18.000 5.0000 0.50000E-01 0.10000
0.0000 0.80000E-01 0.47000 -0.47000 1.1000
0.0000 0.50000 -0.50000 0.50000E-01 0.10000
0.90000 1.1000 120.00 0.50000E-01 0.50000E-01
1.7000 1.1100 1.1100 /
Figure 21-27. Example of Dynamic Data Input File for WT4 with reference to the GE 2.5 MW
wind turbine
21-28
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS(R)E TUE, DEC 02 2008 13:23

TEST SYSTEM
TEST OF WT4 MODEL
WIND MODELS
REPORT FOR ALL MODELS AT ALL BUSES BUS 5 [WT 0.5750] MODELS

5 WT 0.5750 1 3-11 3-5 1-3
TIqCmd TIpCmd VLVPL1 VLVPL2 GLVPL

0.0200 0.0200 0.4000 0.9000 1.1100
VHVRCR CURHVRCR RIp_LVPL T_LVPL

1.2000 2.0000 2.0000 0.0200

5 WT 0.5750 1 12-34 6-15 6-9 1-4
TFV KPV KIV KPP

0.1500 18.0000 5.0000 0.0500
KIP Kf Tf QMX QMN IPMAX TRV

0.1000 0.0000 0.0800 0.4700 -0.4700 1.1000 0.0000
dPMX dPMN T_POWER KQi VMINCL VMAXCL Kvi

0.5000 -0.5000 0.0500 0.1000 0.9000 1.1000 120.0000
Tv Tp ImaxTD Iph1 Iqh1

0.0500 0.0500 1.7000 1.1100 1.1100
ICON(M) - Remote controlled Bus : 5

ICON(M+1) - PFAFLG = 0 (1 if PF fast control is enabled; else 0)
ICON(M+2) - VARFLG = 1 (1 if Qcmd is provided by WindVar; else 0)
ICON(M+3) - PQFLAG = 0 (1 for P priority, =0 for Q priority)
Figure 21-28. Example of Documentation for WT4 with reference to the GE 2.5 MW wind
turbine
21-29
In the course of validation studies for Siemens 2.3 and 3.6 MW Wind Turbines, another set of the
WT4 generator/converter and electrical control models was developed:WT4G2 and WT4E2. These
two models should be used together; neither of them can be used with either WT4G1 or WT4E1.
The main difference between the existing Type 4 and the new set of models is that the converter
current limiter takes into consideration such factors as:
• P, Q converter capability curves.

Note: In load flow, the reactive power output should be set inside the capability curve;
otherwise it will experience an immediate change at the model initialization.
• The effect of the DC link connection between the machine side converter and the grid
side converter.
• Restriction on the magnitude of the internal voltage generated by the grid side
converter.
• Constraints of the internal smoothing reactor.
Several model parameters are used to achieve the above.The recommended values of these
parameters are suggested in example files.
To improve the network solution at abrupt voltage changes (applying or removing a fault), the reac-
tive component of the converter current limit is filtered by a lag with one cycle time constant.
To avoid instability, potential ringing (or hunting), caused by the voltage crossing fault ride-through
(FRT) threshold on the way down and back up, the FRT hysteresis is implemented.
Example parameters for WT4G2 and WT4E2 with the reference to the SMK-213 2.3 MW wind
turbine are provided below.
20 ’WT4G2’ 1
0.0100 0.0020 0.4000 -0.10000 1.1100 1.2500
2.0000 2.0000 0.02000 /
20 ’WT4E2’ 1
20 0 1 0
0.00000 15.000 2.0000 0.08000 0.0100 0.0000
0.08000 1.0000 -1.20000 1.10000 0.0000 0.50000
-0.50000 0.05000 0.0100 0.87500 1.1250 55.0000
0.05000 0.0500 1.1150 1.25000 1.0850 0.0200
0.87500 0.0500 0.0000 2.00000 1.0000 1.0000
1.00000 1.0000 1.1000 0.14154 /
Figure 21-29. Example of Dynamic Data Input File for WT4 with reference to the Siemens 2.3
MW wind turbine
21-30
WIND MODELS
REPORT FOR ALL MODELS AT ALL BUSES BUS 20 [WT4 0.6900] MODELS

20 WT4 0.6900 1 3-11 3-5 1-5
TIqCmd TIpCmd VLVPL1 VLVPL2 GLVPL

0.0100 0.0200 0.4000 -0.1000 1.1100
VHVRCR CURHVRCR RIp_LVPL T_LVPL

1.2500 2.0000 2.0000 0.0200
** WT4E2 ** BUS X-- NAME --X BASEKV MC C O N S S T A T E S V A R S

I C O N S
20 WT4 0.6900 1 12-45 6-16 6-9 1-4
TFV KPV KIV KPP

0.0000 15.0000 2.0000 0.0800
KIP Kf Tf QMX QMN IPMAX TRV

0.0100 0.0000 0.0800 1.0000 -1.2000 1.1000 0.0000
dPMX dPMN T_POWER KQi VMINCL VMAXCL Kvi

0.5000 -0.5000 0.0500 0.0100 0.8750 1.1250 55.0000
Tv Tp ImaxTD Iph1 Iqh1

0.0500 0.0500 1.1150 1.2500 1.0850
Tiqf FRT_Thres FRT_Hys FRT_Droop FRT_Iq_Gain

0.0200 0.8750 0.0500 0.0000 2.0000
Max_FRT_Iq IQMax_Fact1 IQMax_Fact2

1.0000 1.0000 1.0000
DC_Link_Droop VinvMax0 NBR_X

1.0000 1.1000 0.1415
ICON(M) - Remote controlled Bus : 20

ICON(M+1) - PFAFLG = 0 (1 if PF fast control is enabled; else 0)
ICON(M+2) - VARFLG = 1 (1 if Qcmd is provided by WindVar; else 0)
ICON(M+3) - PQFLAG = 0 (1 for P priority, =0 for Q priority)
Figure 21-30. Example of Documentation for WT4 with reference to the Siemens 2.3 MW
wind turbine
21-31
21-32
Chapter 22
Photovoltaic (PV) System
Model
Chapter 22 - Photovoltaic (PV) System Model
22.1 Introduction
The PSS®E Solar PV Unit dynamic stability model was developed to simulate perfor-
mance of a photovoltaic (PV) plant connected to the grid via a power converter. The
model is largely based on the generic type 4 wind model, WT4, with the added ability to
simulate output changes due to solar irradiation.
22.1.1 Load flow Representation

PV panels are decoupled from the grid by a power converter which is actually connected
to the grid.
The wind control mode should be set as 2, i.e. a wind machine which controls a remote
bus voltage within the given range [Qmin; Qmax] of reactive power. As for load flow
models of most power electronic devices, the source reactance of this machine should be
set as infinite: XSORCE = 99999.
All issues of making an equivalent of a PV plant to be modeled is the users' responsibility.
They should make a decision on how many original panels and converters will be lumped
into an equivalent machine presented in the load flow case. For N lumped converters they
need to multiply MBASE of the original converter by N.
They also have to take care of the adequate equivalent of the plant feeders, collectors, and
step-up transformers.

The PV Generic Wind Model comprises the following modules:
• PVGU: power converter/generator module
• PVEU: electrical control module
• PANEL: linearized model of a panel's output curve
• IRRAD: linearized solar irradiance profile
22-1
Photovoltaic (PV) System Model PSS®E 33.4
Introduction Program Application Guide: Volume II
Figure 22-1 shows the interaction between these modules.
Figure 22-1. PV Model Connectivity Diagram
The modules are conventionally designated as wind modules with PVGU as the genera-
tor/converter module, PVEU as the electrical control module, PANEL as the mechanical
module, and IRRAD as the pitch module.
The power converter/generator module calculates the current injection to the grid based on
filtered active and reactive power commands from the electrical control module.
The converter control module includes reactive and active power controls.
The reactive control calculates the reactive current command for the various control
options, which could be any of the following:
• Remote bus voltage control
• Power factor control
• Reactive Power control
The active power control compares the active power injected to the grid against the power
reference, VAR (L+3), and changes the active component of the injected current respec-
tively. The power reference is controlled by the amount of DC power coming from the
PANEL module.
As converter/generator and electrical control modules are very close to respective mod-
ules of the generic WT4 wind model, please refer to the corresponding documentation
given in PSS®E 32 Program Application Guide Volume II.
The panel module calculates the DC power from the PV plant at a given irradiance level.
The user enters the maximum DC power a panel can produce at standard irradiance levels,
which is readily available from a PV manufacturer's I-P curves.
The irrad module allows the user to enter an irradiance profile in the form of up to ten data
points (time, irradiance level) as CONs. At each simulation time step, the module will cal-
culate a linearized irradiance level. The irradiance level is initialized based on the steady-
state power output.
The user has an option of disabling the "IRRAD" module by changing the especially des-
ignated flag.
22-2
PSS®E 33.4 Photovoltaic (PV) System Model
Program Application Guide: Volume II Example Dynamic Input Data
22.2 Example Dynamic Input Data

5 'USRMDL' 1 'PVGU1' 101 1 0 9 3 3
0.20000E-01 0.20000E-01 0.40000 0.90000
1.1100 1.2000 2.0000 2.0000 0.20000E-01 /
5 'USRMDL' 1 'PVEU1' 102 0 4 24 10 4

5 0 1 0
0.15000 18.000 5.0000 0.50000E-01 0.10000
0.0000 0.80000E-01 0.47000 -0.47000 1.1000
0.0000 0.50000 -0.50000 0.50000E-01 0.10000
0.90000 1.1000 120.00 0.50000E-01
0.50000E-01
1.7000 1.1100 1.1100 10.0/
5 'USRMDL' 1 'PANELU1' 103 0 0 5 0 1

0.16 0.38 0.59 0.85 1 /b
5 'USRMDL' 1 'IRRADU1' 104 0 1 20 0 1
1
5 1000 10 900 15 850 20 800 25 700
30 600 35 700 0 0 0 0 0 0 /
22-3
Photovoltaic (PV) System Model PSS®E 33.4
Example Dynamic Input Data Program Application Guide: Volume II
22-4
Chapter 23
Elementary Blocks for
Handling Transfer Functions in
Dynamic Models
Chapter 23 - Elementary Blocks for Handling Transfer Functions in Dynamic Models
23.1 General
Transfer functions are commonly used in dynamic models. The various transfer function
types that are commonly encountered can be classified as follows:
• Integrator
• Integrator with non-windup limit
• First order (lag)
• First order (lag) with non-windup limit
• Washout
• Lead-lag
• Lead-lag with non-windup limit
• Proportional-integrator (PI)
• PI with non-windup limit
• Proportional-integral-derivative (PID) block
• PID with non-windup limit
• Second Order block
One common approach in handling the transfer function blocks is to write the state equa-
tions. In PSS®E implementation, the dynamic models calculate the derivative of the states
(called DSTATE), which are then used to calculate the state variables (called STATE in
PSS®E) using the Modified Euler integration method.
The computation of DSTATEs involves writing out the appropriate equations. Because
there could be several ways of formulating the STATE equations, the expressions for ini-
tializing the STATE variables, and the equations involved in obtaining the DSTATE could
also vary. In addition, implementation of non-windup limits could also prove to be tricky.
23-1
Elementary Blocks for Handling Transfer Functions in Dynamic Models PSS®E 33.4
General Program Application Guide: Volume II
In order to provide for an easy and a consistent method of handling of transfer function
STATE and DSTATE equations along with the associated non-windup limits, the concept
of ‘Elementary Blocks) was introduced in PSS®E. The elementary blocks are simply a
library of functions (provided with PSS®E) that can be invoked in dynamic models to ini-
tialize the model STATE, for the calculation of DSTATE, to impose non-windup limits (if
any), and to calculate the transfer block output.
The description given below illustrates the use of the elementary block FORTRAN func-
tions for all three MODEs for each transfer function configuration.
Some general points about the use of elementary block functions are as follows:
• For each transfer function type given above, there are a set of three FORTRAN func-
tions – one for PSS®E MODE 1 for initialization of STATE, one for MODE 2 for the
calculation of DSTATE; applying non-windup limits (if any); and setting the transfer
function block output, and one for MODE 3 for setting the transfer function block output
using the updated STATE variable following numerical integration, which in turn is used
to calculate the model outputs. Thus, for modeling (say) an integrator without non-
windup limits, the functions that would be invoked in the PSS®E dynamic models are:
INT_MODE1 (for MODE 1), INT_MODE2 (for MODE 2), and INT_MODE3 (for MODE
3). The corresponding functions for an integrator with non-windup limits are
NWINT_MODE1, NWINT_MODE2, and NWINT_MODE3.
• The elementary blocks are implemented as FUNCTION calls (rather than SUBROU-
TINE), meaning that the block input or output is the name of the FUNCTION itself.
• The name of the function to be invoked will depend on the transfer function type, and
the PSS®E MODE (1,2 or 3) for which the calculations are being applied.
• The functions of the MODE1 type returns error codes (IERR). The IERR codes
returned back by the MODE1 calls for each transfer function are given in description
below. Although not mandatory, the model code in turn could have logic to sense the
IERR value and write out appropriate error messages on the progress window. IERR
value of zero indicates absence of any error condition.
• One of the input arguments for all the MODE1 functions (except for the washout
transfer function) is the block output. For the washout block, because the block output
in steady state is zero, the block input is instead specified as one of the input argu-
ments. This is used by the MODE1 function for the calculation of initial value of the
corresponding STATE variable.
• One of the input arguments to the MODE2 and the MODE3 type functions is the block
input. This is used by the MODE2 and MODE3 functions to calculate the DSTATE (in
MODE 2), and calculate the block output.
• The inputs ‘U’ and ‘Y’ shown in the figures are generic designation for block inputs and
outputs respectively. These have to be calculated and set in the model code. Thus, in
the MODE1 function, the block output ‘Y’ is calculated in the model code, and passed
as an argument (VOUT) in the MODE1 function call. The MODE1 function call returns
back the block input. Similarly, for the MODE2 and MODE3 functions, the block input
‘U’ is calculated in the model code and passed as an argument (VINP) in the MODE2
and MODE3 function calls. The MODE2 and MODE3 function calls return the block
output.
• The MODE1, MODE2, and the MODE3 functions all require as inputs the STATE index
of the block for which the function is invoked. Although in the description given below,
23-2
PSS®E 33.4 Elementary Blocks for Handling Transfer Functions in Dynamic Models
Program Application Guide: Volume II Integrator
this index is shown as ‘K’ (if the block has only one STATE, and as K, K1 etc., if the
block has more than one STATE), the appropriate STATE index (in general K+n, where
n could be any number greater than or equal zero) has to be specified.
• The MODE1, MODE2, and the MODE3 functions all require as inputs the model con-
stants (e.g., gains, time constants, maximum and minimum limits etc.), which are
usually in the PSS®E ‘CON’ arrays. The appropriate ‘CON’ values have to be passed
as arguments in the function call.
• The elementary block functions are accessible to user-written models via the INCLUDE
of ‘COMON4.INS’.
In summary, the use of the elementary blocks would greatly simplify the process of writ-
ing PSS®E dynamic models, and in addition, the application of non-windup limits would
be consistent with the IEEE recommendation.
23.2 Integrator
Figure 23-1. Integrator Block
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-2. Variable Declaration for Integrator Block
C_
C____ UNLIMITED INTEGRATOR (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y ! block output
VINP = INT_MODE1( TI, ! time constant (real)
# VOUT, ! block output variable (real)
# K , ! index for state variable (integer)
# IERR ) ! error code (integer)
C IERR = 0 => no error

C IERR = 4 => TI is zero
Figure 23-3. FORTRAN Code for Initialization of Integrator Block (MODE=1)
23-3
Non-windup Integrator Program Application Guide: Volume II
C_
C____ UNLIMITED INTEGRATOR
C_
VINP = U
VOUT = INT_MODE2( TI, ! time constant (real)
# VINP, ! block input variable (real)
# K ) ! index for state variable (integer)
Figure 23-4. FORTRAN Code for Calculation of Derivatives of Integrator Block (MODE=2)
C_
C____ UNLIMITED INTEGRATOR
C_
VINP = U
VOUT = INT_MODE3( TI, ! time constant (real)
Figure 23-5. FORTRAN Code for Calculation of Output Integrator Block (MODE=3)
23.3 Non-windup Integrator
Figure 23-6. Integrator Block with Non-Windup Limits
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-7. Variable Declaration for Integrator Block with Non-Windup Limits
23-4
Program Application Guide: Volume II Non-windup Integrator
C_
C____ NON-WINDUP INTEGRATOR (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y
VINP = NWINT_MODE1( TI, ! time constant (real)
# VRMAX, ! max. limit (real)
# VRMIN, ! min. limit (real)
# K, ! index for state variable (integer)
C
C IERR = 4 => TI is zero
Figure 23-8. FORTRAN Code for Initialization of Integrator Block

with Non-Windup Limits (MODE=1)
C_
C____ NON-WINDUP INTEGRATOR
C_
VINP = U
VOUT = NWINT_MODE2( TI, ! time constant (real)
Figure 23-9. FORTRAN Code for Calculation of Derivatives of Integrator Block

C_
C____ NON-WINDUP INTEGRATOR
C_
VINP = U
VOUT = NWINT_MODE3( TI, ! time constant (real)
Figure 23-10. FORTRAN Code for Calculation of Output Integrator Block

23-5
First Order (Lag) Block Program Application Guide: Volume II
23.4 First Order (Lag) Block
Figure 23-11. First Order Block
Figure 23-12. PSS®E Implementation of the First Order Block
Figure 23-13. PSS®E Implementation of the First Order Block when T = 0
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-14. Variable Declaration for First Order Block
23-6
Program Application Guide: Volume II First Order (Lag) Block
C_
C____ UNLIMITED FIRST ORDER BLOCK (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y
VINP = LAG_MODE1( KI, ! gain (real)
# TI, ! time constant (real)
# K, ! index of state variable (integer)
C
C IERR = 3 => Gain KI = 0 (fatal error at initialization)
C IERR = 4 => TI = 0
Figure 23-15. FORTRAN Code for Initialization of First Order Block (MODE=1)
C_
C____ UNLIMITED FIRST ORDER LAG
C_
VINP = U
VOUT = LAG_MODE2( KI, ! gain (real)
# K ) ! index of state variable (integer)
Figure 23-16. FORTRAN Code for Calculation of Derivatives of First Order Block (MODE=2)
C_
C____ UNLIMITED FIRST ORDER LAG
C_
VINP = U
VOUT = LAG_MODE3( KI, ! gain (real)
Figure 23-17. FORTRAN Code for Calculation of Output of First Order Block (MODE=3)
23-7
First Order (Lag) Block with Non-Windup Limits Program Application Guide: Volume II
23.5 First Order (Lag) Block with Non-Windup Limits
Figure 23-18. First Order Block with Non-Windup Limits

with Non-Windup Limits

with Non-Windup Limits when T = 0
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-21. Variable Declaration for First Order Block

23-8
Program Application Guide: Volume II First Order (Lag) Block with Non-Windup Limits
C_
C____ NON-WINDUP FIRST ORDER LAG BLOCK
C_
VOUT = Y
VINP = NWLAG_MODE1( KI, ! gain (real)
C
C IERR = 1 => initialization above VRMAX
C IERR = 2 => initialization below VRMIN
C IERR = 3 => Gain KI = 0 (fatal error at initialization)
C IERR = 4 => TI = 0
Figure 23-22. FORTRAN Code for Initialization of First Order Block

C_
C____ NON-WINDUP FIRST ORDER LAG
C_
VINP = U
VOUT = NWLAG_MODE2(KI, ! gain (real)
Figure 23-23. FORTRAN Code for Calculation of Derivatives of First Order Block
C_
C____ NON-WINDUP FIRST ORDER LAG
C_
VINP = U
VOUT = NWLAG_MODE3(KI, ! gain (real)
Figure 23-24. FORTRAN Code for Calculation of Output of First Order Block
23-9
Wash-Out Block Program Application Guide: Volume II
23.6 Wash-Out Block
Figure 23-25. Wash-Out Block
Figure 23-26. PSS®E Implementation of the Wash-Out Block
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-27. Variable Declaration for Wash-Out Block
C_
C____ WASHOUT BLOCK
C_
VINP = U
VOUT = WSHOUT_MODE1( KI, ! gain (real),
C
C IERR = 4 => TI = 0
Figure 23-28. FORTRAN Code for Initialization of Wash-Out Block (MODE=1)
23-10
Program Application Guide: Volume II Lead-Lag Block
C_
C____ WASH-OUT
C_
VINP = U
VOUT = WSHOUT_MODE2( KI, ! gain (real)
Figure 23-29. FORTRAN Code for Calculation of Derivatives of Wash-Out Block (MODE=2)
C_
C____ WASH-OUT
C_
VINP = U
VOUT = WSHOUT_MODE3( KI, ! gain (real)
Figure 23-30. FORTRAN Code for Calculation of Output of Wash-Out Block (MODE=3)
23.7 Lead-Lag Block
Figure 23-31. Lead-Lag Block
23-11
Lead-Lag Block Program Application Guide: Volume II
Figure 23-32. PSS®E Implementation of the Lead-Lag Block
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-33. Variable Declaration for Lead-Lag Block
C_
C____ UNLIMITED LEAD-LAG BLOCK
C_
VOUT = Y
VINP = LDLG_MODE1( KI, ! gain (real)
# T1, ! numerator (lead) time constant (real)
# T2, ! denominator (lag) time constant (real)
C
C IERR = 3 => Gain KI = 0
C IERR = 4 => T2 = 0 (Dynamics of the block is ignored)
C IERR = 5 => T1 = 0 (Block treated as first order lag)
Figure 23-34. FORTRAN Code for Initialization of Lead-Lag Block (MODE=1)
23-12
Program Application Guide: Volume II Lead-Lag Block with Non-Windup Limits
C_
C____ UNLIMITED LEAD-LAG BLOCK
C_
VINP = U
VOUT = LDLG_MODE2( KI, ! gain (real)
Figure 23-35. FORTRAN Code for Calculation of Derivatives of Lead-Lag Block (MODE=2)
C_
C____ UNLIMITED LEAD-LAG
C_
VINP = U
VOUT = LDLG_MODE3( KI, ! gain (real)
Figure 23-36. FORTRAN Code for Calculation of Output of Lead-Lag Block (MODE=3)
23.8 Lead-Lag Block with Non-Windup Limits
Figure 23-37. Lead-Lag Block with Non-Windup Limits
23-13
Lead-Lag Block with Non-Windup Limits Program Application Guide: Volume II
Figure 23-38. PSS®E Implementation of the Lead-Lag Block

INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-39. Variable Declaration for Lead-Lag Block

C_
C____ NON-WINDUP LEAD-LAG BLOCK (INITIALIZED WITH OUTPUT = VERR = VINP)
C_
VOUT = Y
VINP = NWLDLG_MODE1( KI, ! gain (real)
C
C IERR = 3 => Gain KI = 0
C IERR = 4 => T2 = 0 (Dynamics of the block is ignored)
C IERR = 5 => T1 = 0 (Block treated as first order lag)
Figure 23-40. FORTRAN Code for Initialization of Lead-Lag Block

23-14
Program Application Guide: Volume II Proportional-Integral (PI) Block
C_
C____ NON-WINDUP LEAD-LAG BLOCK
C_
VINP = U
VOUT = NWLDLG_MODE2( KI, ! gain (real)
# K) ! index of state variable (integer)
Figure 23-41. FORTRAN Code for Calculation of Derivatives of Lead-Lag Block

C_
C____ NON-WINDUP LEAD-LAG BLOCK
C_
VINP = U
VOUT = NWLDLG_MODE3( KI, ! gain (real)
Figure 23-42. FORTRAN Code for Calculation of Output of Lead-Lag Block

23.9 Proportional-Integral (PI) Block
Figure 23-43. Proportional-Integral (PI) Block
23-15
Proportional-Integral (PI) Block Program Application Guide: Volume II
Figure 23-44. PSS®E Implementation of the Proportional-Integral (PI) Block
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-45. Variable Declaration for Proportional-Integral (PI) Block
C_
C____ UNLIMITED PI BLOCK (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y
VINP = PI_MODE1( KP, ! proportional gain (real)
# KI, ! integral gain (real)
# K, ! index of state variable (real)
C
C IERR = 3 => Gain KP = 0 (fatal error at initialization)
Figure 23-46. FORTRAN Code for Initialization of Proportional-Integral (PI) Block (MODE=1)
C_
C____ UNLIMITED PI
C_
VINP = U
VOUT = PI_MODE2( KP, ! proportional gain (real)
Figure 23-47. FORTRAN Code for Calculation of Derivatives of Proportional-Integral (PI)

Block (MODE=2)
23-16
Program Application Guide: Volume II Proportional-Integral (PI) Block with Non-Windup Limits
C_
C____ UNLIMITED PI
C_
VINP = U
VOUT = PI_MODE3( KP, ! proportional gain (real)
Figure 23-48. FORTRAN Code for Calculation of Output of Proportional-Integral (PI) Block
(MODE=3)
23.10 Proportional-Integral (PI) Block with Non-Windup Limits
Figure 23-49. Proportional-Integral (PI) Block with Non-Windup Limits
Figure 23-50. PSS®E Implementation of the Proportional-Integral (PI) Block

23-17
Proportional-Integral (PI) Block with Non-Windup Limits Program Application Guide: Volume II
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-51. Variable Declaration for Proportional-Integral (PI) Block

C_
C____ NON-WINDUP PI BLOCK (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y
VINP = NWPI_MODE1( KP, ! proportional gain (real)
C
Figure 23-52. FORTRAN Code for Initialization of Proportional-Integral (PI) Block

C_
C____ NON-WINDUP PI
C_
VINP = U
VOUT = NWPI_MODE2( KP, ! proportional gain (real)
Figure 23-53. FORTRAN Code for Calculation of Derivatives of Proportional-Integral (PI)

Block with Non-Windup Limits (MODE=2)
23-18
Program Application Guide: Volume II Proportional-Integral-Derivative (PID) Block
C_
C____ NON-WINDUP PI
C_
VINP = U
VOUT = NWPI_MODE3( KP, ! proportional gain (real)
Figure 23-54. FORTRAN Code for Calculation of Output of Proportional-Integral (PI) Block
23.11 Proportional-Integral-Derivative (PID) Block
Figure 23-55. Proportional-Integral-Derivative (PID) Block
Figure 23-56. PSS®E Implementation of the Proportional-Integral-Derivative (PID) Block
23-19
Proportional-Integral-Derivative (PID) Block Program Application Guide: Volume II
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-57. Variable Declaration for Proportional-Integral-Derivative (PID) Block
C_
C____ UNLIMITED PID BLOCK (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y
VINP = PID_MODE1( KP, ! proportional gain (real)
# KD, ! derivative gain (real)
# TD, ! time constant of the derivative channel (real)
# K, ! index of integral state variable (integer)
# K1, ! index of derivative state variable (integer)

Figure 23-58. FORTRAN Code for Initialization of Proportional-Integral-Derivative (PID)

Block (MODE=1)
C_
C____ UNLIMITED PID
C_
VINP = U
VOUT = PID_MODE2( KP, ! proportional gain (real)
# K1 ) ! index of derivative state variable (integer)
Figure 23-59. FORTRAN Code for Calculation of Derivatives of Proportional-Integral-

Derivative (PID) Block (MODE=2)
23-20
Program Application Guide: Volume IIProportional-Integral-Derivative (PID) Block with Non-Wind-
C_
C____ UNLIMITED PID
C_
VINP = U
VOUT = PID_MODE3( KP, ! proportional gain (real)
# K1 ) ! index of derivative state variable (integer)
Figure 23-60. FORTRAN Code for Calculation of Output of Proportional-Integral-Derivative

(PID) Block (MODE=3)
23.12 Proportional-Integral-Derivative (PID) Block with Non-

Windup Limits
Figure 23-61. Proportional-Integral-Derivative (PID) Block with Non-Windup Limits
23-21
Proportional-Integral-Derivative (PID) Block with Non-Windup Limits Program Application Guide:
Figure 23-62. PSS®E Implementation of the Proportional-Integral-Derivative (PID) Block

INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-63. Variable Declaration for Proportional-Integral-Derivative (PID) Block

23-22
Program Application Guide: Volume IIProportional-Integral-Derivative (PID) Block with Non-Wind-
C_
C____ NONWINDUP PID BLOCK (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y
VINP = NWPID_MODE1( KP, ! proportional gain (real)
# TD, ! time constant of derivative channel (real)
# K1, ! index of derivative state variable (integer)
C
Figure 23-64. FORTRAN Code for Initialization of Proportional-Integral-Derivative (PID)

Block with Non-Windup Limits (MODE=1)
C_
C____ NON-WINDUP PID
C_
VINP = U
VOUT = NWPID_MODE2( KP, ! proportional gain (real)
# K1) ! index of derivative state variable (integer)
Figure 23-65. FORTRAN Code for Calculation of Derivatives of Proportional-Integral-

Derivative (PID) Block with Non-Windup Limits (MODE=2)
23-23
Second Order Block Program Application Guide: Volume II
C_
C____ NON-WINDUP PID
C_
VINP = U
VOUT = NWPID_MODE3( KP, ! proportional gain (real)
# K1) ! index of derivative state variable (integer)
Figure 23-66. FORTRAN Code for Calculation of Output of Proportional-Integral-Derivative

(PID) Block with Non-Windup Limits (MODE=3)
23.13 Second Order Block
Figure 23-67. Second Order Block
Figure 23-68. PSS®E Implementation of the Second Order Block
23-24
Program Application Guide: Volume II Second Order Block
INTEGER IERR
REAL VINP
REAL VOUT
Figure 23-69. Variable Declaration for Second Order Block
C_
C____ 2ND ORDER BLOCK (INITIALIZED WITH OUTPUT = Y)
C_
VOUT = Y
VINP = ORD2_MODE1( A, ! parameter A (real)
# B, ! parameter B (real)
# C, ! parameter C (real)
# D, ! parameter D (real)
# E, ! parameter E (real)
# F, ! parameter F (real)
# K, ! index of first state variable (integer)
# K1, ! index of second state variable (integer)
C
C IERR = 3 => D = 0 (dynamics of block is ignored)
C IERR = 4 => C = 0
Figure 23-70. FORTRAN Code for Initialization of Second Order Block (MODE=1)
C_
C____ SECOND ORDER BLOCK
C_
VINP = U
VOUT = ORD2_MODE2( A, ! parameter A (real)
# K1) ! index of second state variable (integer)
Figure 23-71. FORTRAN Code for Calculation of Derivatives of Second Order Block
(MODE=2)
23-25
Second Order Block Program Application Guide: Volume II
C____ SECOND ORDER BLOCK

C_
VINP = U
VOUT = ORD2_MODE3( A, ! parameter A (real)
# K1) ! index of second state variable (integer)
Figure 23-72. FORTRAN Code for Calculation of Output of Second Order Block (MODE=3)
23-26
Chapter 24
Other Models
Chapter 24 - Other Models
24.1 Saturating Reactive Load
24.1.1 Model SAT1

SAT1 is a very simplified model of a saturating reactive device such as a saturated-reactor
form of static-var-generator. SAT1 models no device control dynamics or inherent device
dynamics; it simply models fundamental frequency inductive current consumption as a
piecewise linear function of bus voltage. The initialization procedure is shown in
Figure 24-1. The shape of the current voltage curve is specified by the value of current,
CRO, drawn at the knee, and by the slope, R, of the saturated portion. The voltage at the
knee, VON, is assumed to be an adjustable parameter. (In a real device being modeled by
SAT1, adjustment might occur by tap changing.) The value of VON (VAR(K)) is deter-
mined in activity STRT so that the SAT1 model exactly replaces an equivalent linear reac-
tor.
24-1
Other Models PSS®E 33.4
Saturating Reactive Load Program Application Guide: Volume II
Voltage (pu) Equivalent Linear

Initial Condition from Reactor Characteristic
Power Flow Solution
Vi R SAT1
1 Characteristic
VON
VON =  1 – R V i + R  CRO
CRO Current on
Reactor Base (pu)
Figure 24-1. Initialization of SAT1 Model Knee Voltage
The initial condition power flow case must be solved with a linear reactor, where the rat-
ing must be equal to the rating (CON(J+2)) of the saturating device. This solution estab-
lishes the initial operating point of the saturating device, and hence the value of VON.
Because the SAT1 model replaces the linear reactor, the linear reactor must be removed
from the bus by setting BSHUNT to zero before execution of activity FACT in setting up
the initial condition network model. The power flow case must be solved before the linear
reactor is removed and must not be resolved by any power flow activity after its removal.
An initial condition saved case should be made after removal of the linear reactor from the
bus. The bus at which SAT1 is connected should have no other shunt device connected to
it, but may have a constant admittance component of load. Activity TYSL must not be used
after the linear reactor has been removed from the power flow case. The initialization
sequence for the SAT1 model should be:
CHNG, etc. to place linear reactor on bus
SOLV, etc. to solve power flow case
CONG
CONL as required
ORDR
CHNG to remove linear reactor
SAVE, name
followed, in dynamic simulation, by:

RSTR to recover dynamics data
LOFL
CASE, name
24-2
PSS®E 33.4 Other Models
Program Application Guide: Volume II Models Requiring User-Supplied Logic
FACT
RTRN
STRT
Because it usually has a very rapid variation of reactive current with voltage, SAT1 usu-
ally requires a local deceleration of the iterative network solution. Accordingly, the value
of CON(L) will usually need to be substantially less than unity. A reasonable first trial
value is 0.5 but it is not unusual to need an acceleration factor as small as 0.05 to obtain
good convergence. Typical values of the reactor regulation and knee current might be 0.05
and 0.025, respectively. The saturating device may be disconnected by setting ICON(I) to
zero.
24.1.2 Model SAT2

SAT2, a transformer saturating model, adds an inductive shunt current onto the sum of load
and other currents drawn from a bus. SAT2 should be applied at the bus on the supply side
of the transformer where the magnetizing current is to be modeled. It is permissible to
make multiple applications of SAT2 at a bus if several transformers are connected to it.
The CON values specified for SAT2 must specify the magnetizing current, per unit, of
transformer rated current and the transformer rating must be specified as CON(N).
SAT2 does not initialize itself and so the current drawn by it shows up as a mismatch at the
bus if no manual precautions are taken to recognize it. Such manual action is usually
unnecessary because transformer magnetizing current is usually very small in relation to
load currents at the normal operating voltages existing in the majority of initial conditions.
It is usually sufficient to apply SAT2 to a completely conventional initial condition power
flow setup and to tolerate a small shift in the voltages and reactive power flows due to the
addition of magnetizing current in activity STRT. The network solution in activity STRT
should converge in three or four iterations when SAT2 is used, compared to one iteration
in a normal situation.
SAT2 is de-activated by setting ICON(I) to zero; this should be done when the transformer
with which it is associated is de-energized. SAT2 frequently requires local deceleration of
the network solution. It is not unusual to require a value as small as 0.1 for the accelerating
factor, CON(L), particularly in cases of radial load rejection where the transformer is at
the opened end of a transmission line.
24.2 Models Requiring User-Supplied Logic

The PSS®E model library includes some models that are intended to be called only in
response to some user-supplied logic. This logic may be as simple as the following:
IF (LOGICAL) CALL XXX
in which the user wanted the model to be called at a time equal to 0 sec. This logic may
also be quite complicated if the user wanted to use specific bus or machine quantities (see
Section 19.9, Advanced Uses of CONEC and CONET of the PSS®E Program Operation Man-
ual on Advance Uses of CONEC and CONET). The user usually will not want these mod-
els called by activity STRT or when TIME is less than zero. These models should be called
24-3
Model NETFRQ Program Application Guide: Volume II
when MODE is equal to three. Because the models will take no action during the t+ calls
to subroutines CONEC and CONET. Therefore, the user should include the following as
part of the logic to all the models:
KPAUSE.NE.2
24.2.1 Model LINTRP

This line tripping model (LINTRP), when called, will immediately trip the line described
by its ICONs. When the line is tripped, a message is printed. If the line is not found, an
appropriate message is printed.
24.2.2 Model BSDSCN

BSDSCN, when called, disconnects a bus (makes it Type 4) by tripping all the circuits con-
nected to it, and identifies the bus and circuits in the printout. If the bus cannot be found,
an appropriate message is printed.
24.2.3 Model GENTRP

GENTRP, when called, immediately trips the generator (sets its states to zero). When it
trips the generator, an appropriate message is printed. If the machine cannot be found, an
appropriate message is printed.
24.3 Model NETFRQ

NETFRQ, using selection from dialog opened at Dynamics>Simulation>Simulation Options…,
causes the network parameters to become frequency-dependent. This model should be
called whenever frequency substantially deviates from nominal. It is not automatically
called by PSS®E because it may add 10 to 20% to the running time of a stability case.
Users should verify this themselves and, if running time is not a problem, always include
this model. When the model is called, the following items become frequency-dependent:
1. Line reactances and capacitances.
2. Generator source reactances.
3. Shunts including switched shunts and static var devices.
4. Reactive loads modeled on constant impedance.
The user will typically not see any difference in results during the first swing of a simula-
tion when using this model. Comparing subsequent swings will show better damping
when the model is used (i.e., not using frequency dependence gives conservative results).
PSS®E calculates frequency at each bus independently by taking the instantaneous rate of
change of angle and placing it through a filter time constant. The default value for the time
constant is 0.04 sec. It is this filtered frequency that is used in CMOTOR, CLOAD, etc.
when frequency dependence is invoked. On a few rare instances where the system is very
weak, no generator sources electrically close. This filter constant may be too short and
cause numerical instability. The user should first confirm that the NETFRQ model is caus-
24-4
Program Application Guide: Volume II System-Wide Monitoring Models
ing the numerical instability by deactivating the model and rerunning the simulation
before increasing this filter constant via the solution parameters choice in ALTR.
24.4 System-Wide Monitoring Models

There are several PSS®E models that do not affect the solution at all. These models are
provided mostly for output purposes to aid the user in interpreting results. All the models
are recognized either by activity DYRE, by selection from dialog opened at Dynamics>Simu-
lation>Simulation Options…, or by selection in activity CHSB.
24.4.1 Simulation Option Monitoring Models
24.4.1.1 Model RELANG

PSS®E automatically places the absolute machine angles in the array ANGLE for output
purposes. Model RELANG allows the user to choose a reference angle and then modifies
the ANGLE array to contain all machine angles relative to the reference. The following
choices for the reference are set by dialog:
• Machine average angle.
• Weighted average angle (machine MBASE is used for weighing).
• Bus number and machine identifier.
The model is ignored and does print an appropriate message if the bus or machine is not
found. This model is extremely useful in smaller systems where a fault would cause the
absolute angles of all machines to accelerate so that no good comparison of angles can be
made. Care should be taken when using average angles in trying to find intermachine
oscillations during dynamics stability cases (several seconds) because the average may fil-
ter or introduce new modes.
24.4.1.2 Model RELAY1

RELAY1 is a relay scanning model that checks every circuit from both ends in the system
against the relay characteristics shown in Figure 24-2. This model only notes circuits
within zones; it does not trip any circuits. The model is very good for getting a sense of
what happens on the circuits without having to place a relay on every circuit.
24.4.1.3 Model VSCAN

This model scans to see if any bus voltages are outside the user-specified range and prints
appropriate messages. The model helps to locate potential trouble locations.
24-5
System-Wide Monitoring Models Program Application Guide: Volume II
– 1 X line
 = tan -------------
C R line
2 2
Z line = R +X
B
Center of A 0.46  Z line
A Center of B 0.75  Z line
 Center of C 1.10  Z line
Figure 24-2. Standard RELAY1 and RELAY2 Relay Circles
24.4.1.4 Model OSSCAN

The out-of-step scanning model (OSSCAN) can be used to scan an entire network for out-
of-step conditions. With this model, whenever the apparent impedance as seen from either
end of a line is less than the line impedance, the line is flagged as out of step. OSSCAN is
useful in exploratory simulations, e.g., it is not evident where or if a system will lose syn-
chronism, or the user wishes to avoid placing detailed relay models throughout the system.
The model can be used either to just monitor or to monitor and immediately trip. The
model is activated from the Dynamics > Simulation > Simulation Options… dialog.
24.4.1.5 Model GNSCN1

The generator scanning and tripping model (GNSCN1), which is called by selecting
Dynamics > Simulation > Simulation Options… and specifying the option Scan for generators
exceeding angle threshold from the dialog, scans all generator angles to see if any exceed
the system average angle by more than CON(J) degrees. When it finds a machine with a
large deviation, it trips it immediately and prints an appropriate message. This model is
valuable in doing preliminary studies in which generators lose synchronism and the user
wants to see what effect tripping the generator will have on the system. However, this
model should not be used if the user has multiple ac systems or ac systems connected only
by dc transmission.
24.4.1.6 Model GNSCN2

The generator scanning and tripping model (GNSCN2) is called by selecting Dynamics >
Simulation > Simulation Options… and specifying the option Scan for generators exceeding
power unbalance threshold from the dialog. This model scans all generator angles and
trips immediately any machine where instantaneous per-unit mechanical power exceeds
its per-unit electrical power by more than the threshold, CON(J). The model prints a mes-
sage indicating which generators have been tripped. The value of this model is question-
able and it should be used sparingly.
24-6
Program Application Guide: Volume II System-Wide Monitoring Models
24.4.1.7 Model VIOLSCN

This model checks the bus voltages during dynamic simulation fault recovery and reports
voltage violations if any.
There are two possible voltage violation checks - voltage recovery and voltage dip. These
are collectively called the Voltage Violation check. The two checks can be selected inde-
pendently.
The voltage recovery may have primary voltage criteria (i.e., voltage to recover above
threshold V1 faster than t1 seconds after fault clearing) and secondary voltage criteria
(i.e., voltage to recover above threshold V2 faster than t2 seconds after fault clearing).
The voltage dip check will be based on voltage threshold V3 and time t3 (seconds). Once
voltage has recovered above threshold V3, it should not dip below that value for longer
than t3.
For definitions of V1, V2, V3, and t1, t2 and t3, refer to Figure 24-3.
Figure 24-3. Voltage Violation Check
The dialog for selection of voltage recovery and voltage dip checks will be under Dynam-
ics Simulation Options. Users can select the voltage recovery check and the voltage dip
check independent of each other. Under the voltage recovery check, users can select just
the primary recovery check or the primary as well as the secondary recovery checks.
Selection of the voltage recovery and voltage dip in the Dynamic Simulation Options win-
dow will not automatically trigger the voltage monitoring functions during the dynamic
simulation.
In practice, the voltage violation checks (voltage recovery and/or voltage dip) should start
after fault clearing. The fault clearing time is not pre-determined in a PSS®E simulation,
but is rather a user-defined time that depends on the actual simulation being carried out.
In view of this the triggering of voltage violation check will have to be done by the users.
The steps to be followed for performing voltage violation checks are as follows:
24-7
Playback Model 'TSTGOV1' Program Application Guide: Volume II
1. The first step is to select the required violation check – voltage recovery and/or the
voltage dip from the Dynamics Simulation Options dialog and define the parameters
V1, V2, V3, and t1, t2 and t3.
2. In order to trigger the voltage violation check functions during dynamic simulations,
users have to select the Trigger Voltage Violation Check option, which is available
under Disturbance menu.
The setting of voltage recovery and voltage dip checks as well as the triggering of viola-
tion check will be recordable.
Once the voltage check is triggered, the voltage recovery and/or voltage dip checks will be
performed for all buses. When any bus has a criteria violation, PSS®E will print a mes-
sage to the progress window. Additionally it will be possible to display the violation on the
slider diagram.
24.4.2 Model SYSANG Called from CHSB

This model is extremely useful in monitoring the general results of a stability case. This
model allows the user to put the following outputs in VARs for subsequent output:
1. Average angle.
2. Largest machine angle and its bus number.
3. Smallest machine angle and its bus number, and the angle spread — the difference
between the largest and smallest machine angles.
The largest and smallest angle are extremely useful in helping to locate a machine going
out of synchronism. Any of the subroutine arguments can be set to zero to bypass the logic
and storage of that quantity.
When selected output channels from the activity CHSB dialog, channels are automatically
created, the user can modify the names using Edit>Dynamics/Network Date dialog.
24.5 Playback Model 'TSTGOV1'

The TSTGOV1 model was developed to assist in the evaluation of the performance of the
governor model. The model has two operational modes.
1. Determining the response of the governor model to a defined frequency input, such as
specified in a grid code.
2. Determining the response of the governor model to a frequency input defined in a data
file.
In either mode, the TSTGOV1 model injects the frequency into the simulation as the
machine speed. The governor model will then see this frequency (speed) as its input sig-
nal and will respond according to the dynamics of the governor model.
This is accomplished by creating a simple one bus model as shown in Figure 24-4. In this
case, the bus number was chosen arbitrarily to be 101. In power flow, this bus must be a
code 3 bus (i.e., a swing bus). A constant MW load must be added to the bus (the load
should not be modeled as either constant current load or constant admittance load). The
24-8
Program Application Guide: Volume II Playback Model 'TSTGOV1'
value of this load should be either the initial MW loading of the generator for test mode 1
above, or the value of the generator power at the beginning of the measurements for test
mode 2. In either case, the bus voltage and load reactive power are not of concern. This is
only a test of the power/speed loop and voltage does not play a role. Thus setting the gen-
erator scheduled voltage to 1.0 and the reactive power to zero is recommended, but not
required. When the load flow is solved, the generator output will equal the load.
Figure 24-4 shows an initial loading of 264 MW on the generator.
Figure 24-4. One Bus Test System
The TSTGOV1 model essentially simulates the governor response in an open-loop mode.
That is, it breaks the accelerating power to speed loop by inserting the frequency deviation
into the generator speed array at every time step. The governor model responds to this
speed and adjusts mechanical power. To keep the system balanced, the bus load is
adjusted at each time step in such a way that it follows the governor mechanical power
output. This in turn ensures that the generator electric power output is the same as the
mechanical power input minus any stator losses.
The TSTGOV1 model performs a linear interpolation to calculate the frequency at each
time step in either operational mode. In test mode 2, the model will also calculate the
value of the measured electrical power at each time step, also using linear interpolation.
These values are stored in VARs allowing plotting if desired.
24.5.1 Use of TSTGOV1 to Simulate a Grid Code Frequency Shape

Primary reserves are intended to arrest the initial fall in frequency following a contingency
(e.g. loss of a large unit or an interconnection tie). The ability of a unit to release primary
reserves can be measured by artificially subjecting the unit to a test whereby "measured"
frequency is of the form as illustrated in Figure 24-5. Note that 60 Hz is used as the base
24-9
frequency in Figure 24-5, but the concept is equally applicable to 50 Hz systems.
Figure 24-5. Defined Frequency Excursion for Primary Response Calculation
For example, one definition of primary response capability is defined as the unit's change
in power output in response to this change in frequency, measured at T2 seconds and sus-
tainable for an additional (30-T2) seconds. If the power output measured at T2 seconds is
not sustained for the 30 second period, the primary reserve is the minimum power output
reached during the period from T2 through 30 seconds. The frequency input curve is
intended to represent the system's marginal frequency response following critical contin-
gencies. Parameters in the frequency input curve would be selected to provide some mar-
gin to the frequency at which load shedding is initiated.
Secondary reserves are intended to bring frequencies back to acceptable levels following
unit outages, i.e., frequencies at which most generating units are expected to be able to
operate indefinitely, and high enough so as to reasonably overcome additional unit outages
should they occur in the brief period before frequency is restored through AGC action.
The ability of a unit to release secondary reserves is measured by extending the primary
reserve test as shown in Figure 24-6, again using 60 Hz as the base frequency.
Figure 24-6. Secondary Response Test
24-10
Program Application Guide: Volume II Playback Model 'TSTGOV1'
Secondary response capability is determined, for example, by the lowest governor

response throughout the 30 seconds to 15 minutes period, and is meant to bring the fre-
quency back to F2 Hz within 1 minute after the contingency, and keep it there until
AGC/operator action restore frequencies to nominal.
Of course, other definitions are used by different power systems. Other shapes can be rep-
resented by proper selection of the parameters. If a more complex shape is needed, it can
also be defined using an input data file as explained below.
24.5.2 Use of TSTGOV1 to Simulate a Measured Frequency Response

The following sections illustrate the two methods to use the model and the data required to
perform the respective simulations.
24.5.2.1 Model Data Input

Data records for user-written models of this type must be in the following format (see the
section on Dynamics Model Raw Data File Contents in the PSS®E Program Operation
Manual):
BUSID 'USRMDL' IM 'model name' IC IT NI NC NS NV data list /
Where:
BUSID and IM are the external bus number and the machine id of the machine that is being
subject to test.
IC Is the user-model type code, which in this case would be 505 since the model is classi-
fied as a "Machine Other model".
IT determines the placement in CONET. In this case, IT is 0 since the model does not need
to be called during dynamic simulation network solution.
NI Is the number of ICONs used by the model (NI = 2).
NC Is the number of CONs used by the model (NC = 6).
NS Is the number of STATEs used by the model. (NS = 0).
NV Is the number of VARs used by the model. (NV = 2).
data list specifies NI ICONs, followed by NC CONs.
The data record format for the TSTGOV1 model is thus:

Bus# 'USRMDL' GenID 'TSTGOV1' 505 0 2 6 0 2, ICON(M) to ICON(M+1),
CON(J) to CON(J+5) /
Where the model uses 6 CONs:
CON(J) = TSTART1 (SECONDS)
CON(J+1) = TEND1 (SECONDS)
CON(J+2) = FREQUENCY1 (HERTZ)
CON(J+3) = TSTART2 (SECONDS)
CON(J+4) = TEND2 (SECONDS)
24-11
CON(J+5) = FREQUENCY2 (HERTZ)
The model uses 4 ICONs:

ICON(M) = Flag for grid code test or input from a file
= 0 for grid code test (trapezoidal frequency shape
= 1 to read frequency and power data points from a file
ICON(M+1) = LOAD ID (enter in single quotes, e.g., '1')
And the model uses 2 VARs:

VAR(L) = Measured frequency from file
VAR(L+1) = Measured Pelec of generator from file
As an example for operation in the grid code mode, the data record format for the
TSTGOV1 model using the shape in Figure 24-6 is given below for a 50 Hz system where
F1 is 49.4 Hz, F2 is 49.6 Hz and T1 is 5 seconds and the other times are shown in the fig-
ure. In this case, both the generator and load IDs are 1.
101 'USRMDL' 1 'TSTGOV1' 505 0 2 6 0 2 0 '1' 0.0, 5.0, 49.4, 30.0, 60.0, 49.6 /
As an example for operation in the mode to input data from a file and to determine the
response of the governor model to this frequency input, the data record below can be used
for the TSTGOV1 model. In this example, both the generator and load IDs are 1. Note
that there is no need to input the constants to define the trapezoid and zeros are typically
entered.
101 'USRMDL' 1 'TSTGOV1' 505 0 2 6 0 2 1 '1' 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 /
When the model is used to read data from a file, the data has to be supplied in a file whose
name is 'FTIME1.DAT'. The file must be present in the folder from where PSS®E is
started. The format of this data file must be one data record per line, with each data record
containing the measured data in this order:
Time, Frequency, Power
The data is read free format. The units of time should be seconds; the units of frequency
should be Hz, and the units of power are MW.
There is no restriction in the number of data points supplied in the data file
'FTIME1.DAT'.
The user does not need to calculate and enter the number of data points as the program
automatically reads until the end of the file is reached.
24.5.3 Example of the Use of TSTGOV1 to Simulate a Measured Frequency

Response
The simulation is performed in the usual manner, bringing in a snapshot containing the
model data and a load flow case with the one bus model of 1. Load conversion is not nec-
essary and should not be performed. The load must be a constant MVA load. The snap-
shot should contain the generator, excitation, and governor models for the unit to be
studied and the TSTGOV1 model.
24-12
Program Application Guide: Volume II Equipment Monitoring Models
When the simulation is initialized using activity STRT, PSS®E will read the measurement
data contained in the file FTIME1.DAT and perform an initialization. The user simply
needs to run out to a time consistent with the length of time in the measured data. That is,
there is no disturbance applied by the user - the disturbance is inherently a part of the fre-
quency data driving the simulation.
24.6 Equipment Monitoring Models

There are several PSS®E models provided that monitor specific equipment for output pur-
poses only; they do not affect its solution. These models are not recognized by
activity DYRE; they are ignored and print an appropriate message if the equipment is not
found.
24.6.1 Models Set Up by Activities CHAN and CHSB

Special subroutines have been set up for requesting certain output quantities, not normally
stored by PSS®E. These subroutines are automatically called during a simulation if the
output is requested by activities CHAN and CHSB.
24.6.1.1 Model VOLMAG

This voltage monitoring model (VOLMAG) calculates the magnitude and angle at a bus.
Storage of either quantity can be bypassed by setting the calling argument for each to zero.

RELAY2 calculates the apparent impedance seen from the from bus bus side and checks to
see if circuit is within circles defined in Figure 24-2. If the storage locations are not set to
zero, the apparent impedance is saved in the appropriate location.
24.6.1.3 Model FLOW1

FLOW1 calculates the MW, Mvar, and MVA flowing on the circuit specified out of the
from bus. To calculate total flows, the user should enter a circuit of –1. Flows include line
shunt component at the from bus end. By giving a zero storage location the storing is
bypassed.

RELAY3 calculates the apparent impedance seen from the from bus side looking into a
three-winding transformer and checks to see if the apparent impedance is within the cir-
cles defined in Figure 24-2.

FLOW3 calculates the MW, Mvar, and MVA flowing into the winding from the from bus
on a three-winding transformer. Flows include the magnetizing admittance component at
the from bus end.
24-13
Reactive Switching Devices Program Application Guide: Volume II
24.6.2 Additional Monitoring Models

The following models can be added to subroutine CONET manually (see the PSS®E Pro-
gram Operation Manual, Section 19.9, Advanced Uses of CONEC and CONET).

FLOW2 is identical to FLOW1 except that line shunt components are not included in the
calculation.
24.6.2.2 Model FLOW

FLOW combines the features of FLOW1 and FLOW2 by ICON(I+3), which can be set to 0
to ignore the shunt component or 1 to include it.
24.6.2.3 Model GENTMC

GENTMC calculates the current magnitude and angle on machine base, MBASE, at the
machine terminals.
24.6.2.4 Model GENTMZ

GENTMZ calculates the apparent impedance on machine base, MBASE, from the machine
terminals looking out into the system.
24.7 Reactive Switching Devices

Beyond the constantly varying reactive devices commonly called static var devices
described in an earlier section, the use of high-speed switches and thyristors to gate on
large blocks of either reactors or capacitors is becoming more common. Various unique
schemes defining when these devices will surely be proposed and implemented. This sec-
tion describes models provided in the PSS®E library.
24.7.1 Model SWCAP

SWCAP is a simple model of a capacitor bank that is switched onto a bus if its voltage falls
below a setpoint. SWCAP is completely independent of the power flow switched shunt
elements. The undervoltage relay picks up instantaneously when the bus voltage falls
below the voltage setpoint, CON(J+1), and the circuit breaker closes after a delay speci-
fied in CON(J+2). The delay may be specified as zero to force instantaneous closure of the
breaker when voltage falls below the setpoint.
The switched capacitor may be locked out by setting ICON(I) to a zero or negative value.
After being switched on, the capacitor remains on until removed manually by the user.
The capacitor may be switched off by setting ICON(I) to zero or a negative value and set-
ting VAR(K) to 9999. Because the capacitor is off in the initial condition, SWCAP
requires no special treatment in initialization, except that the setpoint voltage, CON(J+1),
should be less than the initial condition voltage at the bus.
24-14
Program Application Guide: Volume II FACTS Devices
24.7.2 Model SWSHN1

SWSHN1 was designed to use the data stored in power flow switch shunt arrays. It requires
the switch shunt to be at a step and the switch shunt control mode to be discrete (i.e.,
MODSW=1). To be consistent with generator and static var models, this model ignores
the remote bus specified in the power flow.
Pickup timers initiation is dependent on data input. The model allows three different
choices. If the user inputs 0 for the first quantity, the high-voltage pickup and low-voltage
pickup will be the values specified in the power flow. The second input choice is to spec-
ify a voltage deviation of less than or equal to 0.5 for the first CON. The model expects a
negative voltage deviation value for low-voltage switching. Finally, the input values can
be per-unit voltages above and below which pickup times activate. The model prints the
voltage deviation if the pickup timer is activated by deviation; otherwise, it prints the bus
voltage. The pickup timer instantaneously resets if voltage comes within limits. After the
switch timer is activated, it cannot be reset. Both the pick-up timer and the switch timer
reset after each switching action.
The model allows only one timer to be activated at a time either for the pickup or the
switch. One step at a time will be switched on and all steps at a block must be on (or off)
before the model goes on to the next block. If it is intended that this model either trip all
capacitors or apply all reactors at a bus in a single step, then the user must modify the
switched shunt data in the power flow to have only one block with one step having the
proper capacitive or reactive value.
The model allows a different time to be specified for closing and opening of switches, so a
high-speed solenoid close can be represented. Finally, the user may enter a limit on the
number of switches allowed. A large number should be entered (99) to deactivate this fea-
ture.
24.8 FACTS Devices

Flexible ac Transmission System (FACTS) devices constitute a not yet well-defined class
of equipment that resulted from advances in the 80's and early 90's in the field of forced-
commutated thyristors (gate turn-off (GTO), MOS-controlled transistors (MCT), and insu-
lated gate bipolar transistors (IGBT)). As opposed to the conventional thyristors found in
HVDC converters and static var devices, forced commutated thyristors allow switching at
any point of the ac sine wave. The mechanics and control of such high-speed electronic
switches is beyond the bandwidth of PSS®E and is therefore not modeled. Their applica-
tion, however, is leading to a new generation of equipment that can have a significant
impact on the rotor-angle dynamics of interest in PSS®E.
Because FACTS devices are only starting to make their way into the power industry, their
configuration and control is still a matter of debate. For this reason, the models presented
in this section are simplified versions of the actual equipment and are only applicable for
feasibility or scoping studies. They do include, however, some of the main features and
limitations that need to be considered by the system planner contemplating application of
such devices.
24-15
FACTS Devices Program Application Guide: Volume II
24.8.1 CRANI (also known as TCSC)

In these devices, a branch impedance is adjusted by either quick connection and discon-
nection of series capacitor elements or by thyristor control of a series reactor that may or
may not be in parallel with a series capacitor or reactor. The model assumes the impedance
adjustment to be both continuous and linear.
The model can be applied to any branch in the network model except for transformers and
zero-impedance lines. Compensation limits Xmax and Xmin can take any positive (reactor)
or negative (capacitor) value, as long as their range includes the initial branch reactance.
Upon initialization, model output VAR(L+1) and VAR(L+2) are set equal to the power
flow branch reactance. Because of the washout (TW), controller output is zero in the
steady-state. As the simulation progresses, the effective branch reactance is modified in
response to changes in controller input VAR(L), which must be set by appropriate assign-
ment statements in CONEC.
The user should be aware that changes in effective branch reactance are attained by a com-
bination of both branch reactance changes and current injections at the to and from buses.
It is therefore incorrect to monitor with calls from activity CHAN, branch magnitudes such
as active and reactive power or apparent impedances across the device. Instead the user
should instruct activity CHAN to monitor these magnitudes along branches in series with
the CRANI model. These branches can include zero impedance lines on each end of the
CRANI device and would render a true picture of the flows going into and out of the
device.
24.8.2 Static Condenser Model - CSTATT and CSTCNT

In very simplified terms, a STATCON (Static Condenser) consists of a dc capacitor shunt-
connected to the ac network by means of high-speed electronic switches. Given a suffi-
cient number of thyristor bridges it is possible for a converter to synthesize a controlled ac
voltage with acceptable harmonic content behind the converter transformer reactance.
For pure reactive control, this ac voltage is kept in phase with terminal voltage. Because
ideally there is no net exchange of active power with the network, the dc capacitor remains
charged at the desired level. In practice, converters have losses that tend to discharge the
capacitor. In order to replenish the capacitor charging, the synthesized voltage is con-
trolled to slightly lag the terminal voltage, and thus maintain a small flow of active power
between network and condenser. By this same mechanism capacitor voltage is controlled,
which in turn determines the magnitude of the internal voltage and thus determines the
reactive exchange with the network. In CSTATT as well as CSTCNT these active power
flows are assumed negligible, only reactive power exchanges are modeled.
The dynamic models of CSTATT and CSTCNT are the same. The difference between
CSTATT and CSTCNT is that, for power flow CSTATT is modeled as a sychronous con-
denser (i.e., with no active power output), while CSTCNT is modeled as a FACTS device.
Care must be taken to insure that the power flow current output does not exceed CSTATT
current limitations. In order to prevent contributions to short-circuit, a high value of
ZSORCE is recommended.
In the CSTATT and CSTCNT model, the Condenser is modeled as a controlled voltage
behind a transformer reactance. The ac voltage is kept in phase with terminal voltage; i.e.,
24-16
a lossless STATCON is assumed. The resulting STATCON current is limited to the tempo-
rary and steady-state overload limitations of GTO converters.
The STATCON dynamic model (CSTATT and CSTCNT) follows the guidelines proposed
in Reference1. Inputs to the model are voltage reference, and terminal voltage. Because
CSTATT is modeled as a generator in power flow, this model can also accept power sys-
tem stabilizer output (PSSout). Output of the model is STATCON reactive current. Positive
current corresponds to STATCON performance as a capacitor.
The model consists of a voltage regulator with transient gain determined by time constants
T1 through T4 and, more importantly, the integrator gain K. Per-unit steady-state gain is
equal to the inverse of Droop.
AVR gain can be adjusted to attain a conservative crossover frequency of 25 rad/sec:
25
K =
Eterm /Eint
Eterm/Eint is the sensitivity of terminal voltage to internal voltage, and is equal to:
Eterm Zthev
=
Eint SBASE
Zthev + jXtransf ×
STATCONBase
where Zthev is equal to the inverse of the minimum expected per-unit (SBASE) short-cir-
cuit level at the unit terminals.
The model output is subject to limitations of internal voltage and converter current. Typi-
cal values for these limitations are:
Imax = 1.25 pu (i.e., a 25% transient overload capability is assumed)
Vcutout = 0.2 pu
Elimit = 1.2 pu
Other typical model parameters are:

Xtransf = 0.1 pu
Droop = 0.01 to 0.05 pu
T1 = T2 = T 3 = T4
Accl = 0.5
24.8.3 CSMEST
The EPRI CSMEST model was developed with EPRI sponsorship via RP2123-27. It simu-
lates the dynamic characteristics of a SMES. A SMES can be used to improve first-swing
(transient) stability, to provide damping, and/or to limit frequency excursions. Voltages
are normally controlled by an AVR-driven reactive path.
1 "FACTS Device Benefit Assessment on Commonwealth Edison's Power System, Volume 2: A

Study of STATCON Use to Improve Voltage Stability and Power Quality at Schaumburg Sub-
station". EPRI TR-101933 Volume 2, by General Electric.
24-17
First-swing stability applications will normally require stepping the supplementary signal
(VAR(L)) either manually or by a relay model. Damping can be provided by modulation
of SMES active and/or reactive powers. Active power modulation is attained by connec-
tion (via CONEC statements) with supplementary signal controller models such as
CHAAUT. Reactive power modulation is attained from any PSS model such as model
STBSVC. Finally, governor action can be obtained from properly adjusted HVDC control-
ler models driving the active power path. Model CHAAUT, for example, allows simulta-
neous damping and governor control of the active power path.
Because the converter can handle active and/or reactive power but has a finite thermal
capability, the question of sharing converter capability among active and reactive power
arises. Because active power is the primary purpose of the SMES, it is given priority as
shown in the block diagram.
Unlike battery devices, in which energy capacity is likely to be much larger than that
required during a simulation, a SMES device may be precisely sized to a particular appli-
cation. In the EPRI CSMEST model coil energy is modelled explicitly in terms of coil
inductance and coil current. A trial and error process may be needed in SMES sizing stud-
ies to arrive at the appropriate coil energy capacity. Likewise, experimentation may also
be necessary to size the converter.
The EPRI CSMEST model covers two basic types of SMES converters; the voltage-
source and the current-source converters:
• The voltage-source type is characterized by having a chopper+capacitor interface
between the coil and the converter. Much in the way of a STATCON converter, the
SMES converter can be visualized as sequentially switching the dc voltage on the
capacitor so as to synthesize a controlled ac voltage behind the converter transformer
reactance. This internal ac voltage is controlled to regulate both magnitude and phase
with respect to the device terminals. Control of internal voltage magnitude drives reac-
tive power interchange with the network. Control of internal phase angle allows
regulation of active power exchanges with the ac system.
Were it not for the chopper, active power exchanges between capacitor and network
would quickly exhaust or overload the capacitor. It is the role of the chopper to maintain
appropriate voltage (charging) levels on the capacitor by selectively redirecting current
from the coil to the capacitor. The chopper can also be viewed as controlling coil aver-
age dc voltage by rapidly cycling between +/- applications of the capacitor dc voltage
on the coil terminals. This voltage-source design has the advantage that its reactive
power capability is independent of coil charging level, functioning as a Static Con-
denser (STATCON) in the event of complete exhaustion of the coil energy (current) lev-
els.
• In a current-source SMES, coil and converter are connected without any interface.
Such a design is best visualized as an extension of conventional HVDC technology.
The main difference from HVDC converters is that application of self-commutating thy-
ristor technology allows operation in any of the four power quadrants. While
conventional HVDC converters can only consume reactive power, self-commutating
converters can also generate reactive power. As in HVDC terminals, converter ac cur-
rent is proportional to dc (coil) current. Ac current is controlled by having two bridges in
series on the dc side but in parallel on the ac side of the converter. Zero current at the
SMES ac terminals, for example, is attained by having the current on each of the two
bridges 180 degrees out of phase from each other. Maximum ac current, however, is
24-18
still dictated by coil dc current level. Therefore, one disadvantage of this design is that
both active and reactive power capability are limited at low coil charge levels.
In both voltage- and current-source converters the speed of response is very high and a
SMES can make corrections to the ac system much more quickly than synchronous
machines and somewhat more quickly than conventional static var compensators.
The EPRI CSMEST model can simulate both voltage- and current-source SMES devices.
The SMES is modeled as a generator in the power flow. As in the STATCON model, it is
recommended that a high value of ZSORCE be used to eliminate contributions to short-
circuit.
Referring to the block diagram:
• VAR(L) drives the SMES active power path and should be connected by the user to
control logic or controller model output via CONEC assignments.
• Pinit (VAR(L+1)) allows initialization of the model at other than zero active power output.
This, however, will result in an initial condition suspect message that should
be disregarded.
• IDC0 corresponds to initial dc current and, together with KR, drives a slow reset action
that, for 0 power order (VAR(L)+VAR(L+1)=0.), will return the coil current to IDC0 with
time.
• Pmax is a power limit to be used at the user’s discretion.
• VDCMAX and VDCMIN take into account the maximum voltage capability of the con-
verter and coil. This block can also be used to model limitations on maximum rate of
change of coil current; limitations than can be expressed in terms of maximum dc coil
voltages.
• Idcmax1, Idcmax2, Idcmin1 and Idcmin2 limit coil currents to prevent coil overcharging
(Idcmax) or undercharging (Idcmin). To prevent discontinuities, the limits are applied
gradually.
• Iacmax models converter current limitations; similar to those modeled in a STATCON
device. These limits are not applicable to current-source devices because, by design,
the converter must withstand maximum dc current and its corresponding ac current.
• KxVacxIdc models low coil current limitations in current-source converters. Because for
such converters maximum ac current is proportional to dc current both active and reac-
tive power outputs are restricted at low charge levels. In a voltage-source converter this
factor is used to model the inability to sustain maximum dc voltages at low ac voltages.
• Active power output divided by coil current leads to actual (current-source converter)
or average (voltage-source converter) dc voltage on the coil terminals. This dc voltage
is integrated by inductance to obtain coil current.
• The reactive path is comprised of a voltage regulator model. The voltage regulator is
similar to that employed in the STATCON model. Reactive current, instead of internal
voltage, is under direct control of the regulator. Converter current limits are enforced by
the nonwindup limits on the K/S integrator. These current limits are the reactive coun-
terpart of the current limits in the active power path, and are designed to give priority to
active over reactive power orders. As in the STATCON case, the gain K can be
adjusted to attain a crossover frequency of 25 rad/sec:
24-19
K = 25
Eterm /Ireac
Eterm/Ireac is the sensitivity of terminal voltage to SMES reactive current, and is equal
to:
Eterm Zthev × SMESBase

=
Ireac SBASE
where Zthev is equal to the inverse of the minimum expected per-unit short circuit level
at the SMES terminals (on system base, SBASE).
Typical data for a voltage-source SMES can be derived as follows:

• MBASE is selected as maximum coil current times maximum coil voltage. Their product
is typically much larger than the SMES MW rating. dc voltage and current bases are
also set to their maximum permissible values at the coil.
• Inductance, L, is derived from the Power and Energy ratings of the device. Assuming
a rated per-unit power of Pmax, and that rated energy is defined as the maximum time,
T, rated power can be delivered; inductance, L, can be derived from the energy balance
equation:
0.5 × L × (I2 - P2max) = Pmax × T(sec)
Pmax is the SMES MW rating in pu on MBASE (typically much smaller than 1.), but is
also equal to the minimum per-unit current at which rated power can be delivered (i.e.,
with maximum dc voltage (1 pu) applied to coil terminals).
• By definition of per-unit bases, VDCMAX and VDCMIN are 1, and -1, respectively.
• IDCMAX1, IDCMAX2, IDCMIN1, and IDCMIN2 are set to 1, 1.01, 0.02, and 0.01,
respectively.
• Iacmax is calculated as Pmax/Power Factor
• The K factor in KxVacxIdc is set to 1.
• In a SMES solely meant for damping system oscillations, IDC0 could be set midway
between minimum and maximum energy levels available at rated power:
(Pmax2 + I2)
IDCO2 =
2
• Other model parameters are typically set as follows:
KR = 0
T1 = T2 = T3 = T4
VMAX = 9999, VMIN = -9999
Droop = (0.01 to 0.05)/Iacmax
24-20
24.8.4 CBEST (EPRI Battery Energy Storage)

The EPRI CBEST battery model was developed with EPRI sponsorship via RP2123-27. It
simulates the dynamic characteristics of a battery. As in the case of SMES devices, batter-
ies can be used to improve first-swing (transient) stability, provide damping, and/or limit
frequency excursions. Like in the EPRI CSMEST model case, judicious selection of exter-
nal signals is required to model battery performance in any of these functions.
A voltage-source converter is the natural choice for a battery energy system. Power into
and out of the battery is controlled by adjusting battery terminal voltage. The converter
design provides some degree of independence between dc voltage and synthesized voltage
behind converter reactance. This allows independent control of active and reactive pow-
ers.
The active power path in the EPRI CBEST model simulates power limitations into and out
of the battery as well as ac current limitations at the converter. The model assumes that the
battery rating is large enough to cover all energy demands that occur during the simula-
tion. Nevertheless, the model provides an approximate means of computing such require-
ments by separately computing energy out of and into the model, and by adding them in
VAR(L+4). The computations take into account battery inefficiencies both in the retrieval
and in the storage of energy.
The reactive power path is identical to that in the EPRI CSMEST model. AVR gain com-
putation follows the same guidelines as those for the latter model.
The BES is modeled in the power flow as a generator with a large ZSORCE impedance.
The EPRI CBEST model allows initialization at other than zero power factor but will
alarm nonzero state derivatives on the energy-computing states.
If MBASE is set equal to the battery power rating, Pmax is set to 1. Iacmax is set to
Pmax/Power Factor. AVR parameters are set as in the SMES model.
Assuming an 80% turnaround efficiency, retrieval (OutEff) and storage (InpEff) efficien-
cies would typically be set to 1.1 and 0.9, respectively.
24.8.5 CDSMS1
The CDSMS1 model of a D-SMES (Distributed Superconducting Magnetic Energy Stor-
age) device was developed by PTI in conjunction with American Superconductor Corpo-
ration (AMSC), the device’s manufacturer. A D-SMES device, a combination of a SMES
system with a voltage-source IGBT converter, is capable of fast, effectively controlled,
and near instantaneous injection of both real and reactive power into the system. Using D-
SMES is considered as a new option to solve plenty of transmission, generation, and dis-
tribution system problems, including improvement of voltage and angular stability,
increasing power transfer capability of existing grids, damping oscillations, load leveling,
etc.
As in the CSMEST model, in CDSMS1 there exists active (P-path) and reactive (Q-path)
power paths. The model represents the device’s capability of injecting or consuming
active (POUT) and reactive (QOUT) power to hold the voltage at a controlled bus to a
desired level (AVR function). The voltage at a controlled bus is compared against four
threshold values V1, V2, V3, and V4. The device’s controls monitor the voltage to be
held, either at the D-SMES own bus (VDSMES) or remote (Vremote), and determine
24-21
whether it is necessary to inject active and inject or consume reactive power. Active power
injection can also be caused by an auxiliary damping signal, PAUX. In the future the
device, following the damping signal, will also be able to absorb active power from the
system and dissipate it in resistor banks. This device capability is under development,
therefore only its simplified representation is provided.
Any active power injection is formed by discharging the D-SMES magnet. The process of
discharge can be either uninterrupted or repetitive (see model diagram, The Magnet Dis-
charge Curve, datasheet CDSMS1), but during any discharging interval the coil current IL
is ramping down linearly having a given slope under a constant coil voltage VDC. Reac-
tive power injection is formed by the device’s IGBT voltage-source converter. Because
the control of its high-speed electronic switches is beyond the bandwidth of PSS®E, it is
not modeled. There also exists an overload capability activated when the controlled volt-
age is in a specific range; the capability’s definition is given by the Overload Diagram of
datasheet CDSMS1. To have the device’s performance represented more accurately, some
time constants (TON, TOFF) are taken into account that determine the minimum time
intervals needed to activate the magnet or shut it down after the previous switching opera-
tion.
As the model is developed for the device manufactured by AMSC, recommended parame-
ter values and ranges correspond to D-SMES systems currently in production. Comments
are given with reference to the D-SMES Model Diagrams and Data Sheets and only for
specific model parameters that may need clarifying.
Model ICONs:
Notation Description Application
CONV_TYPE Converter type: 0 for current- Set to 1 only.
source converter, 1 for voltage
source converter
IBUS_CONTR Number of remote control bus. Ignored if ICON(M+3) = 0
BOOST_CONTR Boost control flag: 0 if no, 1 if Yes allows the model to change the voltage set
yes point.
VOLT_SEN_LOC Voltage control sensor location
flag: 0 for the IBUS, 1 for con-
trolling the remote control bus
TURN_ON_VOLT Voltage control flag: 0 if no, 1 if No means that the voltage thresholds Vi (i =
yes 1, 2, 3, 4) are ignored by the active power
path, normally set to 1.
TURN_ON_POWER Active power (damping control) No means that PAUX signal is ignored. AMSC is
flag: 0 if no, 1 if yes looking to future technology where D-SMES
systems can be used to dampen phase-angle
swing and improve small signal stability. This
application has not yet been examined deeply
enough to provide commercial solution. Set-
ting to 0 is recommended.
TURN_ON_P Active power output flag: 0 if no, No turns off the active power path (POUT=0).
1 if yes
TURN_ON_Q Reactive power output flag: 0 if No turns off the reactive power path (QOUT=0).
no, 1 if yes
24-22
Model CONs (notation of the diagrams is used):

SRATED rated D-SMES MVA, must be
equal to MBASE
VDC nominal coin voltage (kV)
IINIT Initial dc coil current (kA) Parameters of the Magnet Discharge Curve given
in datasheet CDSMS1. INIT = 1.05 kA, IMIN = 0.4
IMIN Minimum dc coil current (kA)
kA, TDIS = 0.6 sec for AMSC systems currently in
TDIS magnet full-discharge time (sec) production.
TON Minimum time (sec) that real A delay that is established to avoid ’ringing’ in the
power injection must remain off magnet, i.e., quickly turning the real power injec-
before it can be turned back on tion on and off. TON = 0.1 sec for AMSC systems
after being turned off. currently in production.
TOFF Minimum time (sec) that real A delay that is established to avoid ’ringing’ in the
power injection must remain on magnet, i.e., quickly turning the real power injec-
after it is turned on. tion on and off. TOFF = 0.1 sec for AMSC systems
currently in production.
V1 voltage threshold (pu) Voltage thresholds must be V1 > V2 > V3 > V4.
Typical values in per units are 1.03 to 1.05 for V1,
V2 voltage threshold (pu)
0.95 to 0.97 for V2, 0.9 for V3, and 0.5 for V4.
V3 voltage threshold (pu)
V4 voltage threshold (pu) (>0)
KAVR AVR (Q-path) gain Constants of the automatic voltage regulator that
govern the response time of the model. To model
T1 AVR time constant (sec)
quick response, a relatively high gain is required
T2 AVR time constant (sec) (about 700). This high gain has the effect of over-
T3 AVR time constant (sec) (>0) running the effect of the time constants, therefore,
for simplicity, it is recommended to set T1 = T2 =
T4 AVR time constant (sec) (>0) T3 = T4 = 1.0 sec.
AVRDROOP AVR droop Droop setting of the automatic voltage regulator.
PAUX_THRESH A parameter that deals with the as of yet undevel-
oped small signal damping capability of the
PAUX threshold (MW) (>0) systems. Disabled if TURN_ON_POWER is set to
0. Set to 0.1 MW for all AMSC systems currently in
production.
TOVLD time interval of overload, Overload parameters referring to the Overload
when MVA output limit is Diagram of datasheet CDSMS1. Recommended
maximum – SLIMMAX (sec) values are TOVLD = TBACK = 1.0 sec, KOL >
200%.
TBACK time interval when MVA
output limit SLIM is ramping
from maximum value
SLIMMAX to nominal value
SRATED (sec)
KOL overload parameter
(percent)
24-23

TBOOST_BEG boost control starting time Boost voltage control parameters. Setting
(sec) TBOOST_BEG to a value greater than the simula-
tion time interval disables the two other
BOOST_DUR time interval when the boost parameters.
control is active (sec)
STEP_VREF voltage reference step used
by the boost control (pu)
KOV The maximum over voltage level for the low-
voltage winding of the D-SMES step-up
transformer. If during a simulation the remote
parameter determines the
bus voltage Vremote is currently under con-
step-up transformer voltage
trol and the level of KOV is achieved, the
when the remote bus control
device control abandons controlling the
is abandoned (percent)
remote bus voltage and to starts controlling
the D-SMES terminal voltage VDSMES. Rec-
ommended is the value of 110%.
VQMAX maximum limit for AVR state
2 (pu)
VQMIN minimum limit for AVR state
2 (pu)
IACMAX The maximum ac current limit is determined
mainly by the overload parameter KOL.
maximum limit for the D- IACMAX is an additional limit that helps to
SMES ac current (pu) keep the output of the systems within capabil-
ities. Recommended is a value slightly larger
than the KOL.
PMAX maximum limit for POUT
(pu on MBASE)
PMIN minimum limit for POUT
(pu on MBASE)
An arbitrary number of D-SMES models may be connected to any bus. The model is pro-
vided with both state space and extended term simulation mode. The D-SMES device,
although modeled as a FACTS device, is initialized like an electric machine and also
treated as a machine in power flow calculations. In the power flow case MBASE =
SRATED, ZSOURCE = 0 + j99999 to avoid inadequate device’s contribution in short cir-
cuit currents of the system. PGEN must be set to 0, QGEN is set at user’s discretion. How-
ever, the value of QOUT should be coordinated with the voltage thresholds Vi (i = 1, 2, 3,
4), otherwise it may adversely affect the initialization, and the IACMAC limit, otherwise
QOUT will be changed during the initialization to match the limit.
24-24
24.8.6 SVSMO1U1 - Model of continuously controlled SVC

The block diagram of the PSS®E SVSMO1U1 model for a continuously controlled SVC
is shown in Figure 24-7.
Figure 24-7. SVSMO1 Block Diagram
The main features of the model are:

• Proportional-Integral Primary Voltage Regulation Loop with gains Kpv and Kiv.
• Slow Susceptance Regulator: The PI regulator with gains Kps and Kis is the slow sus-
ceptance regulator which slowly biases the SVC reference voltage between the values
of vrefmax and vrefmin to maintain the steady-state output of the SVC within the band-
width of Bscs and Bsis.
• Non-Linear Slope/Droop: If the parameter flag2 is set to 0 then a standard linear droop
of Xs1 is assumed, as is typically in most designs. Alternatively, by setting flag2 to 1,
one can use a three piece piecewise linear droop setting. This can be used to make the
SVC non-responsive in a given bandwidth, similar to deadband control. This control,
however, is more susceptible to limit cycling if not properly tuned.
• Optional Deadband Control.
Note: The deadband control, slow susceptance regulator, and non-linear droop are all
intended for the same purpose - maintaining the SVC at a low steady-state output when
the system voltage is within a given bandwidth. However, these three control strategies
achieve this in quite different ways. Thus, for stable and suitable control response it is
highly recommended to never use any combination of the deadband control, slow-suscep-
24-25
tance regulation and non-linear slope/droop. Only one of the three should be used in the
control strategy. The model checks for this condition during initialization and does not
allow using combinations of these controls.
• Protection: The over/under-voltage strategies and over/under-voltage trip points.
• Short-term rating: the SVC output can exceed its continuous rating up to a given
amount for a short time period.
• Mechanically Switched Shunt (MSS) Logic: detailed MSS logic is implemented which
allows for automated MSS switching based on the SVC VAr output. Two thresholds
(typically, one for fast switching and one for slower switching) are implemented with dif-
ferent delays on switching. The MS capacitor discharge time can also be set, i.e. time
the MSC must be switched off before it can be switched back on. The MSS breaker
delay is also modeled. The models can handle up to 8 MSSs: reactors and/or
capacitors.
Note: the MSS switching must be properly coordinated with the slow-susceptance regula-
tor, if used. Typically, to avoid excessive MSS switching, the slow-susceptance regulator
time constant is chosen such that it acts first to bring the SVC to within the first threshold.
If it is unable to achieve this, then the MSSs switch. Thus, the delay time on the first
(smaller and slower) threshold for MSS switching is chosen to be significantly longer than
the slow-susceptance time constant. Also, the slow-susceptance time constant is much
longer than the primary voltage regulator loop response time.
• There is a provision to add the signal from the auxiliary control, e.g. for power oscillation
damping.
• The upper and lower limits of the PI main controller Bmin/Bmax are parameters of the
model that can be set by an auxiliary program outside of the model by means of the
standard API function.
Load Flow Representation of SVSO1U1

A SVC modeled by the SVSMO1 model must be represented in PSS®E as a continuously
controlled switched shunt.
In PSS®E, switched shunts may include up to 8 blocks with up to 9 steps in each of them,
72 steps total. Each step is sized in MVARs, negative for reactors and positive for capaci-
tors.
The SVC is also capable to regulate voltage by controlling mechanically switched shunts
(MSS), like mechanically switched capacitors (MSC) or reactors (MSR) available in the
system around SVC. MSSs are represented in load flow as up to 8 fixed shunts connected
to the same MSS bus. The status of fixed shunts in load flow can be either on-line or off-
line.
Example Dynamic Data File for the SVSMO1U1 Model

7 'USRSWS' 'SVSMO1U1' 24 1 14 47 5 29
@!ICONs:
@!+++++++++++ SVC REMOTE BUS ++++++++++++
0
@!+++++++++++ MSS BUS ++++++++++++
3
24-26
@!++++ MSS SWITCH FLAG: 0 - NO SWITCH, 1 - SWITCH ON Q +++++++

1
@!+++ DROOP FLAG: 0 - LINEAR DROOP, 1 - NON-LINEAR DROOP ++++++
0
@!+++++++++ FLAGS: ++++++++++++++++++++++++++++++++++++++++
0 0 0 0 0
0 0 0 0 0
@!CONs:
@! CON(J ) = UVSBmax, max cap U/V limit
0.2000
@! CON(J+ 1) = UV1, undervoltage setting 1
0.5000
@! CON(J+ 2) = UV2, undervoltage setting 2
0.3000
@! CON(J+ 3) = UVT, undervoltage trip setting
0.5000
@! CON(J+ 4) = OV1, overvoltage setting 1
1.3000
@! CON(J+ 5) = OV2, overvoltage setting 2
1.5000
@! CON(J+ 6) = UVtm1, undervoltage trip time 1
1.0000
@! CON(J+ 7) = UVtm2, undervoltage trip time 2
7.0000
@! CON(J+ 8) = OVtm1, overvoltage trip time 1
1.0000
@! CON(J+ 9) = OVtm2, overvoltage trip time 2
0.2500
@! CON(J+10) = Xs1, slope/droop
0.0200
0.0000
0.0000
@! CON(J+13) = Vup, up volt break-point for non-linear droop
1.1000
@! CON(J+14) = Vlow,low volt break-point for non-linear droop
0.0000
@! CON(J+15) = Tc1, voltage measurment lead time constant
0.0000
@! CON(J+16) = Tb1, voltage measurment lag time constant
0.0100
@! CON(J+17) = Tc2, pre-controller lead time constant
0.0000
@! CON(J+18) = Tb2, pre-controller lag time constant
0.0000
@! CON(J+19) = Kpv, prop. gain of SVC PI-controller
50.0000
@! CON(J+20) = Kiv, integral gain of SVC PI-controller
250.0000
@! CON(J+21) = Vemax, voltage error max limit
0.5000
24-27
@! CON(J+22) = Vemin, voltage error min limit

-0.5000
@! CON(J+23) = T2, thyristor firing sequence control delay
0.0100
@! CON(J+24) = Bshrt, short-term max. suceptance of SVC
2.2000
@! CON(J+25) = Bmax, continuous max. suceptance of SVC
2.0000
@! CON(J+26) = Bmin, min. suceptance of SVC
-0.5000
@! CON(J+27) = Tshrt, duration of short-term rating
3.0000
@! CON(J+28) = Kps, prop. gain of SSC PI-controller
0.0000
@! CON(J+29) = Kis, integral gain of SSC PI-controller
0.0000
@! CON(J+30) = Vrmax, max. output of slow suceptance control
0.0000
@! CON(J+31) = Vrmin, min. output of slow suceptance control
0.0000
@! CON(J+32) = Vdbd1, steady-state voltage deadband
0.0000
@! CON(J+33) = Vdbd2, inner deadband
0.0000
@! CON(J+34) = Tdbd, recovery time to deadband
0.0000
@! CON(J+35) = PLLdelay, PLL delay in voltage recovering
0.1500
@! CON(J+36) =xeps, delta added to the susceptance bandwidth
0.5000
@! CON(J+37) = Blcs, larger threshold for switching MSCs
40.0000
@! CON(J+38) = Bscs, smaller threshold for switching MSCs
20.0000
@! CON(J+39) = Blis, larger threshold for switching MSRs
-40.0000
@! CON(J+40) = Bsis, smaller threshold for switching
-20.0000
@! CON(J+41) = Tmssbrk, time for MSS breaker to operate
0.1000
@! CON(J+42) = Tdelay1, time delay for larger threshold
0.5000
@! CON(J+43) = Tdelay2, Time delay for smaller threshold
3.0000
@! CON(J+44) = Tout, time MSC should be off before switch back
300.0000
@! CON(J+45) = Vrefmin, lower Vref limit
0.8000
@! CON(J+46) = Vrefmax, upper Vref limit
1.2000
/
24-28
24.8.7 SVSM02U1 - Model of discretely controlled SVC

To better understand the SVC model structure and the way of representing the SVC in
load flow, it is reasonable to start with the description of the arrangement that was used to
test this model. This is shown in Figure 24-8.
The SVC simulated by this model comprises discretely switched elements only, such as
thyristor switched capacitors (TSC) and thyristor switched reactors (TSR). These elements
are connected to two mid-voltage (MV) SVC buses #991 and 992. A composition and a
number of switched elements in both halves of the SVC connected to these two buses is
the same, so both halves of the SVC are identical. The MV buses are connected via step-
up transformers to the SVC high voltage (HV) bus #990 which, in turn, is connected to the
system HV bus #154. This bus is a component of the standard example case SAVNW.SAV
supplied with the PSS®E installation.
In load flow, both halves of the SVC are represented as switched shunts operating in the
discrete mode. In PSS®E, switched shunts may include up to 8 blocks with up to 9 steps
in each of them, 72 steps total. Each step is sized in MVARs, negative for reactors and
positive for capacitors.
The SVC is also capable to regulate voltage by controlling mechanically switched shunts
(MSS), like mechanically switched capacitors (MSC) or reactors (MSR) available in the
system around SVC. The designated buses for MSS connection are shown in the diagram:
5 MSS buses connected to each of HV buses (#15401 through 15405 and 99001 through
99005) and MV buses (#99101 through 99105 and 99201 through 99205) SVC buses, i.e.
up to 20 MSS designated buses total in the test case.
24-29
MSSs are also represented in load flow as switched shunts but, different from SVC dis-
cretely controlled elements TSCs and TSRs, MSSs controlled by the SVC operate in the
fixed mode
Figure 24-8. SVC arrangement used for testing SVSMO2
The block diagram of the SVC dynamic simulation model is shown in Figure 24-9.
This model includes following features:
• Three controllers, namely
- SVC voltage controller
- SVC slow susceptance controller (SSC)
- MSS controller
• The program reads the information from the dynamic data input file and performs a rou-
tine check of topology, data quality, similarity between two halves of SVC, and
consistency between dynamic and load flow data.
• The controlled voltage error has a small deadband necessary for the discretely con-
trolled equipment.
• The SVC voltage controller has a droop and a non-linear gain as a function of the con-
trolled voltage. The output of the voltage controller is the demanded SVC susceptance.
• If the controlled voltage drops below the SVC cut-off voltage, the SVC voltage controller
gets disabled.
• The demanded SVC susceptance is entered into the look-up table. This table is built
by the program at initialization based on the information from the load flow.
24-30
• With the given delay, the output of the look-up table forms the model output that is used
for updating the SVC current injection to the system which is used at each integration
step to update the system voltage vector.
• A reference voltage of the SVC voltage controller Vref is the output of the SVC slow
susceptance regulator. By changing the reference susceptance Bref it is possible to
change Vref.
• The input from an auxiliary control is provided.
MSS bus numbers are entered into the program and checked for the consistency with the
load flow. Also entered is the information of the available size of each MSS. At initializa-
tion, the program puts together this information with the load flow data and makes a judg-
ment whether an MSS is ON or OFF. If, for example, the respective switched shunt size in
load flow is zero but in dynamic data the available size is -0.05 the program figures out
that there is a 5 MVAr reactor available that is off in the beginning of the simulation.
In the dynamic model the number of MSSs controlled by a SVC cannot exceed 8.
• The command for MSS switching comes from the output of the model BSVC.
• The MSS switching logic takes into account the following:
- different switching times for the first and subsequent on-operations and for the off-
operation
- a permissible number of operations for a MSS
- a “cooling” time between switching off and subsequent switching on of a MSS
- reset of timers if the nature of a demand has changed during the switching cycle
- one switching at a time meaning that subsequent switching is allowed only after the
previous switching has been completed
- in case of the positive MVAr demand, each MSS is checked: if this is a reactor and
it is on it will be switched off; if this is a capacitor and it is off it will be switched on;
vice versa in case of the negative demand.
• The “smoothing” mechanism implemented in the model meets two goals:
- Reduction of the system voltage excursion caused by a MSS switching
- better use of the SVC susceptance range available for control
This mechanism is simulated by means of a counter-action of the SVC immediately fol-
lowing a MSS switching. For example, when the 5 MVAR MSC is switched on the same
portion, if available, is subtracted from the SVC demand.
• The overvoltage protection with regard to the SVC MV bus voltage. The SVC usually
controls the HV SVC bus voltage. At the same time it monitors the SVC MV bus voltage
24-31
and, if it exceeds the given threshold, the SVC susceptance demand immediately
drops.
Figure 24-9. SVSMO2U1 Block Diagram
The example dynamic data input file for the SVSMO2 is given below.
991 'USRSWS' 'SVSMO2' 24 1 38 47 3 118
991 0 154 0 0
99101991029910399104
99201992029920399204
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0.0200 0.0300 -0.0100 0.0100 0.9000
0.6000 1.1000 1.2000 0.0000 0.0000
0.0000 0.0000 20.0000 100.0000 20.0000
100.0000 0.0000 20.0000 0.0167 0.0000
0.1000 0.1000 -0.1000 1.1000 0.9000
1.0000 -1.0000 -0.5000 -0.0400 -0.0200
0.0200 0.0400 -0.0400 -0.0200 0.0200
0.0400 0.0000 0.0280 0.0500 0.0500
0.7000 0.0050 -0.0050 -0.0100 10.0000
0.0100 1.5000 /
24-32
24.8.8 SVSM03U1 - Model of VSC based SVC

SVSMO3 is a generic Static Var System (SVS) model for a Voltage Source Converter
(VSC) based STATCOM. This model is coordinated with Mechanically Switched Shunt
(MSS) devices.
The essential difference between the SVSMO3 and SVSMO1, the generic SVS model for
a thyristor-controlled reactor (TCR)-based Static Var Compensator (SVC) is that the
STATCOM is a constant current source once it reaches its limit while the SVC is a con-
stant impedance device when it reaches its limit. Thus, the reactive power output of the
STATCOM is proportional to voltage when it reaches limit. The block diagram of the
PSS®E SVSMO3 model for of STATCOM is shown in Figure 24-10. The block diagram
24-33
for modeling of short-term rating is shown in Figure 24-11.
Figure 24-10. Block Diagram of SVSMO3
Figure 24-11. Modeling of short-term rating
The main objective of the STATCOM is the voltage control of the specified bus.
24-34
The main features the model are2
SVSMO3 Primary Control

• "Proportional-Integral Primary Voltage Regulation Loop with proportional gain, Kp and
integral gain, Ki.
SVSM03 Secondary Control

• Slow-Reset Control: The PI regulator with gains Kpr and Kir is the slow-reset controller
which very slowly biases the STATCOM reference voltage between the values of Vre-
fmax and Vrefmin to maintain the steady-state output of the STATCOM within the
bandwidth of +Idbd and -Idbd. This control is enabled by setting flag1 as 1.
• Non-linear Droop Control: If the parameter flag2 is set to 0 then a standard linear droop
of Xc0 is assumed, as is typically in most designs. Alternatively, by setting flag2 to 1,
one can use a three piece piecewise linear droop setting. This can be used to make the
STATCOM non-responsive in a given bandwidth, similar to deadband control. This con-
trol, however, is more susceptible to limit cycling if not properly tuned.
• Deadband Control: If the dbd parameter is set to 0 then deaband control logic is dis-
abled. On the other hand if dbd is not equal to 0, the STATCOM main control acts only
after the STATCOM bus voltage is outside the given deadband defined by Vref+dbd to
Vref-dbd.
Note: The deadband control, slow-reset control, and non-linear droop are all intended for
the same purpose - maintaining the STATCOM at a low steady-state output when the sys-
tem voltage is within a given bandwidth. However, these three control strategies achieve
this in quite different ways. Thus, for stable and suitable control response it is highly rec-
ommended to never use any combination of the deadband control, slow-reset control and
non-linear droop. Only one of the three should be used in the control strategy. The model
checks for this condition during initialization and does not allow using combinations of
these controls.
Other features
• Protection: The over/under-voltage trip points.
• Short-term rating: the STATCOM output can exceed its maximum continuous current
rating up to the maximum short-term current rating (=Ishrt*Imax1) for a short time
period (= Tdelay1). Alternatively, the short-term rating can be based on the specified
I2t limit.
• Mechanically Switched Shunt (MSS) Logic: detailed MSS logic is implemented which
allows for automated MSS switching based on the STATCOM current output. The
thresholds, Iupr and Ilwr are implemented for switching of capacitor or reactor. The MS
capacitor discharge time can also be set, i.e. time the MSC must be switched off before
it can be switched back on. The MSS breaker delay is also modeled. The model can
handle up to 8 MSSs: reactors and/or capacitors.
2 P. Pourbeik, "User's Manual for ABB STATCOM Model in GE PLSF® and Siemens PTI
PSS/E®", May 8, 2006. ABB Report No. 2006-11241-Rpt4-Rev2
24-35
Under/Over Voltage/Frequency Relay Models Program Application Guide: Volume II
Note: the MSS switching must be properly coordinated with the slow-reset control, if
used. Typically, to avoid excessive MSS switching, the slow- reset control time constant is
chosen such that it acts first to bring the STATCOM to within the two thresholds. If it is
unable to achieve this, then the MSSs switch. Thus, the delay time for MSS switching is
chosen to be significantly longer than the slow-reset time constant. Also, the slow-reset
time constant is much longer than the primary voltage regulator loop response time.
Load Flow Representation

A STATCOM modeled by the SVSMO3 model must be represented in PSS®E as a
FACTS device in load flow.
The STATCOM is also capable of regulating voltage by controlling mechanically
switched shunts (MSS), like mechanically switched capacitors (MSC) or reactors (MSR)
available in the system around STATCOM. MSSs are represented in load flow as up to 8
fixed shunts connected to the same MSS bus. The status of fixed shunts in load flow can
be either on-line or off-line.
'STATCOM ' 'USRFCT' 'SVSMO3U1' 21 1 13 44 6 27
0 1 6 0 0 0 0 0 2 0 0 0 0
0.10000E-01 0.0000 0.10000 0.0000 25.000
0.50000 -0.50000
0.50000E-02 1.0000
0.0000 10.000 0.10000
0.0000 0.10000E-02 0.10000E-01
0.10000 -0.10000
1.4000
0.50000 0.20000 1.0800 1.2000 1.3000 1.0000 0.80000E-01
1.0000 1.0500 0.95000
0.0000 0.0000
0.0000 0.0000 0.0000
0.10000E-01 1.0000 0.10000E-01 1.0250 0.97500
0.10000 300.00 0.50000
0.30000 -0.30000
10.000 /
Figure 24-12. Example Dynamic Data Input File, SVSMO3
24.9 Under/Over Voltage/Frequency Relay Models

The under/over frequency models (FRQDCA and FRQTPA) and under/over voltage mod-
els (VTGDCA and VTGTPA) are protection models, located at the generator bus, that
continuously monitor the frequency/voltage on that bus or a remote bus specified by the
user. They trip the generator for under- and overfrequency/voltage conditions on the gen-
erator (or remote bus).
Both voltage and frequency protection relay models have two entries:
• FRQDCA and VTGDCA relays disconnect generator bus, i.e., disconnect all equipment
attached to generator bus.
• FRQTPA and VTGTPA relays disconnect generators only.
24-36
Program Application Guide: Volume II Voltage Source Converter DC Line Dynamic Model
The relay timer is started during under/over frequency/voltage conditions, i.e., when fre-
quency/voltage is less/greater than or equal to the corresponding pickup threshold. The
relay resets instantaneously if the frequency/voltage restores within the two pickup thresh-
olds. If the relay is not reset, a trip signal is sent to the circuit breaker if the timer reaches
its setting. Frequency/voltage must have remained in an under/over frequency/voltage
condition for the entire time delay for generator tripping to occur. Generator tripping is
delayed by the circuit breaker time.
Several relays can be used to simulate the coordinates of the protection system with the
voltage/frequency versus time function.
24.10 Voltage Source Converter DC Line Dynamic Model

This section describes the dynamic model for a two-terminal, Voltage Source Converter
(VSC) based HVDC system. The model has been developed for the PSS®E dynamics soft-
ware in coordination with the associated VSC DC line power flow PSS®E model.
24.10.1 VSC Dynamic Model Overview

The model described herein is general performance class model developed to evaluate the
interaction of a VSC based HVDC system with the ac network(s) to which it is intercon-
nected. As with other PSS®E HVDC models, the objective is to represent the transient
behavior of the outer, or supervisory controls, because these control actions have the most
influence on the connected ac network. Controls that have been explicitly represented in
the model take into consideration the 0-30 radians/second bandwidth over which PSS®E is
applicable. Other controls and response, particularly those associated with the HVDC side
of the system, have not been specifically modeled due to their rapid response in relation to
the time scale of most PSS®E simulations. Rather, an approximate representation of DC
characteristics has been implemented.
Control of a pair of VSCs and the fundamental dynamic interaction between the pair is
accomplished in the PSS®E representation through the use of a Power Control Order func-
tion. This Power Control Order function utilizes the voltage-sensitive current limits of the
VSCs to coordinate active power transfer across the HVDC system when either VSC goes
into limitation. It should be noted that the model implies direct communication between
VSCs. In the actual system, coordination is accomplished through monitoring and high-
speed control of DC voltage, no direct communication is used.
The dc transmission line dynamics model (VSCDCT) in PSS®E is composed by the inte-
gration of three modules, two voltage source converter modules (VSCDYN) for the VSCs
at each dc line terminals and one dc transmission line module (DCLINE) for the dc link.
VSCDCT is a current injection model. Figure 24-13 shows schematically the modules for
the VSCDCT PSS®E Dynamic model. The data sheet for this model with typical values
for the input parameters is given in the PSS®E Model Library, (Model Configuration). Please
note that an integration time step of ¼ cycle or less should be used when PSS®E simula-
tion cases include a VSC with typical time constants as listed.
24-37
Voltage Source Converter DC Line Dynamic Model Program Application Guide: Volume II
A VSC module can operate in stand-alone mode or in conjunction with a second VSC. In
the latter case, the pair of VSCs is coordinated via a separate module named DCLINE.
This module is discussed subsequently. Stand-alone mode of operation can be set by tak-
ing out of service one of the VSC via power flow parameter input data. Specifying TYPE
= 0, sets the converter out of service. Stand-alone operation may be used if the VSC is
operating as an independent SVC, or when a user is only interested in studying the ac net-
work on one side of a VSC dc system that has no parallel ac ties.
Figure 24-13. VSCDCT PSS®E Model
24.10.2 VSCDYN Module

The VSCDYN module has been developed to represent the control functions of a VSC
converter. The VSCDYN module recognizes the following actions by the controls:
• AC voltage control or reactive power control,
• Active power control or DC voltage control, and
• Current output limitation.
Additionally, the VSCDYN module accommodates the following actions by the user:
• Power ramping, and
• Converter blocking.
24.10.2.1 Active Power Reference Regulation

Typically, the active power order is set manually and power stepping unit carries out the
ramping of power order reference. The VSCDYN module has provision for accepting a
power order directly (P_aux) via VAR(L) for VSC # 1 or VAR(L+12) for VSC #2. The
model also has provision for receiving a power order from the DCLINE module via
VAR(L+7) for VSC # 1 or VAR(L+19) for VSC # 2. When in the stand-alone operating
mode, this power order value remains constant, as set during initialization.
In normal operation of a two-terminal VSC dc system, active power is controlled in one
station while the other converter controls dc voltage. The user specifies whether a VSC is
in active power or dc voltage control by the TYPE parameter in power flow input data.
24-38
24.10.2.2 AC Voltage Control

The AC Voltage Control function varies the reactive power order to control the ac voltage
at the regulated bus. The regulated bus is the ac filter bus by default, however, the user can
specify any remote bus by entering the bus number via ICON(M+1) for VSC # 1 and
ICON(M+3) for VSC # 2. If these values are zero, the model takes the bus numbers from
the power flow input data, REMOT parameter. In a VSC dc system, the AC Voltage Con-
trol is independent in each VSC.
The AC Voltage Control has the provision for droop (CON(J+6) for VSC # 1, or
CON(J+19) for VSC # 2), to be included in the function to effectively alter the voltage ref-
erence point. This feature allows two or more VSCs to regulate the same bus voltage by
sharing reactive power. Without the droop, it would be possible for the reactive output of
the VSCs to drift apart and run the VSCs up to their maximum current limits.
In the VSC power flow data, the parameter MODE could be set to 1 for ac voltage control
or 2 for fixed ac power factor. This information is passed to the VSCDCT model during
dynamics simulations except that when MODE is equal to 2, the VSC holds a constant
reactive power order, rather than ac power factor. If the model is initialized in reactive
power control, the value of reactive power is determined from the power flow conditions.
A VSC can be switched between ac voltage and reactive power control during a simula-
tion. The reactive power order is maintained in VAR (L+6) for VSC # 1 or VAR(L+18)
for VSC # 2 during the course of a simulation. These VARs are updated appropriately
when in ac voltage control.
24.10.2.3 Current Order Limiter

This model converts the active and reactive power orders to current orders by dividing the
values by the magnitude of the ac filter bus voltage.
The function of the Current Order Limiter (COL) is to limit the amplitude of phase current
of the fundamental component to the rated current of the converter. This concept is illus-
trated in Figure 24-14.
Figure 24-14. Characteristic of Current Order
24-39
Voltage Source Converter DC Line Dynamic Model Program Application Guide: Volume II
If the amplitude of the current order is lower than the maximum current (Iacmax),
(CON(J+5) for VSC # 1 or CON(J+18) for VSC # 2), the outputs of the COL will be the
same as the corresponding inputs. If the amplitude of the current order is higher than Iac-
max, the outputs of the COL will be reduced and a feedback signal is sent to the Voltage
Controller to limit wind-up of the integrator.
The reduction of the active and reactive current order is accomplished by using the Power
Weighting Factor (0.0 = PWF = 1.0) fraction. The value for the PWF is taken from the
power flow input data. When PWF is 0.0, only the active current order is reduced; When
PWF is 1.0 only the reactive current order is reduced; otherwise, a weighted reduction of
both active and reactive current order is applied by solving a quadratic constraint and
accepting the minimum real root. When roots are not real values, the reactive current order
is set to zero and the active current order is set to the limit (Iacmax), (CON(J+5) for VSC
# 1 or CON(J+18) for VSC # 2).
If the input setting for the maximum current (Iacmax) [CON(J+5) for VSC # 1 or
CON(J+18) for VSC # 2] is zero then the power flow input data IMAX is used in per-unit
of the converter base. If again, IMAX is zero in the power flow input data (default value)
then unlimited loading is allowed in the VSCDCT model.
24.10.2.4 VSC Blocking

A VSC can be blocked by setting ICON(M) for VSC # 1 or ICON(M+2) for VSC # 2
equal to 1. Blocking a VSC instantly changes the active and reactive current orders to
zero. In a VSC dc transmission system, one converter can be blocked and the remaining
converter can continue controlling ac voltage.
24.10.3 DCLINE Module

The DCLINE module coordinates the active power flow between a pair of VSCs. The
module import from the power flow input data the reference value of DC voltage (pole-to-
pole, DCSET for TYPE 1) to be maintained at the DC voltage controlling VSC. Addition-
ally, the DCLINE module import from the power flow input data the series DC resistance
of transmission system (RDC in ohms). That information is used to estimate the series cur-
rent dependent system losses. Fixed losses are determined at initialization based on the
power difference between the two converters and the calculation of series losses. Both the
fixed (Pzero_loss) and total (Pdc_loss) are routed to VAR(L+32) and VAR(L+33), respec-
tively, in the model for the user's information.
During a simulation run, the DCLINE module produces a power order for each of VSCs of
a two-terminal transmission system accounting for the losses in the HVDC system. The
system losses are always compensated for at the DC voltage controlling VSC except for
the condition of a severe current limitation at this station.
In a two-terminal HVDC Light system, the current limitations in the two converters work
independently. When a limitation is encountered, a difference in active power between
converters can result. In the actual system, this difference creates an error in the DC volt-
age and a fast acting regulator is activated to regulate the power reference in such a way
that the unbalance is eliminated. In the simulation model, this procedure is simplified by
directly controlling the power order reference; the model assumes a fast well-tuned DC
24-40
control system. The references are reduced to the minimum value for the power transmis-
sion capacity of the two VSCs. These values are sent to the DCLINE module from each of
the VSCDYN modules, again by utilizing the VAR(L+7) for VSC # 1 and VAR(L+19) for
VSC # 2, PM variables. If one of the converters is blocked, the DCLINE module is
ignored.
24.10.3.1 Active Power Ramping

Power transfer across a two-terminal, VSC based HVDC system can be changed in a sim-
ulation run by altering VAR(L+24) of the model. At initialization, VAR(L+24) will be set
equal to the network active power injection at the power controlling VSC. The active
power transfer can be ramped up (within the appropriate system limits) by increasing the
absolute value of this variable and applying the same sign. A positive sign indicates power
flow across the HVDC system from the voltage controlling VSC (TYPE 1) to the power
controlling VSC (TYPE 2).
24.10.4 VSC DC Power Flow Representation

One model is required for each HVDC Light Transmission line in the power flow network
to be analyzed. Any transformer associated with the VSC DC line should be modeled by
an explicit transformer record in the PSS®E power flow data. The power flow data pro-
vides the initial conditions for the dynamic simulation of the model. The VSCs that make
up an HVDC system should be dispatched in the PSS®E power flow program assuming a
power transfer from one VSC to the other. Additionally, the converter MVA rating should
be specified in the power flow input data as a non-zero parameter SMAX (Program Oper-
ation Manual, Voltage Source Converter (VSC) DC Transmission Line Data), when this
VSCDCT model is used for dynamics simulations.
AC filters corresponding to a VSC can be modeled as fixed or switched capacitive shunt at
the ac bus to which the VSC is connected. Figure 24-15 illustrates the typical reactive capa-
bility of a VSC assuming a nominal ac bus voltage (1.0 per unit). The curve that cuts off
the right hand side of the circular characteristic is the limit due to the maximum steady-
state internal VSC voltage. As the ac filter bus voltage decreases below 1.0 per unit, this
portion of the curve will shift to the right allowing for more reactive support from the
VSC. This overall capability curve can be used as reference for specifying the VSC reac-
tive power limits in the dyre file, [CON(J+9), CON(J+10), CON(J+22), and CON(J+23)].
Keep in mind, reactive power of one VSC of a HVDC system can operate independent of
the reactive power output of the other.
24-41
Variable Frequency Transformer (VFT) Model Program Application Guide: Volume II
Figure 24-15. VSC Steady-State Reactive Capability Curve
24.11 Variable Frequency Transformer (VFT) Model

The Variable Frequency Transformer (VFT) is a power system device developed by GE to
either interconnect two asynchronous AC power systems or to control the flow on an AC
transmission path. This device has a 3-phase winding on the stator and a motor-driven
rotor which also has a 3-phase winding. This rotary transformer can continuously adjust
the phase shift between the stator and rotor to any desired angle.
The VFT is a medium voltage device with a typical terminal voltage rated at 17kV. Gen-
erator step-up transformers are required on both sides of the VFT to connect it to the trans-
mission system.
Note that this device consumes reactive power. As a consequence, shunt reactive power
compensation might be required for voltage control.
This user guide details the use of the PSS®E dynamic simulation model for the GE VFT
developed by Siemens PTI based on technical information in the GE report "VFT Model-
ing for Planning Studies" Rev. 3, dated June 3, 2009 and previous work from Hydro-Qué-
bec-IRÉQ.
24-42
Program Application Guide: Volume II Variable Frequency Transformer (VFT) Model
24.11.1 Limitations
Note that PSS®E does not presently distinguish a VFT from a standard phase-shifting
transformer, in particular, the VFT's capability of coupling two asynchronous networks.
When using this device to couple two asynchronous power systems, a swing bus (Type 3
bus) is required in EACH island. Without a swing bus in each island, the power flowing
across the VFT would not be regulated. However, if the VFT is used to regulate flow
within a synchronous network (functioning similar to a standard phase shifting trans-
former) only one swing bus is needed in this synchronous system.
When a VFT is used to connect two asynchronous systems, two PSS®E features are
impacted as result of modeling the VFT as a standard phase shifting transformer.
1. Activity TREE
2. The Inertial Power Flow
Activity TREE finds the number of asynchronous islands in a network and verifies that
every island has a swing bus. Since the VFT is represented as a phase-shifting trans-
former, it links its two terminals as an AC branch. Hence, in terms of activity TREE, the
VFT would link the two asynchronous islands into a single island. As a result the output
of activity TREE would not recognize the existence of the two islands and would report
them as one island containing two swing buses.
Activity TREE is implicitly called when solving the inertial power flow to determine the
"synchronous island" that shares the same steady state frequency. Due to the VFT being
modeled as a phase-shifting transformer, the inertial power flow can not distinguish
between the two systems and hence would assume that the two asynchronous systems
share the same frequency and, for example, all generators of both "islands" would contrib-
ute to frequency control.
It is presently planned that this limitation will be addressed in a future release of PSS®E.
That is, logic will be added such that activity TREE, and other activities that use it, will
recognize the VFT as a device capable of linking two asynchronous islands and adjust
their calculations appropriately.
24.11.2 Example
Figure 24-16 shows a single line diagram of the VFT example system.
In this diagram, the
VFT is connected between buses 99003 and 99004. The transformers between 99006 and
99003, and 99007 and 99004 represent the step-up transformers. When entering the trans-
24-43
Variable Frequency Transformer (VFT) Model Program Application Guide: Volume II
former data, the phase shifting element is defined to be on the winding 1 bus. Note the
selection of the tapped bus is arbitrary and the power can flow in either direction.
Figure 24-16. Single Line Diagram of the VFT and Associated GSU and Capacitor Bank
For the phase-shifter to control active power flow, the transformer must be selected as
"Auto Adjust" and the control variable as "Active". as shown in Figure 2 2.
Figure 24-17. Data Record in PSS®E Power Flow Interface
Typical parameters are shown in Table 24-1. These typical parameters are given in per unit
on the rating of the VFT. This rating must be specified under the entry "Winding MVA".
Typically, it is 100 MVA per VFT.
It is important to note that the VFT has a large magnetizing current, typically consuming
18 Mvar per VFT. Therefore, it is important that the magnetizing branch be included in
the transformer model.
Also note that the solution option "Adjust phase shift" needs to be selected when solving
the power flow solution to enable the VFT to control the flow.
Table 24-1. Power Flow Parameter for the VFT
Parameter Typical Value (Based on VFT rating) Note:

RVFT 0.4 % VFT Series Resistance
XVFT 12 % VFT Series Reactance
24-44
Program Application Guide: Volume II Modeling for Stability Analysis
Table 24-1. Power Flow Parameter for the VFT
Xmag -18% VFT Magnetizing Reactance

Xt1 , Xt2 10 % Step-up Transformer Reactance
C1 , C2 20% to 80% depending on needs Shunt Capacitors at VFT Terminals
24.12 Modeling for Stability Analysis

The model of the VFT is suitable for studies of transient stability and other events where
the frequency of the power system or between the two systems, if modeling an asynchro-
nous connection, remains within a few percent of nominal.
The VFT consists of a three-phase stator, a three-phase rotor driven by a DC motor, and
the control system.
Figure 24-18 shows the block diagram of the rotor and the motor-drive system. The input
command (desired rotor speed, Spd Cmd), is the output of the control system shown in
Block Diagram of the VFT Drive System.
TRQ_max_lim
Var(L+17) angle(  )
Var(L)
Kwi TRQ_max_lim 180

X
s Var(L+17) TRQ_rate K + 12 
Spd Cmd K+2
Var(L+12) +
+ m
+ 1 + 1 1 angle
Var(L+18) TRQ_CMD sTdr v 2 HS S (rad)
From Power & - TRQ_m in_lim +
m - -TRQ_rate -
Freq Control Var(L+18) Te K+ 1 K
Kwp TRQ_min_lim m
Var(L+3)
GE VFT: Drive System
Figure 24-18. Block Diagram of the VFT Drive System
The block diagram of the main controller of the VFT is shown in Figure 24-19. It has three
branches:
• a P-I controller that regulates the VFT power,
• a frequency control that regulates a speed input from the difference in Thevenin fre-
quency between the two terminals,
24-45
Modeling for Stability Analysis Program Application Guide: Volume II
• a speed feedback stabilizing path.

VFT Governor
F p_M ax
Output
Pcmd-f
Var(L+11)
Kpi
VFT P ower
s
Pvft +
1 K+3
Var(L+1)
sTd Fpstab -F p_M ax
or - +
Var(L+2) ST AT E(K+13)
Kpp
Fsrlim +
DFth: Va r( L+6) Fratelim Fp_Max
V FT Drive
+ +
1 1 - S ystem
- TFSR S
DFth : Va r(L+7) Va r(L +12)
- -
-Frate lim K+4 -Fp_M ax spd_cmd
-Fsrlim
m
Va r(L+13)
STATE (K  1)
Fp _ Sta blim
+ +
1 + +
sTd F pst ab K pstab
- -
GE VFT: K+5 -Fp_Stablim
Power and Frequnecy

Controller
W pstab
S
K +6
Figure 24-19. Block Diagram of the Main Controller of the VFT
The torque of the DC motor can be limited due to its rotational speed and/or the terminal
voltages. The Drive Motor Torque Limit prevents windup of the speed regulator by limit-
ing its output torque command when the rotor speed is high. The Voltage Dependant
Torque Limit prevents the control from making excessive change when the voltage is low.
The driving torque is also limited by an acceleration limit (Trq_Acel_Lim) which limits
the torque to be within a band around the per-unitized Power Command. The block dia-
gram of these limits is shown in Figure 24-20.
IMTR_lim x TRQ_max_lim
m x y
r M in Var(L+17)
|u | %
m IMTR_taper
S tate( K+1)
y +
p0 Spd_taper
+
1
Var(L+11)
Trq_Acel _Lim
Pcmd _f
V th 1
V ar(L+8)
TRQ _Vdtl_min -
+ 1 +
x x TRQ_min_lim
sTVdtl
Max Var(L+18)
-
K+7 -1
Vth 2
Var(L+9) -1
TRQ _Vdt l_max
GE VFT:
Torque Limiter
Figure 24-20. Block Diagram of the Torque Limiter of the VFT
24-46
Program Application Guide: Volume II Modeling for Stability Analysis
The VFT has a controller that can be used to emulate governor action when connecting
asynchronous systems. If the frequency on one side of the VFT deviates beyond the dead-
band FDB1 on the system 1 side or FDB2 on system 2, the VFT would modulate the
power order command according to the droop settings R1 for system 1 or R2 for system 2.
The VFT also has a Voltage Dependent Power Limiter (VDPL) which limits power trans-
fer during a low voltage event in order to, for example, prevent voltage collapse. This lim-
iter also modulates the power order command. The block diagram of the VFT governor
and the VDPL are shown in Figure 24-21.
DFth1 P0_ cmd

Var(L+6) Dpg _ max Plim0 Var ( L  10) Plim0
gov _ lim
spd c md
if | Var ( L  12) |  F p max
+
the n 1 + To Power &
D pg _ min x sTgov
+ Out  0 Freq Control
E lse Pcmd-f
- +
Out  in K+9 Var( L+11)
DFth2 D pg _ max
Plim 0  Var ( L+10)  Plim0  gov _ lim
Var (L+7)
P0 _ cmd P0_ c md
Var ( L+10)
D pg _ min
V1
Plimo GE VFT:
+ 1 Tr n Base
Min
sTVD SBase
Governor Control
gov _ lim
-
V2 State(K+8)
V PX V P1
TUP
TVD 
TDOWN
Figure 24-21. Block Diagram of the VFT Governor
At model initialization, the VFT model assumes that the VFT is operating in steady state
under nominal voltages. The VFT angle in VAR(L) is initialized as the absolute VFT
angle in degrees and the VFT Power Feedback Pfbk is measured from the tapped side bus
as the power in MW into the VFT.
The VFT adjusts the power flow by rotating its rotor. As this rotation angle becomes
large, the current flowing across the VFT could significantly deviate from the initial value.
When the rotation exceeds the VFT factorizing angle, CON(J+40), the power flow admit-
tance matrix will be re-factorized. This re-factorization speeds up the network solution
and prevents a potential divergent solution when the VFT alters the power flow signifi-
cantly from the initial conditions. As a starting point, Siemens PTI recommends a value
of 20 degrees for this re-factorization angle, but this angle may need to be adjusted based
on the characteristics of the power system modeled.
When connecting asynchronous systems, the VFT links the two systems with different fre-
quencies by rotating its rotor to follow the calculated Thevenin Frequency on each side.
The Thevenin Frequency is calculated at both ends of the VFT.
24-47
Modeling for Stability Analysis Program Application Guide: Volume II
For side 1:
The same calculation is performed for side 2. These signals are the inputs to the Main
Controller of Figure 24-19.
The real power flowing across the VFT is stored in VAR(L+1) and VAR(L+2) for the
from and to side flows respectively, with the convention of positive flow from the bus into
the VFT. Similarly the reactive power is stored in VAR(L+4) and VAR(L+5). Note that
the default channels for branch flow for the VFT should NOT be used as they do not gen-
erally take into account the movement in VFT. That is, if plotting the VTF flow is desired,
the VFT model VARS should be used.
At initialization, the initial active power flow is stored in Pcmd. To change the active
power flow to a new setpoint, Pcmd, VAR(L+10), can be changed to the new setpoint.
Note that the relative direction of the power flow must be taken into account.
A sample PSS®E Dynamic Data Input File for Activity DYRE and DYRE,ADD is shown
below:
99003 ’VFT1’ 99004 1
26.200 0.15000 0.30000 0.20000E-01 0.20000E-01

0.30000E-02 0.52000E-01 0.20000E-01 0.10000 0.55000E-01
0.60000E-01 1.0000 0.10000 0.50000E-01 0.50000E-01
500.00 500.00 0.20000E-01 75.000 4.0000
0.20000 0.25000E-01 1800.0 48.000 3.1500
2.7300 2.5000 5.0000 0.70000 0.95000
3.0000 0.30000 1.1500 0.10000E-01 0.10000E-01
0.10000E-01 0.10000E-01 0.30000 1.5000 1.5000
20.000 0.25000E-01 /
24-48
Chapter 25
Data Verification
Chapter 25 - Data Verification
25.1 General Approach
25.1.1 Stages of Verification

One of the most difficult aspects of power system dynamic simulation work is obtaining
complete and valid data. Problems arise both because of errors in data files and because
parameter values are simply not given by many of the usual sources of data.
Because the data that is normally available on generators, excitation systems, and other
equipment is unreliable in many cases, it is essential to make a careful check of the data
values entered on the model data sheets. The first and most obvious approach to data veri-
fication is to check all parameters for reasonableness after they have been established in
the PSS®E database. This will detect many simple errors, particularly transcription errors,
but is not a conclusive indication of validity of the data.
The second, and most effective, way to check data is to make special simulations of the
response of the generating units in hypothetical isolated operating conditions. Saturation,
excitation system response, and governor effects are then not masked by the electrical syn-
chronization effect but are clearly and unambiguously visible. These special simulations
correspond to the tests that can be made on the generating units in actual practice and,
because the form of the test result needed for satisfactory performance of the unit is well
understood in practice, they assist greatly in detecting those insidious situations where
parameter values appear to be reasonable but are actually incorrect. PSS®E includes activ-
ities for both levels of data checking.
25.1.2 Parameter Range Checking

The PSS®E activity DOCU allows the checking of data for gross errors such as misplaced
decimal points, improper relationships between parameters, and so on. Activity DOCU
lists the parameters and calling arguments of each model. The output listing of DOCU, a
sample is shown in Figure 25-1, confirms that the model-calling arguments and parameters
were actually loaded into the PSS®E program and memory arrays as intended.
Critical examination of the output of DOCU for a large system is difficult because of its
volume. Activity DOCU provides an initial editing of the DOCU output and restricts it to
25-1
Data Verification PSS®E 33.4
General Approach Program Application Guide: Volume II
machines where the values of parameters are suspect. DOCU compares each parameter
with a representative range of values and lists output for a machine only if it has a param-
eter outside the representative range. A typical output of DOCU is shown in Figure 25-2.
.
1100KV DC CASE
REPORT FOR ALL MODELS BUS 200 [HYDRO 345]
** GENSAL ** BUS NAME BSKV MACH C O N ' S STATE'S

200 HYDRO 345 1 29- 40 13- 17
MBASE Z S O R C E X T R A N GENTAP From load

1750.0 0.00000+J 0.25000 0.00000+J 0.10000 1.02500 flow case
T'D0 T''D0 T''Q0 H DAMP XD XQ X'D X''D XL From
5.000 0.050 0.200 5.00 0.00 1.0000 0.7500 0.4000 0.2500 0.1000
CON
S(1.0) S(1.2) data
0.1100 0.6200 array
** SCRX ** BUS NAME BSKV MACH C O N ' S STATE'S SOLID

200 HYDRO 345 1 153- 160 62- 63 FED
0.100 10.000 200.0 0.050 -5.00 5.00 1.0 10.00

200 HYDRO 345 1 215- 226 80- 83 5- 6
R-PERM R-TEMP TR TF TG VELM GMAX GMIN TW AT DTURB QNL

0.050 0.300 5.00 0.050 0.500 0.200 1.00 0.00 1.25 1.20 0.50 0.080
Figure 25-1. Typical Output from Activity DOCU
25-2
PSS®E 33.4 Data Verification
Program Application Guide: Volume II General Approach
1100KV DC CASE
DATA CHECK FOR ALL MODELS BUS 100 [NUCLEAR 345]
BUS 100 MACHINE 1: Suspicious value of T’qo;

T'Q0= 0.1500
that’s right! It should be 1.5.
100 NUCLEAR 345 1 1- 14 1- 6
1100.0 0.00000+J 0.20000 0.00000+J 0.15000 1.02500

6.50 0.060 0.15 0.050 4.00 0.00 1.8000 1.7500 0.6000 0.8000 0.2000 0.1500
S(1.0) S(1.2)
0.0900 0.3800
Warnings!
Excitation system
BUS 100 MACHINE 1: data corresponds
TF1= 2.0000 to a rather high
KF= 0.1000 TF1= 2.0000 TRANSIENT GAIN= 20.0000
transient gain
** IEEEX1 ** BUS NAME BSKV MACH C O N ' S STATE'S VAR
100 NUCLEAR 345 1 121- 136 52- 56 1
TR KA TA TB TC VRMAX VRMIN KE TE KF TF1

0.000 46.00 0.060 0.000 0.000 1.000 -0.900 0.000 0.460 0.100 2.000
SWITCH E1 S(E1) E2 S(E2) KE VAR

0.0 2.3000 0.1000 3.1000 0.3300 -0.0155
BUS 100 MACHINE 2:

T''Q0= 0.0020 T"qo too small in relation to time step
X''D= 0.2000 XL= 0.2300
Something wrong! Xl cannot be larger than X"
100 NUCLEAR 345 2 15- 28 7- 12
1100.0 0.00000+J 0.20000 0.00000+J 0.15000 1.02500

6.50 0.060 0.20 0.002 4.00 0.00 1.8000 1.7500 0.6000 0.8000 0.2000 0.2300
S(1.0) S(1.2)
0.0900 0.3800
Figure 25-2. Typical Output from Data Checking Mode of Activity DOCU
This example warns of an unacceptably small value of q-axis subtransient time constant
and an impossible situation where the value of leakage reactance exceeds that of the sub-
transient reactance. Both parameters must be reviewed and corrected before PSS®E can
give meaningful results. DOCU does not necessarily indicate that parameter values are
wrong; it simply indicates that they deserve to be checked. Conversely, the absence of
warning messages from DOCU does not indicate that all data are valid. Most power sys-
tem parameters can fall within reasonable normal value ranges and still be invalid. For
example, the field rheostat parameter, Ke, of the IEEE type 1 excitation model is often
25-3
General Approach Program Application Guide: Volume II
assigned a value of -0.05 in manufacturers’ listings of typical excitation system parame-

ters. But this parameter changes as the plant operator changes the manual-mode voltage
setpoint of the excitation system, and hence it can vary widely from the typical value, not
only between different plants, but as the loading of a given plant is adjusted. Many other
excitation system parameters are field adjustable, with the normal adjustment ranges being
more than sufficient to cause serious variation of system performance. Section 25.5 con-
tains the parameter range values used by activity DOCU for the various models in the
PSS®E model library.
25.1.3 Performance Verification

The proper criterion for the correctness of generator, excitation system, and turbine gover-
nor data is not that all parameter values lie within typical ranges but rather, that the param-
eter values entered for each item correspond to correct performance of that unit under test
conditions. After data has been checked for gross errors by activity DOCU, the user should
verify that the sets of parameter values correspond to the following:
1. Correct steady-state values of all quantities that are normally measured or documented
in normal operation of a generating unit.
2. Correct reproduction of standard dynamic response tests that may be run on gener-
ating units or their control systems in isolation.
The testing of data with respect to these criteria is handled by running PSS®E simulations
of standard test conditions, observing the steady-state and transient results, and comparing
these with either actual test results if available, or with typical unit performance if test data
is unavailable. Individual unit operating conditions for which actual or typical perfor-
mance data are available for the majority of generating units include:
1. Steady-state generator operation at defined load; measurement of steady-state excita-
tion voltage.
2. Exciter response to step change in voltage regulator reference with voltage feedback
signal held constant; recording of transient exciter output voltage and of response ratio.
3. Excitation system and generator response to step change in regulator reference point
setting with generator at rated speed on open circuit; recording of excitation voltage,
Efd, and generator terminal voltage.
The simulation of these test conditions is facilitated by PSS®E activities ESTR, ERUN,
GSTR, and GRUN and the support program VCV. No special data or setup is needed; just
execute simulation runs using ESTR/ERUN or GSTR/GRUN in place of the normal
STRT/RUN combination. The special activities temporarily override the user’s channel
assignments to place the appropriate unit response quantities into the output channels.
These special data verification simulations are used each time a new set of excitation sys-
tem or governor data is introduced into a study. It is impossible to interpret the results of a
simulation of many interconnected machines without the prior knowledge that the individ-
ual generators, excitation loops, and governing loops are properly modeled with data val-
ues corresponding to an acceptably tuned response for each unit.
25-4
Program Application Guide: Volume II Generator Data
25.2 Generator Data
25.2.1 Reactances and Saturation Function

PSS®E requires the specification of unsaturated synchronous machine reactances together
with specification of two points on the machine’s open circuit saturation curve. The cor-
rect specification of the saturation data is necessary for the generator field voltage, Efd, to
take its correct value at all loadings. The correct initial value of Efd is of major importance
in any simulation run where the excitation system ceiling is a significant parameter; that
is, in virtually all transient stability runs. Any error in the initial value of Efd in such a sim-
ulation produces a corresponding error in the maneuvering range available to raise field
flux linkages following a fault. Because this field-forcing effort is a key factor in deter-
mining transient stability limits, a correct initial value of Efd is just as important as correct
specification of excitation ceiling and generator reactances.
25.2.2 Generator V-Curves

The best form of verification of generator reactance and saturation data is the machine V-
curves that plot generator terminal current versus excitation voltage, Efd, over the
machine’s entire operating range. These curves are commonly provided by generator man-
ufacturers, along with open-circuit saturation curves, as standard design documentation.
Typical manufacturers’ data is shown in Table 1 as well as Figures 25-4 and 25-3. Figure 25-
5 shows program VCV graphic output used to verify that a set of data extracted from
Table 1 and Figure 25-4 does correctly reproduce the generator’s steady-state characteris-
tics.
Table 1: Typical Generator Reactance and Time Constant Data,
Rated Output 69 MW, 0.9 pf
Reactance or Time
Unsaturated Saturated
Constant
xd 1.974
x’d 0.26 0.196
xd 0.171 0.128
xq 1.882
x’q 0.41
xl 0.141
T’do 6.6
Tdo 0.023
T’qo 0.48
Tqo 0.049
25-5
Generator Data Program Application Guide: Volume II
1.4
1.2
1.0
Terminal Current (pu)
0.8
Rated Operating Point
0.6
0.4
0 pf Lead 0.9 1.0 0.9 0 pf Lag

0.2
0
0 100 200 300 400 500 600
Field Current (A)
Figure 25-3. Typical Generator V-Curves as Supplied by Manufacturer

Corresponds to Table 1
Air Gap Line

Open-Circuit
1.2 Magnetization
Curve
1.0
Output Voltage (pu)
0.8
Open-Circuit
0.6
0.4
0.2
0
0 100 200 300
Field Current (A)
Figure 25-4. Saturation Curve as Supplied by Manufacturer; Corresponds to Table 1
25-6
Figure 25-5. VCV Graphical Output for Case Where Full Machine Data is Known
25-7
Examining Figure 25-3 shows that the air gap line field-current for rated voltage, open cir-
cuit is 162 A, and field current at rated load, rated power factor = 448 A. Hence, Efd at
rated output 448/162 = 2.77 per unit. The results computed by VCV on the basis of the
proposed generator data agrees with this value. VCV also allows the user to enter specific
loading information. It calculates and prints the Efd at that loading, as well as the internal
machine angle relative to the terminal voltage assumed at a zero angle. The typical PSS®E
user would not normally use the angle but someone doing machine testing could find it
valuable.
25.2.3 Use of Typical Machine Data

While the complete data of Figure 25-5 is the ideal, it is often necessary to set up machine
data on the basis of a much less detailed machine specification. The traditional approach
in the face of minimal or incomplete machine data abandons the properly detailed syn-
chronous machine models such as GENROU and GENSAL and uses the classical
machine model, as implemented by GENCLS in PSS®E. This approach was justified with
previous generations of stability programs, largely by the argument that the widespread
use of detailed machine models would involve an unacceptable penalty in computational
efficiency. The use of the GENCLS classical model offers no significant computational
advantage in PSS®E, however. More importantly, the use of GENROE, GENSAE, GENROU
and GENSAL for the majority of machines involves no significant penalty in PSS®E. Cor-
respondingly, no justification exists for use of GENCLS to model incompletely specified
machines in PSS®E. Such an approach forces the simulation of these machines to neglect
all generator damping effects, indicate incorrect machine internal voltage and reactive
power output in the postdisturbance steady state, and neglect both steady-state and
dynamic effects of voltage regulators.
It is perfectly reasonable to assume that the majority of machines for which complete data
is unavailable are basically of conventional design, that they do contribute damping
torques, that they are subject to voltage regulator action, and that they do adopt postdistur-
bance conditions according to synchronous reactance values. It follows, therefore, that the
preferred approach to the representation of incompletely specified machines is to use the
detailed machine models GENROE, GENSAE, GENSAL, and GENROU with typical
data that represents the principal response characteristics of normal generators. Further-
more, voltage regulator action at these machines should be represented by a simple excita-
tion system model with data set to give typical conservative response characteristics; a
special excitation system model, SEXS, is provided in the PSS®E model library for this
purpose.
The user can conveniently characterize generator reactance values by the value they imply
for field voltage, Efd, at full load in the absence of saturation. Figure 25-6 shows the values
of Efd corresponding to full load and various power factors. The plot in Figure 25-6 reveals
slight difference in the calculated Efd due to variation in Xq for unity and leading power
factors. Rated lagging power has a difference of about 25%. Figure 25-7 shows the value of
SE(1.2) needed to give a value of Efd exceeding the unsaturated value of Figure 25-6 by
25% at rated load and 0.9 (lagging) power factor. Figures 25-6 and 25-7 apply only for the
25-8
typical machine reactance ratios of Figure 25-8 and Figure 25-9. For machines with known
specific reactances, but no saturation curve, the values of SE(1.0) and SE(1.2) can be esti-
mated by program VCV. The value of SE(1.2) is dependent upon the generator synchro-
nous reactance because increasing machine reactance requires a higher internal flux level
to support a given terminal voltage and loading. Figures 25-6 and 25-7 may be used to esti-
mate a proper value of SE(1.2) and of exciter ceiling level when using the typical data of
Figure 25-8 through 25-10, which show data forms for models GENSAL, GENROU, and
SEXS listing suitable sets of typical data for salient pole generators, turbine generators,
and excitation systems, respectively.
E fd
3.0
2.5
2.0 0.8 Lag
0.9 Lag
1.5 1.0
0.9 Lead X q = 0.96X d

1.0 X q = 0.75X d
0.5 Ld
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Figure 25-6. Excitation Voltages at Rated MVA Using Typical Generator Data
(Saturation Neglected)
25-9
SE (1.2)
1.0 SE (1.2) to give Efd exceeding unsaturated

value by approximately 15%
0.8
GENROU
0.6
SE (1.2) to give Efd exceeding unsaturated
0.4 value by approximately 25%
0.2 GENSAL
0 Ld
0.8 1.0 1.2 1.4 1.6 1.8 2.0
Figure 25-7. Estimation of SE(1.2) for Use with Typical Generator Data
of Figures 25-7 and 25-8
25-10
GENSAL

machine # 1* I.
This model uses CONs starting with #____ J, EFD
Efd ISORCE Source Current
and STATEs starting with #____ K. VOLT at
VT GENSAL ETERM Terminal Voltage
The machine MVA is 5* for each of 1* units = 5* Terminal
MBASE. Bus
ZSORCE for this machine is 0* + j 0.25* ANGLE Angle

on the above MBASE.

J+1 0.05 Tdo (>0) (sec) K+1 q
J+2 0.06 Tqo (>0) (sec) K+2 kd
J+5 1.5 Xd
J+6 1.2 Xq
J+7 0.4 X’d
J+8 0.25 Xd = Xq
J+9 0.12 Xl
J+10 0.03 S(1.0)
J+11 0.25 S(1.2)

machine MVA base.
Figure 25-8. Typical Salient Pole Generator Data
25-11
GENROU
Round Rotor Generator Model (Quadratic Saturation)

machine # 1* I.
Efd ISORCE Source Current
and STATEs starting with #____ K, VOLT at
VT GENROU ETERM Terminal Voltage
The machine MVA is 5* for each of 1* unit = 5* Terminal
MBASE. Bus
ZSORCE for this machine is 0 + j 0.25 on the above ANGLE Angle

MBASE

J+1 0.05 Tdo (>0) (sec) K+1 E’d
J+2 1 T’qo (>0) (sec) K+2 kd
J+3 0.05 Tqo (>0) (sec) K+3 kq
J+4 3 Inertia, H K+4  speed (pu)
J+6 1.4 Xd
J+7 1.35 Xq
J+8 0.3 X’d
J+9 0.6 X’q
J+10 0.2 Xd = Xq
J+11 0.1 Xl
J+12 0.03 S(1.0)
J+13 0.4 S(1.2)
Xd, Xq, X’d, X’q, Xd, Xq, Xl, H, and D are in

pu, machine MVA base.
IBUS, ’GENROU’, I, T’do, Tdo, Tqo, Tqo, H, D, Xd, Xq, X’d, X’q, Xd, Xl, S(1.0), S(1.2)/
Figure 25-9. Typical Solid Rotor Generator Data
25-12
SEXS
Simplified Excitation System

EFD
This model uses CONs starting with #____ J, VUEL SEXS
and STATEs starting with #____ K. VOEL

J 0.1 TA/TB
J+1 10 TB (>0) (sec)
J+2 100 K
J+3 0.1 TE (sec)
K First integrator
K+1 Second integrator
IBUS, ’SEXS’, I, TA/TB, TB, K, TE, EMIN, EMAX/
VREF EMAX
EC – 1 + TAs K
 EFD
(pu) 1 + TBs 1 + TEs
+
EMIN
VS
Figure 25-10. Typical Excitation System Model and Data

for Use when Detailed Data is Not Available
25-13
Excitation System Data Program Application Guide: Volume II
25.3 Excitation System Data centers first on the saturation curve param-
eters and the amplidyne output limit,
25.3.1 General Considerations VRMAX. Activity ESTR determines the
value of the exciter field resistance param-
The verification of excitation system data eter Ke. The amplidyne output limit VRMAX
is complicated by several factors: the exis-
is fixed at 1.05 in the initial trial by setting
tence of a wide variety of types, the fact
the constant, CON(J+7), corresponding to
that key parameter values may depend
Ke, to zero and allowing the initialization
upon the reactances of the generators to
to determine Ke as a variable in each run.
which they are applied, and operating con-
Curve A of Figure 25-12 represents the sat-
ditions. Checking parameter values against
uration curve used in the first trial.
typical values is unrewarding in many
The first step in checking the excitation
cases, but useful checks can be made by
system data, after the use of
using activities ESTR and ERUN to simu-
DOCU,CHECK, is to confirm that the
late excitation system tests. The use of
exciter saturation curve data is properly
these tests can best be illustrated by exam-
related to the excitation requirement of the
ple.
main generator. It is good practice to
assign E1 a value close to rated load excita-
25.3.2 Exciter Ceiling
tion voltage, and E2 a value close to ceiling
The ceiling field voltage obtainable from excitation voltage. (The excitation voltage
many excitation systems is determined at rated load is not 1.0; unity excitation
largely by the saturation characteristic of voltage gives rated main generator voltage
the exciter and, to a secondary extent, by on open circuit in the absence of satura-
the maximum output available from the tion.) The Response Ratio Test option of
excitation control source, particularly in activity ESTR is used to initialize the gen-
the case of dc exciter systems. Figure 25-11 erator at rated output. The preceding
shows the transfer function diagram of a DOCU output confirms that the Ke con-
fairly typical dc excitation system from the stant has been set to zero to tell the model
1950s. This system, represented by the to calculate its own value for Ke, and that
model IEEEX1, uses a dc generator as its this value will be placed in VAR(1). The
main exciter, with control being exerted by ESTR output for the unit at bus 1 shows a
an amplidyne. The excitation system is rated load value of Efd of 2.6665 and thus
presumed, on the basis of contract docu- confirms that the value, 2.47, specified for
ments, to have a specified response ratio of E1 in the exciter data is well placed relative
1.0 and a specified ceiling of twice rated to the rated output of the exciter. A value
excitation. An initial estimate of the data of E1 that is not close to rated exciter out-
for this exciter, taken from manufacturer’s
put should be reviewed because it is likely
typical data, is shown in Figure 25-11. A run
that it was obtained from data on an exciter
of ESTR/ERUN is used to check that this
different from the one under consideration.
data corresponds to the specified perfor-
After initializing for the response ratio test,
mance of this excitation system. (This
activity DOCU would show the value of Ke
method leaves open the question of
determined by the model for the rated
whether the actual performance matches
operating condition (see Section 16.2.4).
the specification, but is better than using
The Ke value, –0.599, is quite reasonable.
typical data with no check at all.Interest
25-14
Program Application Guide: Volume II Excitation System Data
Furthermore, activity DLST would confirm that the amplidyne control element, which is
shown by Figure 25-11 to be STATE(K+2), has been initialized at 10% of its positive range.
25-15
IEEEX1
IEEE Type 1 Excitation System
This model is located at system bus #____ IBUS, ECOMP

machine #____ I, VOTHSG
EFD
This model uses CONs starting with #____ J, VUEL IEEEX1
and STATEs starting with #____ K, VOEL

and VAR #_____ L.

J 0 TR (sec)
J+1 40 KA
J+2 .06 TA (sec)
J+3 1. TB (sec)
J+4 1. TC (sec)
J+5 1.05 VRMAX or zero
J+6 -1.05 VRMIN
J+7 * KE or zero
J+8 .5 TE (>0) (sec)
J+9 .08 KF
J+10 2. TF1 (>0) (sec)
J+11 0 Switch
J+12 2.47 E1
J+13 .035 SE(E1)
J+14 3.5 E2
J+15 .6 SE(E2)
K Sensed VT
K+1 Lead lag
K+2 Regulator output, VR
K+3 Exciter output, EFD
K+4 Rate feedback integrator
VAR # Description
L KE
25-16
IBUS, ’IEEEX1’, I, TR, KA, TA, TB, TC, VRMAX, VRMIN, KE, TE, KF, TF1, 0., E1, SE(E1), E2, SE(E2)/
VREF VS
VRMAX
Regulator
+ +
1 1 + sTC KA 1 EFD
EC –  +  +
 (pu)
(pu) 1 + sTR VERR 1 + sTB 1 + sTA VR sTF
– –
VFB VRMIN
SE + KE
sKF
1 + sTF1
VS = VOTHSG + VUEL + VOEL Damping
Figure 25-11. Sample Data for Excitation System Data Verification
5 B
SE (4.5) = 0.47
Exciter Output Voltage (pu)
4
A
SE (3.5) = 0.6
3
SE (2.47) = 0.035
2
0 1 2 3 4 5
Exciter Field Current (pu)
Figure 25-12. Saturation Curves of Initial dc Exciter Model, (A)

and Proposed Model of Larger Exciter, (B)
The foregoing confirms that the excitation system data is reasonable with regard to the
steady state. The next step is to run activity ERUN. When run after activity ESTR in
response ratio mode, ERUN automatically steps the voltage regulator reference upward by
a large amount, forcing the excitation system as rapidly as possible to ceiling. The
25-17
ESTR/ERUN combination automatically overrides existing channel assignments and

places machine field voltages in the first n channels where n is the number of units in the
power flow case. The output of ERUN shows the rise of units excitation and at t = 0.5 sec
automatically records the response ratio of each exciter. The response ratio, as defined in
the IEEE Transactions, Volume PAS-88, 1969, is then defined. This area is shown in
Figure 25-13. After an adequate time, the ERUN simulation settles with each exciter at its
ceiling.
Using curve A of Figure 25-12 for saturation, ESTR would show that the response ratio is
0.456 and the ceiling is 3.1957. Both values are far below the target values of 1.0 and 2 ×
2.67 = 5.34, respectively. Examination of the IEEEX1 block diagram shows that
• The initial rate of rise of excitation voltage is determined principally by the maximum
output, VRMAX, of the control element.
• The excitation ceiling is determined principally by saturation of the exciter and, second-
arily, by the value of VRMAX.
The low values of response ratio and ceiling shown by an initial ESTR/ERUN test indicate
that both VRMAX and the exciter saturation curve must be revised.
Doubling the size of the amplidyne by changing VRMAX to 2.1 raises the response ratio
from 0.456 to 0.832 and ceiling from 3.1957 to 3.52, which is insufficient. The added
effect of changing the exciter saturation characteristic from curve A to curve B of
Figure 25-12 raises the response ratio to 1.126 and ceiling to a value somewhat above 4.34
per unit. These values are close, but not yet at the target values. Another trial should bring
the test performance very close to target. Figure 25-14 summarizes the results of the three
response ratio tests.
E fd
E fd Ceiling
Response Ratio Area
E fd Full Load
t = 0 t = 0.5
Figure 25-13. Evaluation of Response Ratio by Activity ERUN
25-18
CASE FOR DATA VERIFICATION
08:34
ONE MACHINE INFINITE BUS
POWER
TECHNOLOGIES,
INC.R
FRI JAN 12, 1996

RESPONSE RATIO CEILING
CHNL# 1: [EFD 1 [ ] MC 1]
5.0000 FILE: //tim/user/tfl/pag/gppp.rwf 2.0000
CHNL# 1: [EFD 1 [ ] MC 1]
5.0000 FILE: //tim/user/tfl/pag/gpp.rwf 2.0000
CHNL# 1: [EFD 1 [ ] MC 1]
5.0000 FILE: //tim/user/tfl/pag/gp.rwf 2.0000
0.90000
0.80000
Curve A
Curve A
Curve B
0.70000
0.60000
VRMAX = 1.05
VRMAX = 2.1
VRMAX = 2.1
0.50000
0.40000
TIME(SECONDS)
0.30000
0.20000
0.10000
0.0
-0.1000
Figure 25-14. Exciter Responses in Response Ratio and Ceiling Tests for IEEEX1
25-19
25.3.3 Checking of Automatically Initialized Parameters
25.3.3.1 Example Based on Model IEEEX1

The automatic calculation of those excitation system parameters that are dependent on
generator loading is intended to handle the task in a reasonable manner in the majority of
plants in a large system study. It is emphasized, however, that it is the implementation of
reasonable assumptions only, and that the results should be checked by the engineer in
critical situations.
An example of the need for a specific check arises in the example completed above. Here
the size of the amplidyne represented by the IEEEX1 model was doubled and the model
was reinitialized at full load with automatic calculation of the parameter, Ke. A DOCU
output obtained just after activity ESTR shows a positive value of Ke, Ke = 0.168. A posi-
tive value of Ke implies that the manual setting of the exciter field resistance is above the
self-excitation limit and hence that the exciter is dependent upon the amplidyne to main-
tain its steady state. This implication is not necessarily incorrect, but is contrary to the nor-
mal intention that the exciter should be in self-excited mode with the amplidyne providing
only a trimming effect.
The questionable situation can be corrected by recognizing that the key item in determin-
ing Ke is not the value of Ke itself but, rather, is the value of the amplidyne output signal.
The model data sheet, Figure 25-11, shows that this quantity is available as STATE(K+2).
The criterion for correctness of the parameter, Ke, is that the value of this state should be
near to the middle of its range. A Ke value requiring Ke to be near the upper end of its
range would correspond to an operation where the control system is using nearly all of its
control effort capability to maintain a steady state and hence has little capability to
increase excitation when this is required. (The converse applies with regard to excitation
decrease.) Such operation is bad practice and would surely not be intended. The recom-
mended approach in this case of a positive value of Ke would be to:
• Change Ke from the positive value as calculated by the IEEEX1 model to a proper neg-
ative value by placing the desired nonzero value in CON(J+7); see Figure 25-11.
• Rerun activity ESTR or STRT to reinitialize all state variables.
• Recheck the amplidyne output signal, STATE(K+2), to ensure that it is reasonably
placed in the range between VRMAX and VRMIN.
• Adjust the value of Ke in CON(J+7) and repeat the process if necessary.
25.3.3.2 Example Based on Model EXST2

A second example of the importance of checking automatically calculated parameters is
provided by the setup of model EXST2 to represent an SCPT-type excitation system where
exciting power is obtained from compounded potential and current transformers.
Figure 25-15 shows the EXST2 data sheet filled in with representative first trial data. The
value of KI, CON(J+10), has been set to zero to tell the model to determine a value for KI.
KI, unlike Ke in model IEEEX1, is a design constant and does not vary with machine load-
ing or operator action. The value of KI should, therefore, be determined once and then
fixed by placing an appropriate positive value in CON(J+10)
25-20
EXST2
IEEE Type ST2 Excitation System
ECOMP
This model is located at system bus #____ IBUS,
ETERM
machine #____ I,
ITERM
XADIFD EXST2
and STATEs starting with #____ K, VOTHSG
VUEL
and VAR #____ L.
VOEL

J 0 TR (sec)
J+1 120 KA
J+2 .15 TA (sec)
J+3 1.2 VRMAX
J+4 -1.2 VRMIN
J+5 1.0 KE
J+6 0.5 TE (>0) (sec)
J+7 .02 KF
J+8 1. TF (>0) (sec)
J+9 1.19 KP
J+10 0 KI or zero
J+11 .7 KC
J+12 4.5 EFDMAX
K Sensed VT
K+1 Regulator output, VR
K+2 Exciter output, EFD
K+3 Rate feedback integral
VAR # Description
L KI
25-21
IBUS, ’EXST2’, I, TR, KA, TA, VRMAX, VRMIN, KE, TE, KF, TF, KP, KI, KC, EFDMAX/
Figure 25-15. Initial Data Estimate for Compounded Transformer-Type Excitation System
(Sheet 1 of 2)
VREF VS
VRMAX EFDMAX
+ +
1 + KA + + 1
–  
EC   EFD
1 + sTR VERR 1 + sTA VR sTE
– + –
VRMIN 0.
VF VB
KE
sKF
1 + sTF
VT VE
V E = K P V T + jK I I T 
IT
IFD
IN = KC FEX = f(IN)
IFD VE IN FEX
If I N  0. F EX = 1
If I N  0.433 F EX = 1 – 0.577 I N
2
F EX = 0.75 – I N
IN If 0.433  I N  0.75 FEX
If I N  0.75 F EX = 1.732  1 – I N 
If I N  1 F EX = 0
Figure 0-0 (Cont). Initial Data Estimate for Compounded Transformer-Type Excitation System (Sheet
2 of 2)
Figure 25-16 shows the first execution of ESTR to initialize a generator at 0.9 power factor.
The resulting value is shown by DOCU to be 3.084, which is nearly twice the main genera-
tors quadrature synchronous reactance; a high value for KI. The second execution of
ESTR, also shown in Figure 25-16, initializes the generator at unity power factor. The ensu-
ing DOCU output shows that KI is calculated at 2.78 in this condition. These two automat-
25-22
ically calculated values of KI indicate what KI would have to be to produce the required
excitation voltage at full load with zero output, Vr, from the voltage regulating element.
This assumption is now overridden and, in Figure 25-17, the KI constant (CON(J+7) or
CON(27), in this setup, is set to 2.5; a value that will require some output from the regulat-
ing element in the steady state at high generator loads. Figure 25-17 next shows a reinitial-
ization at 0.9 pf to determine the value of regulating element output required by the value
KI = 2.5. The following DLST output shows that VR (STATE(10) in this setup) has the ini-
tial steady-state value of 0.6146. This is above one half of the regulating elements operat-
ing range; it draws attention to the value of the parameter, KC, which characterizes
commutating drop in the excitation rectifier.
Figure 25-18 shows the effect of changing KC, CON(28) in this setup, from 0.7 to 0.5. Ini-
tialization at 0.9 pf now results in a very reasonable value of 0.2328 for the output of the
regulating element and suggests that a reasonable set of data values has been reached.
A statement of typical excitation system parameters is of little value in this example unless
it is coordinated with data on the reactances and rated power factor of the main generator.
As a general rule, excitation system data stated in isolation from corresponding main gen-
erator data should be treated with considerable distrust.
25.3.4 Excitation Loop Tuning

The data checks and excitation system ceiling tests described above pertain mainly to the
sizing and output limits of the excitation system power components; they reveal little
about the correctness of parameters, such as KA and KF in IEEEX1 or EXST2, involved in
the voltage regulator loop. These parameters must be checked by a test where the result
depends on the inherent dynamic characteristics of the regulator loop, rather than on
equipment ratings and limits.
Voltage regulator parameters can be checked readily by simulating the response of the
excitation system and generator to a step change in voltage regulator reference when the
generator is running at rated speed on open circuit. This test is simulated automatically by
ESTR and ERUN when the OPEN CIRCUIT SETPOINT STEP TEST option of activity
ESTR is selected. Its use is best illustrated by an additional example.
The open-circuit step test on the generator and EXST2 excitation system model combina-
tion was run on the data refined in Section 25.3.3.2. The plotted result of the test is shown
in Figure 25-19. The excitation system is much too oscillatory. A review of Section 16.4.1
suggests that the voltage regulator gives too high a gain at the frequency of the oscillation,
about 1 Hz, and that an increased value of the rate feedback gain, KF, in the EXST2 model
would improve the tuning of the system. Figure 25-20 shows the plotted result from re-exe-
cution of the open circuit step test with the value of KF increased from 0.02 to 0.05. The
increased rate feedback gain produces the desired result of quenching the oscillatory
behavior while retaining a quick, well-tuned tracking of the step change.
25-23
ACTIVITY? estr Use ESTR to initialize

ENTER 0 FOR RESPONSE RATIO TEST generators at rated load
1 FOR OPEN CIRCUIT SETPOINT STEP TEST: 0
ENTER DEFAULT POWER FACTOR: .9 0.9 pf for first trial

ENTER BUS NUMBER, MACHINE ID, POWER FACTOR:
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E FRI JAN 12, 1996 10:35

1 1 1.0000 2.6665 90.00 43.59 0.9000 38.19 0.8990 0.4379
ENTER CHANNEL OUTPUT FILENAME: Efd = 2.67 defines rated size

DOCU is convenient way to examine value of excitation system if
ACTIVITY? docu
of KI as calculated by EXST2 model generator is rated for 0.9 pf
1 1 1- 14 1- 6
100.0 0.00000+J 0.19800 0.00000+J 0.00000 1.00000

5.00 0.050 1.00 0.050 3.50 0.00 1.8000 1.7280 0.2700 0.4500 0.1980 0.0800
S(1.0) S(1.2) KI is calculated as 3.084 to hold required excitation

0.1000 0.5000 voltage with zero voltage regulator output
** EXST2 ** BUS NAME BSKV MACH C O N ' S STATE'S VAR
1 1 15- 27 7- 10 1
TR KA TA VRMAX VRMIN KE TE
0.000 120.0 0.150 1.200 -1.200 1.000 0.500
KF TF KP KI KC EFDMAX KI VAR
0.020 1.000 1.190 0.000 0.700 4.500 3.084
KI CON is zero to force
ACTIVITY? estr automatic calculation
ENTER 0 FOR RESPONSE RATIO TEST Second trial with generator operated
1 FOR OPEN CIRCUIT SETPOINT STEP TEST: 0 at unity power factor
ENTER DEFAULT POWER FACTOR: 1

1 1 1.0000 2.1452 100.00 0.00 1.0000 57.36 0.8421 0.5394
ENTER CHANNEL OUTPUT FILENAME: Efd = 2.15 defines smaller rating required of excitation
system if generator rated power factor is raised to unity
ACTIVITY? docu

1 1 15- 27 7- 10 1
Unity of main generator
0.000 120.0 0.150 1.200 -1.200 1.000 0.500 operation would require
KI = 2.77 to hold required
0.020 1.000 1.190 0.000 0.700 4.500 2.780 excitation with zero
regulator output
Figure 25-16. Use of ESTR to Determine Value of EXST2 KI Parameter

Corresponding to Different Generator Rated Power Factors
25-24
ACTIVITY? estr
Reinitialize at 0.9 pf to check
ENTER 0 FOR RESPONSE RATIO TEST steady-state values of all
1 FOR OPEN CIRCUIT SETPOINT STEP TEST: 0
excitation system variables
ENTER DEFAULT POWER FACTOR: .9

1 1 1.0000 2.6665 90.00 43.59 0.9000 38.19 0.8990 0.4379
ACTIVITY? docu
With KI = 2.5,
voltage regulator
** EXST2 ** BUS NAME BSKV MACH C O N ' S STATE'S VAR must have nonzero
1 1 15- 27 7- 10 1 steady-state
TR KA TA VRMAX VRMIN KE TE output to hold
0.000 120.0 0.150 1.200 -1.200 1.000 0.500 required
0.020 1.000 1.190 2.500 0.700 4.500 2.500
ACTIVITY? dlst
ENTER CON RANGE: 0

ENTER VAR RANGE: 0
ENTER STATE RANGE: 7 10 Nonzero KI CON forces
ENTER ICON RANGE: 0
ENTER OUTPUT CHANNEL RANGE: 0 KI to its value
ENTER CRT PLOT CHANNEL RANGE: 0
STATES:
7: 1.000 8: 0.6146 9: 2.667 10: 0.5333E-01
DOCU shows that voltage regulator

output
Figure 25-17. Presetting of KI = 2.5 and Checking for Reasonableness

of Required Voltage Regulator Output
25-25
ACTIVITY? estr
Initialize again at 0.9 pf to
ENTER 0 FOR RESPONSE RATIO TEST
1 FOR OPEN CIRCUIT SETPOINT STEP TEST: 0 recheck voltage regulator
output needed in steady state
ENTER DEFAULT POWER FACTOR: .9

1 1 1.0000 2.6665 90.00 43.59 0.9000 38.19 0.8990 0.4379
ACTIVITY? docu

1 1 15- 27 7- 10 1
0.000 120.0 0.150 1.200 -1.200 1.000 0.500 Required excitation
KF TF KP KI KC EFDMAX KI VAR voltage is now
0.020 1.000 1.190 2.500 0.500 4.500 2.500 maintained with
voltage regulator
ACTIVITY? dlst
output of 0.24 pu
ENTER CON RANGE:
ENTER VAR RANGE:
ENTER ICON RANGE:
ENTER OUTPUT CHANNEL RANGE:
ENTER CRT PLOT CHANNEL RANGE:
STATES:
7: 1.000 8: 0.2328 9: 2.667 10: 0.5333E-01
Figure 25-18. Change of KC Parameter of EXST2 Model

to 0.5 to Reduce Commutation Drop Modeled in Excitation Rectifier Unit,
and Recheck of Steady-State Values of Excitation System Variables
25-26
13:57
POWER
TECHNOLOGIES,
INC.R FILE: gp
FRI JAN 12, 1996

CHNL# 2: [ETRM 1 [ ] MC 1]
1.1000 0.90000
CHNL# 1: [EFD 1 [ ] MC 1]
2.8000 0.80000
10.000
9.0000
8.0000
7.0000
6.0000
Terminal Voltage
Field Voltage
TIME(SECONDS)
5.0000
4.0000
3.0000
2.0000
1.0000
0.0
Figure 25-19. Plotted Response from Open-Circuit Step Test of Model EXST2
25-27
14:02
POWER
TECHNOLOGIES,
INC.R FILE: gpp
FRI JAN 12, 1996

CHNL# 2: [ETRM 1 [ ] MC 1]
1.1000 0.90000
CHNL# 1: [EFD 1 [ ] MC 1]
2.8000 0.80000
10.000
9.0000
8.0000
7.0000
Terminal Voltage
Field Voltage
6.0000
TIME(SECONDS)
5.0000
4.0000
3.0000
2.0000
1.0000
0.0
Figure 25-20. Plotted Response of Open-Circuit Step Test for Model EXST2
but with Rate Feedback Gain, KF, Increased to 0.05
25-28
Program Application Guide: Volume II Turbine Governor Data Verification
25.3.5 General Notes on Excitation 25.4 Turbine Governor Data

System Data Review Verification
The examples illustrate the principles of
excitation system data checking. The fol- 25.4.1 Governor Response Test
lowing notes summarize a number of
Activity GSTR and GRUN run simulations
points to consider when using ESTR and
of governing response of individual units
ERUN.
1. ESTR overrides channel assign- in isolation. The principal purpose of the
ments, placing field voltages and governor response test is to ensure that the
terminal voltages in output chan- governor gain and time constant parame-
nels. Consequently, while it is valid ters correspond to correctly tuned well-
to make a snapshot after use of damped response. The user initiates the
ESTR, channel assignments made
governor test with activity GSTR, which
by the user will be lost. Do
ESTR/ERUN work before the initializes each governor to a load level
setup of channels for output of specified by the user. Activity GRUN is
study results; use of activity CHAN then used to simulate the response of the
then overrides the assignments governors to a step change in load. The
that have been made by ESTR. load electrical power is held constant
2. Activity ERUN automatically prints (independent of frequency) after the step
out the response ratio of all so that the response indicates the damping
exciters at time 0.5 sec, but its due to the turbine and governor loop only.
main significance is for rotating The governors should be initialized to
machine exciters. SCR-bridge type about 0.8 per-unit load and the load step
exciters (as modeled by SCRX for
should be approximately 0.1 per unit. The
example) will generally exhibit very
high response ratios because of damping of hydro governing loops is usu-
their ability to reach ceiling output ally decreased with increasing load and
practically instantaneously. hence that response tests should normally
be made near full load for these units.
3. Response ratio/ceiling tests should
Governor response tests should be run for
be run for at least 1 sec; 2-sec runs
may be needed to allow older, at least 5 sec for steam turbine units
slower responding excitation sys- (TGOV1) and for at least 15 sec for gas tur-
tems to settle at their ceilings. bine (GAST) and hydro units (HYGOV). All
three types of units should have well
4. Open-circuit step response tests
damped response. Hydro governors will
should be run for at least 5 sec to
allow slower responding excitation generally show a somewhat greater over-
systems to reveal their dynamic shoot than steam turbine governors, but
behavior. should still be well damped, showing no
persistent oscillations. The governor test-
5. A final value of Efd exceeding 1.3
ing simulation output shows the transient
pu for a 1.05 pu terminal voltage in
the open-circuit step response test variation of turbine power. A small nega-
usually indicates suspect main tive change in hydro turbine power before
generator saturation data. it follows a positive change of load power
is normal and does not indicate incorrect
governor tuning.
25-29
Turbine Governor Data Verification Program Application Guide: Volume II
25.4.2 Governor Data Verification Example

For an example of governor data verification, consider that the data for the governor of
generator 1 was entered according to Figure 25-21, passed checking by DOCU, and must
be verified. Hydro governors always require careful attention to the values of temporary
droop (CON(J+1)) and relaxation time constant (CON(J+2)), because the values required
for these parameters vary over quite wide ranges depending upon the civil engineering
characteristics of the plant.
Figure 25-22 shows output from activity DOCU when the check option was chosen, which
indicates nothing unusual about the data assigned to the hydro turbine governor model,
HYGOV. The DOCU data review is followed in Figure 25-22 by the use of GSTR to initial-
ize the unit at 80% load in preparation for simulation of a test where the electrical load is
suddenly increased to 90% of unit rating. GSTR is followed by GRUN to run the simula-
tion. Figure 25-23 shows the plotted response of the turbine-governor to the test and reveals
oscillatory performance. This response is unsatisfactory to the point where plant operators
would not tolerate it. It indicates that the governor temporary droop, CON(J+1), and/or
relaxation time constant, CON(J+2), are incorrect even though their values fall within rea-
sonable ranges. Figure 25-24 shows the plotted response of the same GSTR/GRUN test
made after changing:
CON(J+1) from 0.25 to 0.5
CON(J+2) from 5.0 to 7.5
The step change response is now much improved, although the return of speed to nominal
is more sluggish than may be desired. It would be interesting to try another test with tem-
porary droop increased still further to 0.75 and the relaxation time returned to 5.0 sec.
25-30
HYGOV
This model is located at system bus #_____ IBUS,

machine #_____ I,
This model uses CONs starting with #_____ J, SPEED PMECH
HYGOV
Speed
and STATEs starting with #_____ K,
and VARs starting with #_____ L.

J R, permanent droop
J+1 r, temporary droop
J+2 Tr (>0) governor time
constant
J+3 Tf (>0) filter time constant
J+4 Tg (>0) servo time constant
J+5 + VELM, gate velocity limit
J+6 GMAX, maximum gate limit
J+7 GMIN, minimum gate limit
J+8 TW (>0) water time
constant
J+9 At, turbine gain
J+10 Dturb, turbine damping
J+11 qNL, no power flow
K e, filter output
K+1 c, desired gate
K+2 g, gate opening
K+3 q, turbine flow
VARs # Description
L Speed reference
L+1 h, turbine head
25-31
VAR(L) + 1 e 1 + T rs c 1 g
nref 
1 + T fs rTrs 1 + Tgs
– Velocity and
SPEED  SPEED
+
R Dturb
X
–
g – 1 q +
Tws +
q h
+ –
1. qNL
Figure 25-21. Initial Trial Data for Hydro Turbine Governor
25-32
DATA CHECK FOR GOVERNOR MODELS
BUS 1 MACHINE 1: Reminder only, OK for medium head hydro turbine

DTURB= 0.5000
Governor adjustable parameters (always deserve careful review)
1 1 28- 39 11- 14 2- 3
DOCU,CHECK shows
nothing suspicious in
R-PERM R-TEMP TR TF TG
0.050 0.250 5.00 0.050 0.500
VELM
0.200
GMAX
1.00
GMIN
0.00
TW
1.75
AT DTURB QNL
1.15 0.50 0.080
hydro turbine governor
0
ACTIVITY? gstr Activity GSTR to run hydro plant isolated step load test
ENTER INITIAL LOADING, STEP (P.U.): .8 .1 Run test from 80% load with 10% increase

1 1 1.0000 1.8371 80.00 0.00 1.0000 51.53 0.6264 0.4977
ENTER CHANNEL OUTPUT FILENAME: gs
ACTIVITY? grun Activity GRUN to run the test simulation

AT TIME = -0.017 ENTER TPAUSE, NPRT, NPLT, CRTPLT: 0,2,2
CHANNEL OUTPUT FILE IS gs Print/plot intervals commensurate with slow response characteristic of governors
-0.0167 0.00000 SPD 1 [ ] MC 1 0.80000 PMEC 1 [ ] MC 1
0.0000 -0.88697E-10 0.80000
ACTIVITY? grun
Note that GSTR overrides
prior channel assignments
CHANNEL OUTPUT FILE IS gs
0.0000 -0.88697E-10 SPD 1 [ ] MC 1 0.80000 PMEC 1 [ ] MC 1
2.0000 -0.28219E-01 0.79830
4.0000 -0.33245E-01 0.91676
6.0000 -0.70223E-02 1.0034
Figure 25-22. Use of Activity GSTR/GRUN to Check Hydro Turbine Governor Data
by Simulation of Isolated Load Step Response Test
25-33
16:31
POWER
TECHNOLOGIES,
INC.R FILE: gs
FRI JAN 12, 1996

CHNL# 2: [PMEC 1 [ ] MC 1]
1.0000 0.75000
CHNL# 1: [SPD 1 [ ] MC 1]
0.05000 -0.0500
20.000
18.000
Turbine Speed
16.000
14.000
Turbine Power
12.000
TIME(SECONDS)
10.000
8.0000
6.0000
4.0000
2.0000
0.0
Figure 25-23. Plotted Response of Governor Step Load Test

Corresponding to Figure 25-22
25-34
16:34
POWER
TECHNOLOGIES,
INC.R
FILE: gss
FRI JAN 12, 1996

CHNL# 2: [PMEC 1 [ ] MC 1]
1.0000 0.75000
CHNL# 1: [SPD 1 [ ] MC 1]
0.05000 -0.0500
20.000
18.000
16.000
14.000
Turbine Power
Turbine Speed
12.000
TIME(SECONDS)
10.000
8.0000
6.0000
4.0000
2.0000
0.0
Figure 25-24. Plotted Response of Governor Step Load Test

with Revised Governor Tuning
25-35
Parameter Ranges for Activity DOCU Program Application Guide: Volume II
25.5 Parameter Ranges for Activity DOCU

DOCU, when in the checking mode, assumes that data is specified on machine base. The
user will probably see generator inertias, damping factors, reactances, turbine governor
limits and regulating gains flagged for machines not on machine base. The following
checks are made by each standard PSS®E equipment model:
• BBGOV1
• BBSEX1
• CDC1
• CDC4 and CDC6 for Appropriate Data
• CIMTR1, CIMTR2, CIMTR3 and CIMTR4 for Appropriate Data
• CLOAD
• COMP
• COMPCC
• CPAAUX
• CRANI
• CRCMGV for Both High- and Low-Pressure Turbines
• CSTCON
• CSVGN1, CSVGN3, and CSVGN4 for Appropriate Data
• CSVGN5 and CSVGN6 for Appropriate Data
• DCTC1
• DEGOV, DEGOV1
• ESAC8B
• ESDC1A, IEEET1, IEET1A, IEEEX1, IEEX2A for Appropriate Data
• ESST4B
• EX2000
• EXAC1, EXAC1A, EXAC2, EXAC3, ESAC1A, ESAC2A, ESAC3A, ESAC6A for Appro-
priate Values
• EXAC4, EXST1, ESAC4A, ESST1A for Appropriate Data
• EXBAS
• EXDC2, ESDC2A, ESAC5A for Appropriate Data
• EXELI
• EXPIC1
• EXST2, EXST2A, ESST2A
• EXST3, ESST3A
• EXTLD2
• GAST
• GAST2A, GASTWD
25-36
Program Application Guide: Volume II Parameter Ranges for Activity DOCU
• GENROU, GENSAL, GENROE, GENSAE, GENDCO, GENCLS, GENTRA, FRECHG

for Applicable Data
• GNSCN1
• GNSCN2
• HVDCAU
• HYGOV
• IEE2ST, ST2CUT
• IEESGO
• IEEEG1, WSIEG1
• IEEEG2
• IEEEG3
• IEEEST
• IEEET3, IEEEX3 for Appropriate Data
• IEEET4, IEEET5, IEET5A, IEEEX4 for Appropriate Data
• IEEEVC
• LOEXR1
• MAXEX1, MAXEX2
• MNLEX1, MNLEX2, MNLEX3
• MTDC01, MTDC03
• OLTC1, OLPS1
• PAUX1
• PAUX2
• PIDGOV
• PSS2A
• PTIST1
• PTIST3
• REXSYS and REXSY1
• RXR
• SCRX
• SEXS
• SHAF25
• SQBAUX
• STAB1
• STAB3, STAB4
• STAB2A
• STBSVC
• TGOV1
25-37
• TGOV2
• TGOV3
• TGOV4
• TGOV5
• WEHGOV
• WESGOV
• WPIDHY
• WSHYDD, WSHYGP for Appropriate Data
GENROU, GENSAL, GENROE, GENSAE, GENDCO, GENCLS, GENTRA, FRECHG for Appli-
cable Data
1 < H < 10 0 < S1.0
0D<3 S1.0 < S1.2
xd = IMAG(ZSORCE) for GENROU,
1 < T’do < 10
GENSAL, GENDCO, FRECHG
4 × DELT < Tdo < 0.2 0.2  T’qo  1.5
4 × DELT < Tqo < 0.2 x’q < xq
xd < 2.5 x’d < x’q
x’d < 0.5 × xd xd < x’q
xq < xd 0.025  Ta  0.1
x’d < xq for GENSAL, FRECHG and
x’d = IMAG(ZSORCE) for GENTRA
GENTRA only
xd < x’d 0 < Acceleration Factor  1.0
H1 × MBASE1 = H2 × MBASE2 for
x1 < xd
FRECHG
COMP
0 < Xe  < 1
COMPCC
X1 > 0 and X2 > 0
25-38
IEEEVC
0  Rc  < 1 0 < Xc  < 1
STAB1
4 × DELT < T 4 × DELT < T4
4 × DELT < T3
STAB2A
DELT < T2 DELT < T5
DELT < T3
STAB3, STAB4
0 = Tt or 4 × DELT < Tt 4 × DELT < Tx2
4 × DELT < Tx1 4 × DELT  TC for STAB4
IEEEST
0 < IC  6 where IC is stabilizer input code
Remote bus = 0 when IC = 2 or IC > 5 where IC is stabilizer input code
0  T1  10 4 × DELT < T6 < 2
0  T2  10 0 < LSMAX < 0.3
0  T3  10 - 0.3 < LSMIN < 0
0  T4  10 0  VCU < 1.25
0 < T5 < 10 0  VCL < 1.0
IEE2ST, ST2CUT
0 < IC  6 where IC is either of the stabilizer inputs
Remote bus = 0 when IC = 2 or IC > 5 where IC is either of the stabilizer inputs
0  T1  10 0  T8  10
0  T2  10 0  T9 < 10
0  T3  10 0  T10 < 2
3 × DELT < T4  10 0 < LSMAX < 0.3
0  T5 < 10 - 0.3 < LSMIN < 0
0  T6 < 2 0  VCU < 1.25
0  T7  10 -0.1  VCL < 1.0
25-39
PSS2A
1.5  TW1  15 2 × DELT < T9  2.0
1.5  TW2  15 2 × DELT  T1  2.0
1.5  TW3  15 2 × DELT  T3  2.0
1.5  TW4  15 2 × DELT  T2  6.0
2 × DELT < T6 2 × DELT  T4  6.0
2 × DELT < T7 0 < VSTMAX < 0.99
2 × DELT < T8  2 - 0.3  VSTMIN  0
PTIST1
TP > 4 × tc T2 > 4 × tc
TF > 4 × tc T4 > 4 × tc
PTIST3
TP > 4 × tc T4 > 4 × tc
TF > 4 × tc T6 > 4 × tc or T6 = 0
T2 > 4 × tc B2 > 4 × tc or B2 = 0
STBSVC
KS1  > 0 0 < KS3
2 × DELT < TS9 0 < KS2 if second signal is used
0 < TS13
2 × DELT < TS12 if second signal is used
2 × DELT < TS14
SCRX
0.05 < TA / TB < 0.3 - 5 < EMIN  0
5 < TB < 25 2 < EMAX < 10
50 < K < 400 CSWITCH = 0 or 1
5  K × TA / TB  15 0  rc / rfd  10
0  TE < 1
SEXS
0.05 < TA / TB < 1 5  K × TA / TB  15
5 < TB < 20 EMIN = 0
20 < K  100 3  EMAX  6
0  TE < 0.5
25-40
ESDC1A, IEEET1, IEET1A, IEEEX1, IEEX2A for Appropriate Data

0  TR < 0.5 0 < KF < 0.3
10 < KA < 500 4 × DELT < TF or TF1 < 1.5
5  TF / KF  15 for ESDC1A, IEEET1,
0  TA < 1
IEEEX1, IEET1A, IEET1S
0.5 < VRMAX < 10 4 × DELT < TF2 < 1.5 for IEEET2, IEEEX2
- 10 < VRMIN < 0 5  TFN / KF  15 for IEEET2, IEEEX2
- 1  KE  1 MINIMUM(S,S2) < 0.2
4 × DELT < TE < 1
where:
S = 1 - TF1 / TE  E1 < E2 or EFDMAX
S2 = 1 - TF2 / TE  SE (E1) < SE (E2) or SE (EFDMAX)
TFN = TF1 if S2  S EFDMIN = 0 and EFDMIN < EFDMAX
TFN = TF2 if S  S2 0  TB
0  E1 0  TC
0  SE (E1) < 1 TB = 0 and TC  0
IEEET3, IEEEX3 for Appropriate Data

0  TR < 0.5 4 × DELT < TF < 1.5
10 < KA < 200 5  TF / KF  15
0  TA < 1 Kp = 1.19
0.5 < VRMAX < 1.5 0.9 < K1 < 1.1
- 1.5 < VRMIN < -0.5 1.0 < VBMAX < 4.0
4 × DELT < TE < 2 0 < KE  1.0
0 < KF < 0.3
25-41
IEEET4, IEEET5, IEET5A, IEEEX4 for Appropriate Data

0.01 < KR < 0.05 0  E1
2.0 < TRH < 100 or TRH = 0 for IEET5A 0  SE (E1) < 1
0.02 < KV < 0.10 E1 < E2
K R < KV SE (E1) < SE (E2)
2.0 < VRMAX < 10.0 0 < KA < 25
0  VRMIN < 2.0 EFDMIN < EFDMAX
4 × DELT < TE < 1.0 E2  EFDMAX
-1.0 < KE < 0.10 TR = 0 or 4 × DELT < TR < 0.5
EXDC2, ESDC2A, ESAC5A for Appropriate Data

0  TR < 0.5 - 1.0  KE  1.0
10 < KA < 500 4 × DELT < TE < 2.0 for EXDC2, ESDC2A
0  TA < 1.0 4 × DELT < TE < 1.0 for ESAC5A
0  TB 0 < KF < 0.3
0  TC 4 × DELT < TF1 < 1.5
TB = 0 and TC  0 0  TF2
0.5 < VRMAX < 10.0 0  TF3
-10 < VRMIN < 0 5.0  TF1 / KF  15.0 for EXDC2, ESDC2A
For ESAC5A:
5.0  TFN / KF  15.0 0  E1
0  SE (E1) < 1.0
MINIMUM(S,S2) < 0.2
where: E 1 < E2
S = 1 - TF1 / TE  SE (E1) < SE (E2)
S2 = 1 - TF2 / TE 
TFN = TF1 if S2  S
TFN = TF2 if S  S2
EXAC1, EXAC1A, EXAC2, EXAC3, ESAC1A, ESAC2A, ESAC3A, ESAC6A for Appropriate
Values
0 < VAMAX  10 for EXAC3, ESAC3A,
0  TR < 0.5
EXAC2, ESAC2A, ESAC6A
0  TB < 20 0 < VAMAX  15 for ESAC1A
- 500  VRMIN < 0 for EXAC2, ESAC2A,
0  TC < 20
ESAC6A
- 10  VAMIN < 0 for EXAC3, ESAC3A,
0 < KA < 1000
EXAC2, ESAC2A, ESAC3A, ESAC6A
25-42
0  TA < 10.0 - 15  VAMIN < 0 for ESAC1A

0 < VRMAX  10.0 for EXAC3, ESAC3A,
0  KB 
EXAC1A
- 10  VRMIN < 0 for EXAC3, ESAC3A,
0  KL  1.1
EXAC1A
0 < VRMAX < 15.0 for EXAC1, ESAC1A 0  KH  1.1
- 15.0 < VRMIN < 0 for EXAC1, ESAC1A 5 < VLR  10
0 < VRMAX  100 for ESAC6A only 0  KLV  1.1
- 100  VRMIN < 0 for ESAC6A only 0  KFA  1.1
2 × DELT < TE < 2 0  KR < 1.1 for EXAC3
2 × DELT <TE < 4. for ESAC3A only 0  KR < 75 for ESAC3A
0  KF  0  KN 
2 × DELT < TF < 1.5 0 < EFDN  10
0  KC  1 - 5.0 < VLV  5.0
0  KD  1 - 5 < VFEMAX  20
0 < KD  2 for ESAC6A only 0  VEMIN  1.1
0 < KE  1 0  KLI  1.0
0 < KE  2 for ESAC6A only 0 < TK < 10
0  E1 0 < VFELIM  20
0  SE (E1) < 1 0 < KH  100
E 1 < E2 0 < VHMAX  100
SE (E1) < SE (E2) 0  TH  1
0 < VRMAX  500 for EXAC2, ESAC2A,
0  TJ  1
ESAC6A
25-43
EXAC4, EXST1, ESAC4A, ESST1A for Appropriate Data

0  TR < 0.1 5  KA × TC / TB  15
0 < VIMAX  0.2 3  VAMAX  8
- 0.3 < VIMIN  0 for EXAC1, ESST1A - 8 < VAMIN  - 3
- 0.2 < VIMIN  0 for EXAC4, ESAC4A 3  VRMAX  8
0  TC < 10.0 - 8  VRMIN  - 3
4 × DELT < TB < 20 0  KC < 0.3
0  TC1 < 10 0 < KF  0.3
4 × DELT < TB1 < 20 0.3 < TF  1.5
TB > 4 × DELT or TA > 4 × DELT for EXST1 0 < KLR  5.0
50 < KA  1000 0 < ILR  5.0
0  TA < 0.5
EXST2, EXST2A, ESST2A

0  TR < 0.5 0 < KF < 0.3
10 < KA < 1000 4 × DELT < TF < 1.5
0  TA < 1 5  TF / KF  20
0.5 < VRMAX < 1.5 Kp = 1.19
- 1.5 < VRMIN < 0.5 0  KI  8.0
0 < KE  1 0 < KC < 2
4 × DELT < TE < 2 and 4 × DELT < KE < 2 1 < EFDMAX < 10
EXST3, ESST3A
0  TR < 0.5 0  KI  1.1
0 < VIMAX < 1 1 < EFDMAX < 20
- 1 < VIMIN < 0 0  KC < 1
0 < KJ < 1000 0 < XL < 0.5
0 < TC < 20 0 < VGMAX < 20
4 × DELT < TB < 20 - 90 < p < 90
0 < KA  200 0  TM < 1.0
0  TA < 1.0 0.5 < VMMAX  1.5
0.5 < VRMAX  10 - 1.5 < VMMIN < 0.5
- 10  VRMIN < 0.5 0 < KM < 1000
0  KG < 1.1 0 < VBMAX < 20
1 < Kp < 10
25-44
ESST4B
0  TR < 0.5 - 118.8  VMMIN  0
0  KPR  75 0  KG < 1.1
0  KIR  75 1  KP < 10
0.8  VRMAX  10 0  KI  1.1
- 6  VRMIN  0 1 < VBMAX < 20
0  TA < 1 0  KC < 1
0  KPM  1.2 0  XL < 0.5
0  KIM  18 - 90 <THETAP < 90
0.8  VMMAX  118
EXPIC1
0  TR < 0.5 0  KC < 2
1 < KA < 500 0  Ke  1
0  TA1 < 10 1 < EFDMAX < 10 if EFDMAX < 20
0.5 < VR1 < 10.0 if VR1 < 20 - 6 < EFDMIN  - 0,5 if EFDMIN > -20
- 6 < VR2  - 0.5 if VR2 > - 20 1  VRMAX  15 if VRMAX < 50
0  Te < 2 - 6 < VRMIN  0 if VRMIN > - 50
0  KF < 0.3 0  E1
4 × DELT < TF1 < 15 0  SE (E1) < 1
5  TF1 / KF  25 SE (E1) < SE (E2)
0  KP < 5 E1  E2
0  KI  1.1 0  TF2 < 5
EXBAS
0  TR < 0.5 0  TF2
0 < KP < 5 TB = 0 and TC > 0
0  KI < 1.1 –1  KE  1
10 < KA  4000 4DELT < TE < 10
0  TA  10 5  TF/KF  15 or 5 TF2/KF  15
0  TB < 20 KC  1
0  TC < 20 KD  2
0.5 < VRMAX < 20 0  E1
–20 < VRMIN < 0 0  SE (E1) < 1.0
4DELT < TF < 1.5 E1 < E2
0  TF1 SE(E1) < SE (E2)
25-45
ESAC8B
0  TR < 0.5 –1 < VRMIN < 1.5
10 < KP < 500 0 < TE
10 < KI < 500 –1  KE  1
10 < KD < 500 0  E1
0  TD < 0.5 0  SE (E1) < 1.0
0 < KA  1 E 1 < E2
0  TA  1 SE (E1) < SE (E2)
0 < VRMAX  10
EXELI
Tfv  0 Xe  0
Tfi  0 Tw  0
Tnv > 0 Ts1  0
Vpnf  0 Ts2  0
Dpnf  0 Smax > 0
EFDmax  EFDmin
BBSEX1
0  TF < 0.5 0 < T4
10 < K < 500 T4 = 0 and T3  0
2 × DELT <T1 <10 0.5 < VRMAX < 10
2 × DELT <T2 <10 - 10 < VRMIN < 0
0 < T3 0.5  K × T2 / T1  25
25-46
EX2000
0  KPR < 50 0  KRCC < 0.2
0.5  KIR < 15 0  TR < 0.1
0.5  VRMAX < 3 2.7  IFDREF1 < 4
- 3 < VRMIN < - 0.5 2  IFDREF2 < 3
0  KPA < 50 1.8  IFDREF3 < 2.7
0.5 < KIA  20 2.5  IFDREF4 < 4
0.7  VAMAX  1.5 2  I1 < 3
- 1.5  VAMIN  - 0.7 30  T1 < 90
0  KP  50 2  I2 < 4
5  KL  50 20  T2 < 40
Delta/2 < TE < 7 where Delta is the time 3.5  I3 < 5

step 10  T3 < 20
4  VFEMAX < 6 4  I4 < 6
0  KE  1.5 5  T4 < 9
90  KC  0.5 0  TLEAD < 3
0 < KD  0.5 0  TLAG < 0.5
0  KF1 < 1 0  KPIFD < 5
0  KF2 < 1 0.5  KIIFD < 10
2.5  E1 < 4 0.5  IFDLIMP < 1.5
0  S(E1) < 0.4 - 1.5  IFDLIMN < - 0.5
3  E2 < 5 0  IFDADVLIM < 1.5
0  S(E2) < 0.7 -10  VEMIN < 0.1
1.05  KVHZ < 1.2 0.9  REFLIMP  1.3
REXSYS and REXSY1

0  TR < 0.5 0  KII < 0.5
0  KVP < 200 KIP = 0 and KII = 0
0  KVI < 200 0  TP
KVI = 0 and KVP = 0 0.5 < VFMAX < 100
0 < VIMAX  20 - 100 < VFMIN < 0
0  TA  10 - 10  KH  10
0  TB1 < 20 - 10  KE  10
0  TC1 < 20 DEL2 < TE < 10
0  TB2 < 20 KC  1
25-47
0  TC2 < 20 KD  2
0.5 < VRMAX < 20 0  E1
- 20 < VRMIN < 0 0  SE (E1) < 1
0 < KF < 0.3 E 1 < E2
DEL2 < TF < 1.5 SE (E1) < SE (E2)
0  TF1 0  FLIMF  1
0  TF2 0 < XC < 6
0  FBF  2 0  VCMAX  10
0  KIP < 0.5
MNLEX1, MNLEX2, MNLEX3

0 < KF2 < 0.3 0 < K < 10 for MNLEX1
4 × DELT < TF2 < 1.5 0.5 < Qo < 1 for MNLEX2
5  TF2 / KF2  15 1 < R < 2 for MNLEX2
0  KM < 5 - 0.5 < Qo < 0 for MNLEX3
0  TM  4 0  B < 0.5 for MNLEX3
0  MELMAX < 0.2
TGOV1
0 < R < 0.1 0 < T2
4 × DELT < T1 < 0.5 4 × DELT < T3 < 10.0
0.5 < VMAX < 1.2 T2 < T3 /2.0
VMIN < VMAX 0  Dt < 0.5
0  VMIN < 1.0
TGOV2
0 < R < 0.1 1.0 < T3 < 10.0
4 × DELT < T1 < 0.5 0  Dt < 0.5
0.5 < VMAX < 1.2 4 × DELT < Tt < 0.5
VMIN < VMAX 4 × DELT < TA < 0.25
0  VMIN < 1.0 TA + 0.1 < TB < 50.0
0.1 < K < 0.5 TB + 1.0 < TC < 50.0
25-48
TGOV3
5.0  K  30 0 < T4  1
0  T1 < 5 - 2.0  K1  1
0  T2 < 10 4 × DELT < T5 < 10
4 × DELT < T3  1.0 0  K2 < 0.5
0.01  Uo  0.3 0  T6 < 10
- 0.3  Uc < 0 0  K3 < 0.35
0.5  PMAX  1.0 4 × DELT < TA < 0.25
0  PMIN < 0.5 TA + 0.1 < TB < 50.0
PMIN < PMAX TB + 1.0 < TC < 50.0
GAST
0 < R < 0.1 0 < Kt < 5.0
4 × DELT < T1 < 0.5 0.5 < VMAX < 1.2
4 × DELT < T2 < 0.5 0  VMIN < 1.0
4 × DELT < T3 < 5.0 VMIN < VMAX
0 < Ambient Load Limit  1.0 0  Dturb < 0.5
HYGOV
0 < R < 0.1 0 < GMAX  1.0
0 < r < 2.0 0  GMIN < 1.0
R<r GMIN < GMAX
4 × DELT < Tr < 30 0.5 < Tw < 3.0
4 × DELT < Tf < 0.1 0.8 < At < 1.5
4 × DELT < Tg < 1.0 0  Dturb < 0.5
0 < VELM < 0.3 0 < qn1 < 0.15
DEGOV, DEGOV1
0  T1 < 25.0 0  TD < 0.125
0  T2 < 0.5 0  TMAX < 1.5
0  T3 < 10 -0.05  TMIN < 0.5
15  K < 25.0 0  DROOP < 0.1
0  T4 < 25.0 0  TE < 1.0
0  T5 < 10 If T1 = 0, then T3 = 0
0  T6 < 0.5
25-49
IEESGO
0  T1 < 100 5  K1  30
0  T2 < 10 0  K2  3.0
4  DELT < T3  1.0 -1.0  K3  1.0
0  T4  1.0 0.5  PMAX  1.5
0  T5  50 0  PMIN  0.5
0  T6  1.0 PMIN < PMAX
CRCMGV for Both High- and Low-Pressure Turbines

0.5 < PMAX < 1.2 4  DELT < T4
0 < R < 0.1 4  DELT < T5
4  DELT < T1 DH = 0
4  DELT < T3
IEEEG1, WSIEG1
5.0  K  30 0  K3 < 0.5
0  T1 < 5.0 0  K4 < 0.5
0  T2 < 10.0 0  T6 < 10.0
4  DELT < T3  1.0 0  K5 < 0.35
0.01  Uo  0.3 0  K6 < 0.55
-0.3  Uc < 0 0  T7 < 10.0
0.5  PMAX  2.0 0  K7 < 0.3
0  PMIN < 0.5 0  K8 < 0.3
PMIN < PMAX 0  DB1  0.02
0 < T4  1.0 0  ERR  0.02
-2.0  K1  1 0  DB2  0.02
K2 = 0 GV1  GV2  GV3  GV4  GV5
0  T5 < 10.0
IEEEG2
5.0  K  30 0.5  PMAX  1.5
0  T1 < 100 0  PMIN  0.5
0  T2 < 10 PMIN < PMAX
4  DELT < T3  1 4  DELT  T4  5.0
25-50
IEEEG3
4  DELT  TG  1.0 0 <   1.2
4  DELT  Tp < 0.1 1.0  TR < 50
0  Uo < 0.3 4  DELT < Tw < 10
-0.3 Uc  0 0 < a1 < 1.5
0.5  PMAX  1.0 0 < a13 < 1.5
0  PMIN  0.5 0 < a21  1.5
0 <   0.1 0 < a23 < 1.5
SHAF25
Hi < 5 where i is any mass 0  Di  1.0 where i is any mass
Hi < 10 0  Kshaft i-j  200 where i and j are masses
0  Power Fractioni < 1.0 where i is any
0.4  Xd - Xd  2.5
mass
 Power Fraction = 1.0 2  Tdo  10
Exciter and generator mass numbers are valid.
TGOV4
5  K  30 0  CV#3  0.5
0  T1  5 CV#2  CV#3
0  T2  10 0  CV#4  0.75
0 < T2 and T1 = 0 CV#3  CV#4
4  DELT < T3  1 IV = 0.8 or 0.2  IV  0.7
0.01  Uo  0.3 0  IV#2  0.5
-0.3  Uc  -0.01 CV CHAR = 0.8 or 0.2  CV CHAR  0.7
0.05  KCAL  1.5 IV CHAR = 0.8 or 0.2  IV CHAR  0.7
0  T4  1 0  CV START  0.1
0.1  K1  0.5 1  CV CLOSE  10
4  DELT < T5  10 0  TIME CV1 CLOSED  5
0  K2  0.5 0  TIME CV2 CLOSED  5
0  T6  2.0 0  TIME CV3 CLOSED  5
1.1  PRMAX  1.5 0  TIME CV4 CLOSED  5
0.2  KP  1 0  IV START  0.1
0.01  KI  0.1 1  IV CLOSE  10
0  TFUEL  30 0  TIME IV1 CLOSED  5
0  TFD1  30 0  TIME IV2 CLOSED  5
25-51
0  TFD2  30 4  DELT < TRPLU  0.1

0.1  Kb  0.3 0  PLU RATE  1
150  Cb  300 0  TIMER  0.1
4  DELT < TIV  1 0  PLU UNBALANCE LEVEL  1
0.01  UOIV  0.3 4  DELT  TREVA  0.1
-0.3  UCIV  -0.01 0  EVA RATE LEVEL  1
0.01  R  0.05 0  EVA UNBALANCE LEVEL  1
0  Offset  0.60 0  Load Ref  0.5
CV = 0.8 or 0.2  CV  0.7 0  Ramp Rate  0.1
0  CV#2  0.25
BBGOV1
0  fcut < 0.01 0  T4  1
5  KS < 30 0 < K2 < 1.0
0.05 < KLS  0.3 0  T5  50
0.1  KP  2.0 0  K3  1.0
1.0  TN  10.0 0  T6  1.0
0.1  KD  2 4  DELT < T1  5
4  DELT < TD  2 0.5  PMAX  0.5
25-52
GAST2A, GASTWD
0 < W < 30 for GAST2A 0.5 < K4 < 1
0  X for GAST2A TC  T R
4  DELT < Y < 0.5 for GAST2A 0.8  MBASE  TRATE  1.05  MBASE
Z = 0 and Z  1 for GAST2A 0  T  0.05
0  KDROOP  0.1 for GASTWD 0  Cf2  1
0  KP  20 for GASTWD 10 < T3 < 25
0  KI  10 for GASTWD 1 < T4 < 5
0  KD  20 for GASTWD 100 <t < 600
Max > Min 1 < T5 < 5
0.5 < Max < 1.8 500 < af1 < 1000
-0.2 < Min < 0.1 300 < bf1 < 700
0.5 < K3 < 1 -1 < af2 < 1
0.5 < a < 50 0.9 < bf2 < 1.5
4  DELT < b < 2 700 < TR < 1050
0  c  1.01 0.1 < K6 < 0.5
0.05 <f < 0.8 0  TD < 0.5
0 Kf  1.0 0  CR < 0.5
0.05 < K5 < 0.5 0  TCD < 0.5
WPIDHY
0.05  TREG < 5.0 0.3  GATMX  1
0 < REG < 0.1 0  GATMN  0.5
0  KP < 10 0.5  TW  3.0
0  KI  5 0.5  PMAX  1.1
0  KD  5 0  PMIN  0.5
4  DELT < TA  2 0 < D < 0.5
4  DELT < TB  2 G0  G1  G2
0  VELMX  1 P1  P2  P3
-1  VELMN  0
25-53
TGOV5
5  K  30 0  RMAX  0.5
0  T1 < 5 -0.5  RMIN  0
0  T2 < 10 0  LMAX  2.2
4  DELT < T3  1 0  LMIN  1.1
0.01  Uo  0.3 LMAX > LMIN
-0.3  Uc < 0 0.1  C1  1
0.5  VMAX  2 -1  C2  20
0  VMIN < 0.5 0  C3  1
VMAX > VMIN 0  B  30
0 < T4  1 50  CB  300
0  K1  1 0  K1  0.5
K2 = 0 20  T1  300
0  T5 < 10 0  TR  100
0  K3 < 0.5 0  TR1  10
0  K4 < 0.5 0  CMAX  2.4
0  T6 < 10 0  CMIN  0.5
0  K5 < 0.35 CMAX > CMIN
0  K6 < 0.55 TD = 0 or 0.5  TD  100
0  T7 < 10 0  TF  30
0  K7 < 0.3 0  TW  15
0  K8 < 0.3 0 < PSP  2
0  K9 < 1 0  TMW
0  K10  1 KL = 0 or KL = 1
0  K11  1 KMW = 0 or KMW = 1
0  K12  10 KL = 0 and KMW = 0
0  K13  1 0  PE  0.2
0  K14  5
25-54
WSHYDD, WSHYGP for Appropriate Data

0  DB1  0.02 0  Tt  0.5
0  ERR  0.02 0  KG  5.0
0  Td < 10.0 0  TP  0.5
0 < K1  1.0 0  VELOPEN  0.01
0  Tf  1.0 0  VELCLOSE  0.01
0  K2  1.0 0  PMAX  1.05
0  KD  1 0  PMIN  0.5
0  KP  200 PMIN < PMAX
0  KI  200 0  DB2  0.02
0  R  0.5 GV1  GV2  GV3  GV4  GV5
WESGOV
0 < TP  0.25 0  T1 < 0.2
0 < TC  0.25 0.2  T2  0.6
0 < DROOP < 0.10 0.15  ALIM < 0.4
10  KP < 25 0  Tpe < 0.2
1.0  TI < 10
WEHGOV
0 < R-PERM-GATE < 0 0  DTURB < 0.5
0 < R-PERM-PE < 0.1 0.5 < TW < 3
4  DELT < Tpe < 0.5 0 < DBAND < 0.005
1 < KP < 10 0  DIPV < 0.1
1 < KI < 20 0  DICM < 0.1
0 < KD < 20 GATE1  GATE5 ascending value order
4  DELT < TD < 0.1 0.8 < GATE5 < 1.3
4  DELT < TP < 0.2 FLOWG1  FLOWG5 ascending value order
4  DELT < TDV < 0.2 0.8 < FLOWG5 < 1.3
4  DELT < TG <1 FLOWP1  FLOWP10 ascending value order
0 < GTMXOP < 0.3 0.8 < FLOWP10 < 1.3
-0.3 < GTMXCL < 0 PMECH1  PMECH10 ascending value order
0  GMAX < 1 0.8 < PMECH10 < 1.3
-0.1 < GMIN < 0.3
25-55
PIDGOV
2  DELT < Ta < 1 P1  P3 ascending value order
2  DELT < Tb < 1 0.01  Velmax  0.3
2  DELT < TW < 1 -0.3  Velmin < 0
G0  G2 ascending value order
PAUX1
4 DELT < Tr Kc  0
0  TD  10.5  DELT MIN < MAX
PAUX2
4 DELT < TR 0 < T2
0  TD  9.5  DELT 4 DELT < T3
Kc  0 4 DELT < T4
0 < T1 MIN < MAX
SQBAUX
4 DELT < T2 0  TD  10.5  DELT
0 < B2 0  TL < 0.25
IMIN < IMAX
CPAAUX
Cm  0 PMIN < PMAX
4  DELT < TB < 1 0 < PMAX < 50
4  DELT < TA < 1 -50 < PMIN < 0
GNSCN1
50 < Angle < 200
GNSCN2
1.0 < Power Threshold < 1.2
25-56
RXR
0 < R1 R10  R9 < R8
X 1 < X2 R10  R6
R3 < R2 X6 < X10
0 < X3 X10 < X9
X 1 < X4 X11 < X12
R5 < R4 R13 < R12
0 < X5 R14 < R11
R6 < R7 X14 < X13
X 7 < X8
LOEXR1
0  TZ1 < 6 0 < REACH2 < 0.6
0 < REACH1 < 0.4 80 < ANGLE2  100
80 < ANGLE1  100 -0.4  CDIST2 < 0
-0.2  CDIST1 < 0 0.5  VPICKUP
3  TZ2 < 20
CDC1
T1 = 0 or 4  DELT < T1 < 0.2 I1 < I2 < I3
T2 = 0 or 4  DELT < T2 < 0.2 V 2 < V3
0 < IMIN < 300 0.05 < DELTI < 0.2
CDC4 and CDC6 for Appropriate Data

ALFDY  ALFMX from power flow C1  C2  C3
GAMDY  GAMMX from power flow V 1  V2  V3
TVDC = 0 or 4  DELT < TVDC < 0.2 0  TCMODE < 0.2
TIDC = 0 or 4  DELT < TIDC < 0.2 TVRDC = 0 or 4  DELT < TVRDC < 0.2
0 < C  300
CLOAD
0  % LARGE MOTOR  100 0  % CONSTANT POWER  100
0  % SMALL MOTOR  100 0 < KP < 5
0  % TRANSFORMER EXC CUR  5 0  R < 0.5
0  % DISCHARGE LIGHTING  100 0 < x < 1.0
25-57
CSVGN1, CSVGN3, and CSVGN4 for Appropriate Data

50  K < 1000 VMIN < VMAX
4  DELT < T3 0  CBASE < 500
0  RMIN < MBASE 0.1  VOV  0.5
1.0  VMAX  2.0 0.4  VL  0.8
-0.5  VMIN 1.05  VH  2.0
CSVGN5 and CSVGN6 for Appropriate Data

TS1 < 0.4 sec -2BMAX < BMIN
VE MAX < 0.3 per unit 2  DELT < TS6  0.2
TS2 < 2 sec 0  DV < 0.5
0 < TS3 < 5 sec VE MIN < 0.3
TS4 < 2 sec VMAX < 0.3
TS5 < 5 sec VMIN < 0.3
50  KSVS < 1000 BIAS < 1.0
0 < KSD < 1000 DV2 < 0.3
0 < BMAX < 10 | BSHUNT | < 2.0
BMAX < BMAX TDELAY < 0.1
BMIN < BMIN
CIMTR1, CIMTR2, CIMTR3 and CIMTR4 for Appropriate Data

0.2 < T < 10 x1 < x
T = 0 or 4  DELT < T < 0.2 0  E1
0.5 < H < 10.0 E1 < E2
0 < x < 5.0 S(E1) < S(E2)
x < 0.5  x 0.2 < D < 5 for motor
T  0 ZSORCE = x
x < x
MTDC01, MTDC03
DYi  ANGMAXi 0 < c - i  300
TVACi = 0 or 4  DELT  TVACi < 0.2 c1-i  c2-i  c3-i
TVDCi = 0 or 4  DELT  TVDCi < 0.2 v1-i  v2-i  v3-i
TIDCi = 0 or 4  DELT  TIDCi < 0.2 0  TCMODE  0.2
25-58
Program Application Guide: Volume II Other Data Assumptions Made By Models
HVDCAU
0 < IC  5 where IC is input code T4  0 if T5 > 0
MININ < MAXIN B = 0 or B = 1
A = 0 or A = 1 MINOUT < MAXOUT
T3  0 if T2 > 0 4  DELT  T1 < 0.2
CRANI
0  T1 < 5 0 < TW < 2
0  T2 < 5 XMIN < XMAX
1 < T3 < 20
EXTLD2
0 < KP  0.1 0 < KQ  0.1
1.0  PMLTMX  1.2 1.0  QMLTMX  1.2
0.8  PMLTMN  1.0 0.8  QMLTMN  1.2
DCTC1
5 < TDR 5 < TDI
TSDR < TDR TDI < TCI
TCR < TSDR TSDI < TDI
OLTC1, OLPS1
5 < TD TC < TSD
TSD < TD
MAXEX1, MAXEX2
1.0 < EFDRATED < 4.0 40  TIME1 < 120
1.0 < EFD1 < 2.0 40  TIME2 < 120
1.0 < EFD2 < 2.0 40  TIME3 < 120
1.0 < EFD3 < 2.0
CSTCON
T3 > 0 Xt > 0
T4 > 0 VMAX > VMIN
25.6 Other Data Assumptions Made By Models

In certain models, not only the time constant itself but its effective time constant may
cause numerical instability in the modified Euler used in RUN. In these models, an inner
25-59
Other Data Assumptions Made By Models Program Application Guide: Volume II
loop has been added in which ten time steps have been simulated rather than one. For
these models, the time constants are still set to zero if they are less than or equal to 0.5 
DELT. Table 2 lists these models and the time constants.
Table 2: Time Constants Assumed Zero if Less Than or Equal to 1/2 Times DELT
Model Time Constant

EXAC1 TR, TA
IEET1S TR
EXAC2 TR, TA
EXST1 TR, TB, TA
IEEX2A TR, TA
CSVGN5 TS1, TS5
The only area where PSS®E models make data assumptions is with small time constants.
In Chapter 11 - Fault Analysis, numerical integration stability was discussed and it was
stated that T should be kept smaller than 20 to 25% of the shortest time constant in the
process being simulated. This guideline is not always practical, especially because data
often comes from a central bureau and small filter time constants are often entered. To
avoid numerical instability in the modified Euler used in RUN, PSS®E will assume a time
constant to be zero if it is less than or equal to twice T in certain models. Table 3 lists the
models and time constants where this assumption is made.
Table 3: Time Constants Assumed Zero if Less Than or Equal to Two Times DELT
Model Time Constant

STAB3 TT
STAB4 TT, Td, Te
IEE2ST T1, T2
ST2CUT T1, T2
STBSVC TS7, TS10
IEEET1 TR, TA
IEEET3 TR, TA
IEET1A TA
IEEEX1 TR, TA
IEEEX3 TR, TA
IEEEX4 TR
EXDC2 TR, TA
EXAC1A TR, TA
25-60
Program Application Guide: Volume II Other Data Assumptions Made By Models
Model Time Constant

EXAC3 TR, TA
EXAC4 TR, TA, TB
EXST2, EXST2A TR, TA
EXST3 TR, TB, TA
SCRX TE
SEXS TE
EXPIC1 TR, TA4, TF2
EXBAS TR, TA
ESAC8B TR, TA, TD
EXELI Tfv, Tfi, Tw, Ts1, Ts2
BBSEX1 TF, T2
MNLEX1, MNLEX2, MNLEX3 TR, TM, TA2, TA1
TGOV3 T 1, T 4, T 6
TGOV4 T1, T4, T6, TFUEL, TFD1, TFD2
TGOV5 TMW, TF, TW, T1, T4, T5, T6, T7
IEESGO T 1, T 4, T 6
IEEEG1 T 1, T 4, T 5, T 6, T 7
IEEEG2 T1
DEGOV T5, T6
BBGOV1 T 4, T 5, T 6, T 1
GAST2A, GASTWD b, f, TCD
WEHGOV Tpe, TP, TDV
WESGOV Tpe, T1, T2
WPIDHY TREG
CSVGN5 TS1, TS5
CSVGN1, CSVGN3, CSVGN4 T5
CDC1 T1, T2
CDC4 TVDC, TIDC
CDC6 TVDC, TIDC, TVRDC
MTDC01, MTDC03 TVACi, TVDCi, TIDCi
PAUX2 TR
SQBAUX TL
25-61
Other Data Assumptions Made By Models Program Application Guide: Volume II
Model Time Constant

HVDCAU T1
CRANI T1
25-62
Chapter 26
Linear Analysis
Chapter 26 - Linear Analysis
26.1 Overview
The linear system functions that are described in this chapter are designed for studies on
small disturbance (dynamic) stability of power systems. The activities handle the state
space matrix determination, eigenvalue, and eigenvector calculations and frequency
response predictions.
The functions are not designed to solve any specific problem but rather comprise an arse-
nal of computational tools that can be directed by the user in the solution of a wide class of
problems associated with the dynamic response of power systems.
The basic premise in the design of these functions is that the intimate control over the
application of analytical tools be retained by the user. So as throughout PSS®E, the results
of each stage of the computation are available to the user for interpretation in order to
decide whether to proceed to the next step.
The capacity of these functions has been set at 20 input and 50 output variables.
LSYSAN auxiliary program, binary matrix file incompatibility with older versions of
PSS®E. Because PSS®E-30 allows longer bus names, up to 18 characters verses 12
characters maximum in older versions, the binary matrix file that is produced by activity
ASTR will not be compatible with earlier versions of PSS®E. Likewise, ASTR binary
files produced by older versions of PSS®E cannot be read in by the BCAS activity of
LSYSAN in PSS®E-30.
26.1.1 State Space Formulation of a Linear System

Execution of activity ASTR in the dynamics section of PSS®E is usually the first step a
user would perform. This converts the system modeled by the user's CONEC and CONET
and data contained in the working files into the following linear equations
x· = Ax + Bu (1.1)
v = Hx + Fu (1.2)
where:
A, B, H, F = Constant matrices.
26-1
Linear Analysis PSS®E 33.4
Overview Program Application Guide: Volume II
x = Real state vector.

u = Real input vector.
v = Vector of real system outputs.
This is the state space formulation, and equations (1.1) and (1.2) are called state equations.
Assume the following:
The number of states = n.
The number of inputs = m.
The number of outputs = k.
Then,
x is an nxl vector.
u is an mxl vector.
v is a kxl vector.
A is an nxn matrix.
B is an nxm matrix.
H is a kxn matrix.
F is a kxm matrix.
26.1.2 Eigenvalues
The solution of the state equations (1.1) and (1.2) can be obtained by taking the Laplace
Transforms. Thus:
 x  0  + Bu  s    Adj  SI – A  
x  s  =  SI – A  – 1  x  0  + Bu  s   = -----------------------------------------------------------------------
Det  SI – A  (1.3)
and
Y  s  = H  SI – A  – 1  x  0  + Bu  s   + Fu  s 
(1.4)
where:
I = Is the identity matrix of the same dimension as matrix A.
Det (SI - A) = Denotes the determinant of matrix (SI - A).
Adj (SI - A) = Denotes the adjoint of matrix (SI - A).
X(0) = Is the state at time t = 0.
The poles of X(s) are the roots of:

Det  SI – A  = 0 (1.5)
or equivalently the roots of:

Det  A – SI  = 0 (1.6)
26-2
PSS®E 33.4 Linear Analysis
Program Application Guide: Volume II Overview
The values of S that satisfy equation (1.6) are called eigenvalues. The eigenvalues, typi-
cally denoted by , are functions of the state matrix A. The number of eigenvalues is equal
to the number of states. For a stable system, all the eigenvalues have negative real parts.
26.1.3 Eigenvectors
For each eigenvalue (i), there exists a vector ti that satisfies an equation of the form
At i =  i t i (1.7)
The vector ti is called the eigenvector of A associated with eigenvalue i. The dimension-
ality of vector ti is equal to the number of states.
26.1.4 Linear Independence of Eigenvectors

Let t1, ..., tn be the eigenvectors corresponding to eigenvalues 1, ..., n. If C1, ..., Cn are
scalars (real or complex numbers), then the sum
C 1 t 1 + ... + C n t n
is called a linear combination of the eigenvectors.

The vectors t1, ..., tn are said to be linearly independent if no one of them is a linear combi-
nation of the others.
26.1.5 Modal Formulation of a Linear System

It is common in linear system analysis to transform the original set of system differential
equations into the modal form
y· = y + u (1.8)
x = Ty (1.9)
where:
 = Diagonal matrix of eigenvalues.
y = Complex system state vector in modal domain.
T = Complex matrix for which the columns are eigenvectors of A.
 = T-1B = Complex input matrix.
Each element (yi) of vector y is termed a mode of the system.

Activities EIGN and DIAG of program LSYSAN obtain the L, , and T matrices given A,
B, H, and F. The L, , and T matrices themselves convey a great deal of information on
inherent system dynamics characteristics. The alternative to the modal or time-domain
information provided by (1.8) and (1.9) is to produce plots of the frequency response of key
outputs of the system when selected inputs are subjected to sinusoidal perturbations.
The frequency response activity FRRS, obtains the complex output vector, v', correspond-
ing to a unit input signal by using the equations
y = –   – jI  – 1 u' (1.10)
26-3
Overview Program Application Guide: Volume II
s = Ty (1.11)
v' = Hs + Fu' (1.12)
where:
u' = u vector with one unity element and all others equal to zero.
s = Complex state vector in frequency domain.
v' = Complex output vector in frequency domain.
26.1.6 Participation Factors

In practice, it is helpful to determine the magnitude of the influence of a particular state,
and hence device, on a given mode of oscillation. A controller should be located at a
device that strongly influences the mode to be controlled.
Before introducing the notion of a participation factor, it is necessary to define the right
and left participation matrices.
The matrix T, which appears in equation (1.11), is referred to as the right eigenmatrix.
Let:
Z = T –1 (1.13)
The matrix Z is referred to as the left eigenmatrix.

A participation factor is defined as
P ik = t ik z ki
where:
tik = The ith element of the kth column of matrix T.
zki = The ith element of the kth row of matrix Z.
The participation factor Pik is the measure of the influence of state i on mode k.
The description that follows presents one logical sequence a user may use in solving a
problem. Figure 26-1 is a tree graph that shows prerequisites for each activity. If the user
runs an inappropriate activity a message is printed.
26-4
Program Application Guide: Volume II PSS®E Interface - Inferring Linear System Matrices
ASTR
LSYSAN
MCAS/BCAS
MCHG
LSTR MLIS EIGN
LRUN DIAG EIGP PLEI
FRRS PRFA
BODE POLR
Figure 26-1. Tree Structure Showing Flow of Functions
26.2 PSS®E Interface - Inferring Linear System Matrices

The basic function of ASTR is to sequentially perturb the differential and algebraic equa-
tions representing the power system and subsequently infer the state space equations
describing the power system dynamic performance.
Executing activity ASTR allows the user to obtain the A, B, H, and F matrices describing
the small perturbation behavior of a system that has already been set up. Linear analysis is
illustrated by a small system. The power flow raw data file of the system is shown in
Figure 26-2. Figure 26-3 shows the dynamics raw data file.
Consider a system described by the ordinary differential equations:
26-5
PSS®E Interface - Inferring Linear System Matrices Program Application Guide: Volume II
x = f  x  u v  (1.14)
and the algebraic equations:

v = g  x u v  (1.15)
that constitute a dynamics set-up.

Let xo be an x vector giving a valid equilibrium condition, that is, dx/dt is zero when eval-
uated at x = xo. Now, let xj be a vector in which all elements except the jth one are identical
to those of xo, and in which the jth element differs from the corresponding value by xj.
Then, if xj is sufficiently small, the jth columns of A and H may be estimated from
j dx o
 dx
------ –  ------ = A j x j (1.16)
 dt   dt 
v j – v o = H j x j (1.17)
where (dx/dt)o and (dx/dt)j are evaluated from (1.14) using xo and xj; vj and vo are system
outputs evaluated on the basis of xo and xj; and Aj and Hj are the jth columns of A and H.
The columns of B and F are estimated similarly by evaluating (dx/dt) and v with individ-
ual elements perturbed in the input vector u.
ASTR, using the user's set-up first does a standard initialization to set up the equilibrium
condition (see PSS®E Program Operation Manual, Section 14.15, Building a State Variable
Matrix for Linear Dynamic Analysis (LSYSAN)). During this initialization all equipment values
are calculated for each model's boundary conditions.
Activity ASTR then instructs the user to ENTER THE LARGEST DERIVATIVE
CHANGE ALLOWED. If no value is entered, the derivative defaults to 0.01. ASTR then
checks the values of all state variable derivatives, DSTATE, against a user entered limit.
All elements of the DSTATE array should ideally be zero, but small deviations may exist
in practice due to a small error in the initial estimate in the power flow or in the x, v, and u
vectors. ASTR is different from the standard STRT in that the actual DSTATE is checked
against the perturbation specified by the user rather than a default value. A new specified
perturbation is required because it is felt that the method used should ideally be in equilib-
rium for utmost accuracy and engineering judgment is required to decide if it is close
enough. If any DSTATE is above the derivative change allowed, it is printed and the activ-
ity is aborted.
26-6

PSS®E 33.4
0 100.00 / TUE, OCT 27 1998 11:33
TWO MACHINE EXAMPLE WITH INFINITE BUS
1 'INF BUS ' 0.0000 3 0.000 0.000 1 1 1.00000 0.0000 1

2 'GEN 1 ' 0.0000 2 0.000 0.000 1 1 1.00000 16.3116 1
3 ' ' 0.0000 1 0.000 0.000 1 1 1.01351 13.0876 1
4 'GEN 2 ' 0.0000 2 0.000 0.000 1 1 1.00000 22.7724 1
0
3 1 1 1 1 400.000 -100.000 0.000 0.000 0.000 0.000 1
0
1 1 -150.000 1.877 9999.000 -9999.000 1.00000 0 100.000 0.00000 0.10000 0.00000 0.00000 1.00000 1 100.0
9999.000 -9999.000 1 1.0000
2 1 50.000 -15.443 9999.000 -9999.000 1.00000 0 200.000 0.00000 0.25000 0.00000 0.00000 1.00000 1 100.0
9999.000 -9999.000 1 1.0000
4 1 500.000 -7.259 9999.000 -9999.000 1.00000 0 600.000 0.00000 0.30000 0.00000 0.00000 1.00000 1 100.0
9999.000 -9999.000 1 1.0000
0
1 3 1 0.00000 0.15300 0.13000 0.00 0.00 0.00,,, 0.00000 0.00000 0.00000 0.00000 1 0.0 1 1.0000
2 3 1 0.00000 0.11400 0.10000 0.00 0.00 0.00,,, 0.00000 0.00000 0.00000 0.00000 1 0.0 1 1.0000
3 -4 1 0.00000 0.03410 0.20000 0.00 0.00 0.00,,, 0.00000 0.00000 0.00000 0.00000 1 0.0 1 1.0000
0
0
0
0
0
0
0
0
PSS®E Interface - Inferring Linear System Matrices

0
26-7
0
0
Figure 26-2. Small Example System for Linear Analysis Examples
1 'GENCLS' 1 0. 0./
2 'GENROU' 1 6. .035 1. .07 4. 0. 2.2 2.0 .4 .47 .25 .1 .12 .3667/
4 'GENROU' 1 4.8 .035 1.5 .07 3.2 0. 1.8 1.75 .37 .47 .3 .15 .12 .3667/
2 'SCRX' 1 .1 10. 150. .05 -3.2 4.0 1.0 0/
4 'SCRX' 1 .1 10. 150. .05 -3.2 4.0 1.0 0/
Linear Analysis
Figure 26-3. Dynamics Raw Data File for Setup of Simulation Model of Small Example System
The activity then instructs the user to enter an output filename. It then instructs the user to
enter single or range of state variables to be used and whether or nor the user would like to
perturb them. The overall flow chart of the process is shown in Figure 26-4.
The elements of the output vector, v, are defined as the first fifty (or less) output channels
the user selects.
Before determining the A,H or B,F matrices, ASTR allows the user to designate up to 20
system variables as inputs for the B matrix. It does this by instructing the user:
ENTER INPUT QUANTITY CATEGORY AND
AMOUNT OF PERTURBATION:
0=EXIT 1=EFD 2=PMECH
3=VOTHSG 4=VREF 5=VAR
Each input is specified in response to questions by ASTR by typing in

1. Input quantity and amount of perturbation, 0.01 per unit is default for each input. This
perturbation default value is reset for each quantity.
2. Bus number and machine ID for machine quantities or VAR number. If a bus entered
is not found or a machine is not available, an appropriate message is printed and the
user is asked to specify the bus and machine again. No checks are made on VAR
inputs.
Note that specifying EFD for a machine with an exciter or PMECH for a machine with a
governor as inputs is invalid. The activity however does not print any messages indicating
the invalid requests because the information is not known by the program.
The user response is the output file for storing the matrices for use by the linear analysis
program. A carriage return exits the activity without storing the matrices.
The activity now instructs the user to
ENTER SINGLE STATE OR STARTING AND ENDING STATES TO BE
INCLUDED IN MATRICES
because the user may not always want all the states represented. If the user specifies the
last state greater than the number used, the question is repeated. After the user input, the
activity writes the matrices or parts of the matrix to the appropriate file.
If state space matrices corresponding to different system conditions are required, the sys-
tem may now be updated (i.e., new power flow case, different system parameters via
ALTR, etc.) and ASTR can be rerun.
26-8
From
Figure
11.10
Perturb
STATEi
Mode=3
Call CONEC
Call SITER
Mode=2
Call CONEC
Calculate
Aji, j=1, NINTEG
Mode=3
Call CONEC
Calculate
Hji, j=1, NCHANS
Reset
STATE
i=i+1
Is
No
i > NINTEG
?
Yes
Do Reset
You Want No A, B, H, F
An Input Matrices
? File
Yes
Mode=3
Calculate Call CONEC
Fj, i, j=1, NCHANS Call SITEC
Mode=2
Call CONEC
Mode=3 Calculate
Call CONEC Bi, i, j=1, NINTEG
Figure 26-4. Flow Chart of Activity ASTR
26-9
$ PSSDS4
INITIATED AT DYNAMICS ENTRY POINT ON THU, OCT 15 1992 16:40
ACTIVITY? RSTR SNAP2 Pick up Snapshot

SNAPSHOT SNAP2.SNP WAS SAVED ON TUE, OCT 13 1992 09:22

46 18 0 0 5
ACTIVITY? LOFL Pick up power flow
ACTIVITY RTRN TO RETURN TO DYNAMICS--ACTIVITY?

CASE CVLF2
CASE CVLF2.SAV WAS SAVED ON TUE, OCT 13 1992 09:22
ACTIVITY? STRT Do a standard STRT to

make sure all is okay

1 INF BUS 1 1.0000 0.9968 -150.00 -14.51-0.9954 -8.65 0.0823-1.5048
2 GEN 1 1 1.0000 1.0839 50.00 -15.47 0.9553 38.38 0.0507 0.2567
4 GEN 2 1 1.0000 1.9046 500.00 -7.33 0.9999 69.60 0.6592 0.5100
ENTER SNAPSHOT FILENAME:
ACTIVITY? DOCU,ALL Do a DOCU to note

state variables
2 FOR A FILE 3 FOR KMW_UP
4 FOR QMS_LG 5 FOR HARD COPY TERMINAL
Figure 26-5. Creating the Linear Analysis Matrices

(Sheet 1 of 4)
26-10
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E THU, OCT 15 1992 16:40

REPORT FOR ALL MODELS AT ALL BUSES BUS 1 [INF BUS ] PLANT MODELS
** GENCLS ** BUS NAME BSKV MACH C O N ' S STATE'S We will want to delete
1 INF BUS 1 1- 2 1- 2 these state variables
MBASE Z S O R C E X T R A N GENTAP H DAMP
100.0 0.00000+J 0.10000 0.00000+J 0.00000 1.00000 0.00 0.000
ENTER 0 TO END LIST, 1 FOR NEXT PAGE: 1

REPORT FOR ALL MODELS AT ALL BUSES BUS 2 [GEN 1 ] PLANT MODELS

2 GEN 1 1 3- 16 3- 8
200.0 0.00000+J 0.25000 0.00000+J 0.00000 1.00000

6.00 0.035 1.00 0.070 4.00 0.00 2.2000 2.0000 0.4000 0.4700 0.2500 0.1000
S(1.0) S(1.2)
0.1200 0.3667

2 GEN 1 1 31- 38 15- 16 FED

0.100 10.000 150.0 0.050 -3.20 4.00 1.0 0.00
REPORT FOR ALL MODELS AT ALL BUSES BUS 4 [GEN 2 ] PLANT MODELS

4 GEN 2 1 17- 30 9- 14
600.0 0.00000+J 0.30000 0.00000+J 0.00000 1.00000

4.80 0.035 1.50 0.070 3.20 0.00 1.8000 1.7500 0.3700 0.4700 0.3000 0.1500
S(1.0) S(1.2)
0.1200 0.3667 We will also want to drop
these state variables
4 GEN 2 1 39- 46 17- 18 FED

0.100 10.000 150.0 0.050 -3.20 4.00 1.0 0.00
ACTIVITY? DLST List output channels
Figure 0-0 (Cont). Creating the Linear Analysis Matrices (Sheet 2 of 4)
26-11

ENTER CON RANGE: 0
ENTER VAR RANGE: 0
ENTER STATE RANGE: 0
ENTER ICON RANGE: 0

ENTER OUTPUT CHANNEL RANGE: 1,5

OUTPUT CHANNELS:
#:ADDR 1: 4002 2: 4003 3: 24002 4: 24003 5: 16002

P ELECTRICAL P ELECTRICAL SPD DEVIAT'N SPD DEVIAT'N FIELD VLTAGE
IDENT P 1GEN 1 P 1GEN 2 SPD1GEN 1 SPD1GEN 2 EFD1GEN 1
VALUE 0.50000 5.0000 0.00000E+00 0.00000E+00 1.0839
ACTIVITY? ASTR Now create linear analysis matrices


1 INF BUS 1 1.0000 0.9968 -150.00 -14.51-0.9954 -8.65 0.0823-1.5048
2 GEN 1 1 1.0000 1.0839 50.00 -15.47 0.9553 38.38 0.0507 0.2567
4 GEN 2 1 1.0000 1.9046 500.00 -7.33 0.9999 69.60 0.6592 0.5100
ENTER LARGEST DERIVATIVE CHANGE ALLOWED (DEFAULT IS 0.01):
ENTER MATRIX OUTPUT FILE NAME (0 TO EXIT): MATOUT
18 STATES IN USE. ENTER SINGLE STATES OR STARTING AND

ENDING STATES TO BE INCLUDED IN MATRICES (CR FOR FIRST 18)
ENTER STARTING AND ENDING STATES: 3 16
ENTER STARTING AND ENDING STATES: 0
ENTER AMOUNT TO PERTURB SELECTED STATES (DEFAULT IS 0.01):
5 OUTPUT CHANNELS IN USE. ENTER SINGLE CHANNELS OR STARTING AND

ENDING CHANNELS TO BE INCLUDED AS LINEAR SYSTEM OUTPUTS (CR FOR FIRST 5)
26-12
Program Application Guide: Volume II Accessing Linear System Activities/Picking Up the Matrices
ENTER STARTING AND ENDING OUTPUT CHANNELS: 1 5
ENTER STARTING AND ENDING OUTPUT CHANNELS:

ENTER INPUT QUANTITY CATEGORY AND AMOUNT OF PERTURBATION:
0 = EXIT 1 = EFD 2 = PMECH
3 = VOTHSG 4 = VREF 5 = VAR: 2

0 = EXIT 1 = EFD 2 = PMECH Two of the inputs are
mechanical powers

3 = VOTHSG 4 = VREF 5 = VAR: 4 The third input is

ENTER BUS NUMBER, MACHINE ID, 'IDENTIFIER': 2 reference voltage
Figure 26-5 shows the dialog to create the matrices of the small example system. In this
problem the user had full representation of the generators and excitation systems. For the
linear analysis calculations it was not useful to keep the excitation states of the machine on
bus 4 or those of the classical generator. The interactive dialog reflects this decision for the
starting and last state variables.
All the linear system activities keep the state variable in the order they are assigned in
CONEC. When the user enters a starting integrator greater than one, the order is kept but
the index will be reduced so that the first retained state is numbered one.
Note that activity ASTR does not calculate bus frequency because the perturbation tech-
nique used to build the matrices is a t-minus, t-plus calculation in which frequency will not
change. Approximations obtained by using bus angle change tend to produce meaningless
results. Therefore, any model such as a stabilizer where the input is bus frequency will not
contribute to the eigenvalues (it will be as if the device is not there).
26.3 Accessing Linear System Activities/Picking Up the Matrices

Before any of the linear system activities can be run, it is necessary for the user to leave
PSS®E and enter the linear system activities program, LSYSAN. If a frequency response
calculation has been run and the output stored, the user can then run a BODE or POLAR
plot. Most likely, however, the linearized matrices describing the problem will be read into
memory. This is accomplished by either activities MCAS or BCAS. Activity MCAS reads
data in ASCII form while BCAS reads binary data.
The program activity selector will print the following:
26-13
Listing the A, B, H, and F Matrices Program Application Guide: Volume II
ACTIVITY?
Activity MCAS asks the user to

ENTER MATRIX INPUT FILE NAME:
This file was either created by activity ASTR or by the user via the CRT or another pro-
gram. User input files are allowed so that all the linear system analysis activities are made
available to the user for other type studies. The file should contain the following:
IHA
IHB
NINTEG, NOUTS, NINPUT
A matrix
B matrix
H matrix
F matrix
IDENT
NINDEN
where:
IHA and IHB = Records describing the case. Each line may contain up to 60 characters
that are entered in columns one through 60.
NINTEG = Number of integrators.
NOUTS = Number of outputs.
NINPUT = Number of inputs.
IDENT = The 12 character identifiers of the outputs.
NINDEN = The 12 character identifiers of the inputs.
These values are entered in free format except for IDENT and NINDEN for which the for-
mat is 5(1X,A). Matrices should be input by rows. Activity BCAS should be used if the
matrix data is in binary form. Note that the activities of LSYSAN use frequency response
techniques that require that no row of the A matrix is comprised of only zero elements. If
the user reads in an A matrix that has rows that are composed of only zero elements,
LSYSAN will eliminate the zero rows and the user will be informed.
26.4 Listing the A, B, H, and F Matrices

Often the user will want to list these matrices to check their reasonableness. This is per-
formed via activity MLIS. The activity allows all the standard PSS®E output destinations.
Figure 26-6 shows the dialog to start program LSYSAN, to pick up and to list the matrices
for the simple system created in Figure 26-5. Figure 26-7 lists the matrices for our sample
system.
26-14
Program Application Guide: Volume II Calculating Eigenvalues and Eigenvectors
26.5 Calculating Eigenvalues and Eigenvectors

Activity EIGN uses EISPACK1 subroutines to determine the eigenvalues and eigenvec-
tors. The eigenvalues are saved in order of decreasing magnitude of their imaginary part,
followed by eigenvalues with zero imaginary part in order of decreasing magnitude of real
part. The eigenvectors are saved in order corresponding to the original state variable order
with the corresponding eigenvalues. The eigenvectors are normalized so that the largest
element of each is unity.
The activity will print messages if the algorithms cannot find the eigenvalues or eigenvec-
tors. The messages indicate that the results of EIGN cannot be trusted as input to activities
using the eigensystem of A. If such messages appear, the user should check the validity of
the A matrix, and should check for a very wide range of magnitudes of elements of A, or
other evidence of ill-conditioning.
26.6 Documenting the Eigensystem

Activity EIGP automatically lists the eigenvalues of the system. The activity allows all the
standard PSS®E output destinations. As an option, the user can list eigenvectors also.
Eigenvalues and eigenvectors listings can be limited to those associated with synchronous
machine angles. These are often the ones in which the user is most concerned about. When
listing eigenvalues, the damping factor is also printed. This is done so that the users atten-
tion can focus on the lightly damped modes.
Activity PLEI will give a planar plot of all the eigenvalues. Activity PLEI also allows the
user to plot individual elements of each eigenvector. This is often very valuable when try-
ing to analyze the modes of oscillation of the system.
Figure 26-8 shows the dialog used to create, list, and plot the eigenvalues and eigenvectors
as well as their values for our sample system. Figure 26-9 shows a planar plot for our sam-
ple system while Figure 26-10 shows a plot of three elements of the second eigenvector.
1 Matrix Eigensystem Routines - EISPACK Guide, Springer-Verlag, 1974.
26-15
Documenting the Eigensystem Program Application Guide: Volume II
LSYSAN
LINEAR SYSTEM ANALYSIS PROGRAM INITIATED ON FRI OCT 16, 1992 09:49
ACTIVITY? BCAS
ENTER MATRIX INPUT FILE NAME (0 TO EXIT): MATOUT
ACTIVITY? MLIS

2 FOR A FILE 3 FOR PRINTER-1
4 FOR PRINTER-2 5 FOR HARD COPY TERMINAL
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E FRI OCT 16, 1992 09:50
ROW BUS X-- NAME --X MC MODEL STATE

1 2 GEN 1 1 GENROU K
2 2 GEN 1 1 GENROU K+1
7 4 GEN 2 1 GENROU K
8 4 GEN 2 1 GENROU K+1 Identification of the system state
13 2 GEN 1 1 SCRX K
14 2 GEN 1 1 SCRX K+1
Figure 26-6. Linking to Linear Analysis Activities Picking-Up and Listing Matrices
26-16
Program Application Guide: Volume II Documenting the Eigensystem
A MATRIX COLUMNS STARTING AT 1

1 -0.8637 -0.2818E-01 0.3030 -0.4142E-01 0.0000 -0.1314 0.1156 0.2954E-01 0.5392E-01 0.3347E-01
2 -0.1323 -3.975 -0.1323 1.701 0.0000 0.9114 -0.1776 0.3285 -0.8289E-01 0.3723
3 21.42 -0.2763 -35.73 -0.4064 0.0000 -7.511 6.603 1.688 3.081 1.912
4 0.2106 10.71 0.2108 -19.53 0.0000 7.766 -1.514 2.799 -0.7064 3.172
5 -0.6338E-01 0.7128E-01 -0.6338E-01 0.1045 -0.3094E-01 -0.2089 0.7151E-01 -0.4899E-01 0.3337E-01 -0.5551E-01
6 0.0000 0.0000 0.0000 0.0000 377.0 0.0000 0.0000 0.0000 0.0000 0.0000
7 0.3031E-01 -0.2214E-01 0.3031E-01 -0.3250E-01 0.0000 0.7358E-01 -0.8736 -0.7249E-01 0.3219 -0.8225E-01
8 0.5382E-01 0.4841E-01 0.5381E-01 0.7099E-01 0.0000 -0.6282E-01 -0.1777 -2.421 -0.8241E-01 1.034
9 0.9380 -0.6853 0.9379 -1.006 0.0000 2.277 24.89 -0.6309 -30.29 -0.7151
10 0.6151 0.5534 0.6151 0.8114 0.0000 -0.7178 0.6677 12.45 0.3117 -16.37
11 -0.1697E-02 -0.2611E-01 -0.1695E-02 -0.3830E-01 0.0000 0.5619E-01 -0.1318 -0.8287E-02 -0.6149E-01 -0.9412E-02
12 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
13 -0.2351E-01 -0.9470E-02 -0.2351E-01 -0.1389E-01 0.0000 0.1592E-02 -0.1220E-01 -0.9561E-02 -0.5691E-02 -0.1084E-01
14 -78.35 -31.57 -78.36 -46.31 0.0000 5.307 -40.67 -31.87 -18.97 -36.12
1 0.0000 0.5827E-01 0.0000 0.1667
2 0.0000 -0.7292 0.0000 0.0000
3 0.0000 3.330 0.0000 0.0000
4 0.0000 -6.214 0.0000 0.0000
5 0.0000 0.1511 0.0000 0.0000
6 0.0000 0.0000 0.0000 0.0000
7 0.0000 -0.1819 0.0000 0.0000
8 0.0000 0.6362E-01 0.0000 0.0000
9 0.0000 -5.628 0.0000 0.0000
10 0.0000 0.7272 0.0000 0.0000
11 -0.1289 -0.1103 0.0000 0.0000
12 377.0 0.0000 0.0000 0.0000
13 0.0000 0.4822E-02 -0.1000 0.0000
14 0.0000 16.07 3000. -20.00
a. A Matrix
B MATRIX COLUMNS STARTING AT 1

1 0.0000 0.0000 0.0000
2 0.0000 0.0000 0.0000
3 0.0000 0.0000 0.0000
4 0.0000 0.0000 0.0000
5 0.1250 0.0000 0.0000
6 0.0000 0.0000 0.0000
7 0.0000 0.0000 0.0000
8 0.0000 0.0000 0.0000
9 0.0000 0.0000 0.0000
10 0.0000 0.0000 0.0000
11 0.0000 0.1562 0.0000
12 0.0000 0.0000 0.0000
13 0.0000 0.0000 0.9000E-01
14 0.0000 0.0000 300.0
b. B Matrix
H MATRIX COLUMNS STARTING AT 1

1 1.014 -1.140 1.014 -1.673 0.0000 3.343 -1.144 0.7838 -0.5340 0.8882
2 0.6515E-01 1.003 0.6514E-01 1.470 0.0000 -2.158 5.062 0.3182 2.361 0.3614
3 0.0000 0.0000 0.0000 0.0000 1.000 0.0000 0.0000 0.0000 0.0000 0.0000
4 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
5 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
1 0.0000 -2.418 0.0000 0.0000
2 0.0000 4.237 0.0000 0.0000
3 0.0000 0.0000 0.0000 0.0000
4 1.000 0.0000 0.0000 0.0000
5 0.0000 0.0000 0.0000 1.000
c. H Matrix
F MATRIX COLUMNS STARTING AT 1

1 0.0000 0.0000 0.0000
2 0.0000 0.0000 0.0000
3 0.0000 0.0000 0.0000
4 0.0000 0.0000 0.0000
5 0.0000 0.0000 0.0000
d. F Matrix
Figure 26-7. Sample System Matrix
26-17
ACTIVITY? eign The A matrix is in memory; run EIGN

ACTIVITY? eigp List eigenvalues and eigenvectors
ENTER DESIRED OUTPUT OF EIGEN ANALYSIS RESULTS:
0 TO EXIT 1 FOR EIGENVALUES
2 FOR EIGENVECTORS 3 FOR EIGENVALUES AND EIGENVECTORS
4 FOR EIGENVALUES AND EIGENVECTORS OF SYNCHRONOUS MACHINE ANGLES: 3

PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E FRI DEC 03, 1993 10:30
EIGENVALUES:
NO. REAL IMAG DAMP FREQ

1 -1.2322 8.8473 0.13794 1.4081
2 -1.2322 -8.8473 0.13794 1.4081
3 -0.22455 4.2638 0.52591E-01 0.67860

4 -0.22455 -4.2638 0.52591E-01 0.67860
5 -0.63916 0.59114 0.73415 0.94082E-01

6 -0.63916 -0.59114 0.73415 0.94082E-01
ENTER 0 TO EXIT, 1 FOR NEXT PAGE: 1
EIGENVALUES:
NO. REAL TIME CONSTANT

7 -36.673 0.27268E-01
8 -30.123 0.33197E-01
9 -20.246 0.49393E-01
10 -18.908 0.52888E-01
11 -16.427 0.60874E-01
12 -2.1915 0.45631
13 -1.3913 0.71877
14 -0.15747 6.3503
ENTER EIGENVECTOR NUMBER (0 FOR ALL, -1 TO EXIT): 5 List the elements of eigenvector #5
EIGENVALUE 5: REAL= -0.63916 IMAG= 0.59114

DAMP= 0.73415 FREQ= 0.94082E-01
ROW X---------------- VECTOR ELEMENT ----------------X STATE MODEL BUS X-- NAME --X ID
MAGNITUDE PHASE REAL IMAGINARY
1 0.20242 -114.94 -0.85369E-01 -0.18354 K GENROU 2 GEN 1 1
2 0.20078 32.187 0.16992 0.10695 K+1 GENROU 2 GEN 1 1
3 0.19922 -123.56 -0.11013 -0.16601 K+2 GENROU 2 GEN 1 1
4 0.22236 37.007 0.17756 0.13384 K+3 GENROU 2 GEN 1 1
5 0.84297E-03 -173.60 -0.83773E-03 -0.93924E-04 K+4 GENROU 2 GEN 1 1
6 0.36502 49.162 0.23870 0.27616 K+5 GENROU 2 GEN 1 1
7 0.16010E-01 137.34 -0.11773E-01 0.10850E-01 K GENROU 4 GEN 2 1
8 0.27666E-01 -124.37 -0.15616E-01 -0.22837E-01 K+1 GENROU 4 GEN 2 1
9 0.15954E-01 175.29 -0.15900E-01 0.13091E-02 K+2 GENROU 4 GEN 2 1
10 0.33390E-01 -117.67 -0.15508E-01 -0.29570E-01 K+3 GENROU 4 GEN 2 1
11 0.18656E-03 -168.45 -0.18279E-03 -0.37354E-04 K+4 GENROU 4 GEN 2 1
12 0.80785E-01 54.314 0.47126E-01 0.65616E-01 K+5 GENROU 4 GEN 2 1
13 0.83905E-02 -56.851 0.45881E-02 -0.70249E-02 K SCRX 2 GEN 1 1
14 1.0000 0.00000 1.0000 0.00000 K+1 SCRX 2 GEN 1 1
ENTER EIGENVECTOR NUMBER (0 FOR ALL, -1 TO EXIT): -1
Figure 26-8. Obtaining and Listing Eigenvalues and Eigenvectors

(Sheet 1 of 2)
26-18
Program Application Guide: Volume II Documenting the Eigensystem
ACTIVITY? plei

0 = NONE 2 = HP 7221A
5 = TEKTRONIX 4014 W/EGM 6 = TEKTRONIX 4662
7 = TEKTRONIX 4663 11 = GRPG FILE
25 = X Window (B&W) 27 = HP 7470A
28 = HP 7475A 31 = QMS LASERGRAFIX
33 = TEKTRONIX 4111 37 = TEKTRONIX 41XX FILE
38 = HP-GL FILE 39 = X Window
40 = TEKTRONIX 4010/4014 FILE 41 = POSTSCRIPT
99 = INDE. PLOT FILE
SELECT PLOTTING DEVICE [,PARM FILE]: 41
ENTER FIRST OPTIONAL 60 CHARACTER TITLE:
planar plot
ENTER SECOND OPTIONAL 60 CHARACTER TITLE:
ENTER 1 FOR PLANAR PLOT OF EIGENVALUES

2 TO PLOT SELECTED EIGENVECTORS
3 TO PLOT SELECTED EIGENVECTORS OF SYNCHRONOUS MACHINE ANGLES: 1 Do a planar plot first
ENTER MINIMUM, MAXIMUM FOR ABSCISSA: -40 5
ENTER MINIMUM, MAXIMUM FOR ORDINATE: -10 10
ENTER 25 CHARACTER LABEL:
eigenvalues
ENTER:
0 = RETURN
1 = SELECT NEW PLOTTING DEVICE
2 = NEW PLOT ON THE POSTSCRIPT
3 = REPEAT PLOT ON NEW DEVICE: 2
plot an eigenvector
ENTER 1 FOR PLANAR PLOT OF EIGENVALUES

2 TO PLOT SELECTED EIGENVECTORS
3 TO PLOT SELECTED EIGENVECTORS OF SYNCHRONOUS MACHINE ANGLES: 2
ENTER EIGENVECTOR NUMBER (0 FOR NEW PLOT): 2 Plot eigenvector #2
EIGENVALUE 2: REAL= -1.2322 IMAG= -8.8473
DAMP= 0.13794 FREQ= 1.4081
EIGENVECTOR:
MAGNITUDE PHASE REAL IMAGINARY STATE MODEL BUS X-- NAME --X ID
1 0.28007E-01 -52.091 0.17208E-01 -0.22097E-01 K GENROU 2 GEN 1 1
2 0.20720 87.941 0.74441E-02 0.20706 K+1 GENROU 2 GEN 1 1
3 0.23212 -162.48 -0.22136 -0.69878E-01 K+2 GENROU 2 GEN 1 1
4 0.49586 37.328 0.39430 0.30068 K+3 GENROU 2 GEN 1 1
5 0.23695E-01 -97.929 -0.32684E-02 -0.23468E-01 K+4 GENROU 2 GEN 1 1
6 1.0000 0.00000 1.0000 0.00000 K+5 GENROU 2 GEN 1 1
7 0.18566E-01 90.381 -0.12343E-03 0.18565E-01 K GENROU 4 GEN 2 1
8 0.13855E-01 -100.32 -0.24812E-02 -0.13631E-01 K+1 GENROU 4 GEN 2 1
9 0.12508 15.539 0.12051 0.33507E-01 K+2 GENROU 4 GEN 2 1
10 0.45545E-01 -161.30 -0.43142E-01 -0.14599E-01 K+3 GENROU 4 GEN 2 1
11 0.87780E-02 80.882 0.13910E-02 0.86671E-02 K+4 GENROU 4 GEN 2 1
12 0.37047 178.81 -0.37039 0.76871E-02 K+5 GENROU 4 GEN 2 1
13 0.47342E-03 -8.3200 0.46844E-03 -0.68505E-04 K SCRX 2 GEN 1 1
14 0.67312 -74.584 0.17894 -0.64890 K+1 SCRX 2 GEN 1 1
ENTER EIGENVECTOR ELEMENT TO PLOT (0 FOR NO MORE): 2
ENTER EIGENVECTOR ELEMENT TO PLOT (0 FOR NO MORE): 3 Choose these elements
ENTER EIGENVECTOR ELEMENT TO PLOT (0 FOR NO MORE): 9
ENTER EIGENVECTOR ELEMENT TO PLOT (0 FOR NO MORE): -1
ENTER VECTOR SCALING OPTION:
0 TO USE ALL VECTORS
1 TO USE ONLY VECTORS TO BE PLOTTED: 1
#2
ENTER:
0 = RETURN
2 = NEW PLOT ON THE POSTSCRIPT
3 = REPEAT PLOT ON NEW DEVICE: 0 No other plots
ENTER NUMBER OF COPIES (0 TO 5), DEVICE NAME FOR PS DEVICE: 1
Figure 0-0 (Cont). Obtaining and Listing Eigenvalues and Eigenvectors (Sheet 2 of 2)
26-19
Figure 26-9. Plot of Eigenvalues
26-20
Program Application Guide: Volume II Modal Form Transformation of Activity DIAG
Figure 26-10. Plot of an Eigenvector
26.7 Modal Form Transformation of Activity DIAG

This activity calculates the matrix  of equation (1.8) corresponding to the basic linear sys-
tem equation (1.1). While the EIGN activity constructs the  and T matrices,
activity DIAG completes the transformation of (1.1) and (1.2) into the modal form (1.8) and
(1.9).
The  matrix is given by:
 = T –1  B
26-21
Calculating and Listing Participation Factors Program Application Guide: Volume II
DIAG checks for isolated natural modes that may cause the frequency response calcula-
tion to be inaccurate and prints an appropriate message. If the error message is received,
the user should carefully check the model and data and be cautious of results of
activity FRRS. Note the eigenvalues and eigenvectors previously calculated are correct.
DIAG also checks the accuracy of the transformation by computing the residual terms of
(B - T). These are printed if they exceed a tolerance of 0.001.
This activity can only be run after EIGN had been completed with no error messages.
The activity requires that the eigenvectors are linearly independent (see Section 26.1.4 for
the definition of linear independence). If two eigenvalues have very nearly equal values,
the eigenvectors produced by EIGN may be linearly dependent. This produces a singular
T matrix.
If this occurs, the user will see the message
CANNOT DIAGONALIZE THE MATRIX---ABORT ACTIVITY
As an option the user can list at all the standard PSS®E output destinations both  and .
Both  and  are saved even if the user specifies that no output is desired.
Figure 26-11 lists  and  for our sample system.
26.8 Calculating and Listing Participation Factors

Activity PRFA calculates the participation factors associated with the user-specified
modes. The activity selectively lists the participation factors according to magnitude.
PRFA allows all the standard PSS®E output destinations.
As a prerequisite activity PRFA requires that the matrix eigenvalues and eigenvectors
have been calculated (activity EIGN has been run) and the matrices have been diagonal-
ized (activity DIAG has been run). If they have not, an appropriate message will be
printed.
Figure 26-12 shows the dialog used to calculate and list the participation factors for mode
4.
Note that the largest participation factors are for STATE(K+4) and STATE(K+5) of the
GENROU model at bus 4. These are the speed and angle states for this model. Placing a
device, such as a stabilizer, at this machine would have the greatest effect on mode 4.
26-22
Program Application Guide: Volume II Calculating and Listing Participation Factors
ACTIVITY? DIAG

NABLA THETA
-0.123E+01 J 0.885E+01 -0.378E+00 J-0.198E+01 0.534E+00 J 0.251E+01

0.307E+00 J-0.314E+00
-0.123E+01 J-0.885E+01 -0.378E+00 J 0.198E+01 0.534E+00 J-0.251E+01
0.307E+00 J 0.314E+00
-0.224E+00 J 0.426E+01 -0.674E+00 J-0.216E+01 -0.251E+01 J-0.736E+01
0.119E+01 J-0.677E+00
-0.224E+00 J-0.426E+01 -0.674E+00 J 0.216E+01 -0.251E+01 J 0.736E+01
0.119E+01 J 0.677E+00
-0.639E+00 J 0.591E+00 0.232E+00 J-0.118E+01 0.138E+01 J 0.423E+00
-0.841E+00 J 0.580E+01
-0.639E+00 J-0.591E+00 0.232E+00 J 0.118E+01 0.138E+01 J-0.423E+00
-0.841E+00 J-0.580E+01
-0.367E+02 J 0.000E+00 -0.102E+01 J 0.278E-06 0.856E+00 J-0.351E-06
0.651E+01 J-0.131E-04
-0.301E+02 J 0.000E+00 0.372E+00 J-0.273E-07 -0.144E+01 J 0.727E-06
0.241E+01 J 0.122E-04
-0.202E+02 J 0.000E+00 0.117E+02 J-0.256E-06 -0.116E+02 J-0.115E-06
-0.320E+02 J 0.295E-04
-0.189E+02 J 0.000E+00 -0.840E+01 J-0.355E-06 0.136E+02 J-0.134E-05
0.280E+03 J-0.322E-04
-0.164E+02 J 0.000E+00 -0.111E+01 J 0.158E-06 0.238E+01 J 0.317E-07
0.471E+02 J-0.447E-05
-0.219E+01 J 0.000E+00 -0.246E+01 J-0.158E-07 0.345E+00 J 0.290E-06
-0.475E+01 J-0.441E-05
-0.139E+01 J 0.000E+00 -0.656E+00 J-0.371E-07 0.145E+00 J 0.102E-06
-0.380E+00 J-0.589E-06
-0.157E+00 J 0.000E+00 0.289E-01 J 0.266E-07 0.999E+00 J-0.117E-06
-0.419E+00 J-0.617E-06
Note that the eigenvalues have not changed
Figure 26-11. Obtaining Modal Equivalents for Sample System
26-23
Frequency Response Calculations Program Application Guide: Volume II
ACTIVITY? prfa

ENTER MODE NUMBER (0 TO EXIT): 4 Select mode #4
ENTER OUTPUT CODE FOR PARTICIPATION FACTORS FOR MODE 4
0 = NO OUTPUT
1 = ALL FACTORS IN DESCENDING ORDER OF MAGNITUDE
2 = ONLY LARGEST FACTORS: 1
NORMALIZED PARTICIPATION FACTORS FOR MODE 4: -0.22455 -4.2638
FACTOR ROW STATE MODEL BUS X-- NAME --X ID

1.00000 11 K+4 GENROU 4 GEN 2 1
0.99887 12 K+5 GENROU 4 GEN 2 1
0.38487 5 K+4 GENROU 2 GEN 1 1
0.38473 6 K+5 GENROU 2 GEN 1 1
0.10884 7 K GENROU 4 GEN 2 1
0.02742 1 K GENROU 2 GEN 1 1
0.01883 9 K+2 GENROU 4 GEN 2 1
0.01233 14 K+1 SCRX 2 GEN 1 1
0.01214 13 K SCRX 2 GEN 1 1
0.01075 3 K+2 GENROU 2 GEN 1 1
0.00704 2 K+1 GENROU 2 GEN 1 1
0.00437 8 K+1 GENROU 4 GEN 2 1
0.00302 4 K+3 GENROU 2 GEN 1 1
0.00161 10 K+3 GENROU 4 GEN 2 1
ENTER MODE NUMBER (0 TO EXIT): 0
Figure 26-12. Obtaining and Listing Participation Factors
26.9 Frequency Response Calculations

Activity FRRS calculates the frequency response of selected system outputs to unit sinu-
soidal perturbations of the selected system inputs. The outputs and inputs were defined
during the execution of either activity BCAS or MCAS and both are reflected in the H and
B matrices. They may be changed only by completely rebuilding the A, B, H, and F matri-
ces via activity ASTR or by editing the matrix input file.
To illustrate frequency response calculations, the linear system was modified to include
only the two states (15 and 16) of the exciter for the machine at bus #2. The inputs and
outputs were not changed.
The matrices that comprise the new linear system are shown in Figure 26-13. The eigenval-
ues and eigenvectors of the linear system are shown in Figure 26-14.
FRRS computes the complex vector v'(j) resulting from a sinusoidal perturbation of each
individual element of u' at a frequency of  radians per second. In our sample system, the
outputs are:
v'1 = Electric power of the machine at bus #2.
v'2 = Electric power of the machine at bus #4.
26-24
Program Application Guide: Volume II Frequency Response Calculations
v'3 = Speed deviation of the machine at bus #2.

v'4 = Speed deviation of the machine at bus #4.
v'5 = Field voltage of the exciter for the machine at bus #2.
The system input variables for our sample system are:

u'1 = Mechanical power of the machine at bus #2.
u'2 = Mechanical power of the machine at bus #4.
u'3 = Reference voltage of the exciter for the machine at bus #2.
FRRS deposits the output-input quantities directly into the plotting file so that the fre-
quency response ratios, vi(j)/uj(j) may be displayed in Polar or Bode plots.
Frequency inputs are specified by the user by entering a lower frequency, upper frequency,
and an increment ratio x. The value of x must be greater than unity. The value of fre-
quency is then scanned logarithmically, that is
 new =  old  x
FRRS compares the new value of each output for each new frequency value with the cor-
responding previous output. If any has changed in magnitude ratio by more than .3 or
phase by more than 10 degrees, the frequency change is halved and the frequency
response is recalculated. For cases where a smooth plot is not obtained because of sharp
resonances, the user will have to run to a specific frequency, reduce the increment, and
then run to just past the resonance where the increment could again be increased.
26-25
Frequency Response Calculations Program Application Guide: Volume II
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS®E MON OCT 19, 1992 14:43
ROW BUS X-- NAME --X MC MODEL STATE The linear system incorporates only the two
1 2 GEN 1 1 SCRX K
2 2 GEN 1 1 SCRX K+1
states of the exciter for the machine at bus 2

1 -0.1000 0.0000
2 3000. -20.00
B MATRIX COLUMNS STARTING AT 1

1 0.0000 0.0000 0.9000E-01
2 0.0000 0.0000 300.0

1 0.0000 0.0000
2 0.0000 0.0000
3 0.0000 0.0000
4 0.0000 0.0000
5 0.0000 1.000
F MATRIX COLUMNS STARTING AT 1

1 0.0000 0.0000 0.0000
2 0.0000 0.0000 0.0000
3 0.0000 0.0000 0.0000
4 0.0000 0.0000 0.0000
5 0.0000 0.0000 0.0000
Figure 26-13. The Matrices of the Linear System Used in the Frequency Calculations
26-26
Program Application Guide: Volume II Frequency Response Calculations
EIGENVALUES:
NO. REAL TIME CONSTANT

1 -20.000 0.50000E-01 Both eigenvalues have negative real parts
2 -0.10000 10.000
ENTER EIGENVECTOR NUMBER (0 FOR ALL, -1 TO EXIT): 1
EIGENVALUE 1: REAL= -20.000

TIME CONSTANT= 0.50000E-01
ROW X---------------- VECTOR ELEMENT ----------------X STATE MODEL BUS X-- NAME --X
ID
1 0.00000 0.00000 0.00000 0.00000 K SCRX 2 GEN 1
1
2 1.0000 0.00000 1.0000 0.00000 K+1 SCRX 2 GEN 1
1
ENTER EIGENVECTOR NUMBER (0 FOR ALL, -1 TO EXIT): 2
EIGENVALUE 2: REAL= -0.10000

TIME CONSTANT= 10.000
ROW X---------------- VECTOR ELEMENT ----------------X STATE MODEL BUS X-- NAME --X
ID
1 0.66333E-02 0.00000 0.66333E-02 0.00000 K SCRX 2 GEN 1
1
2 1.0000 0.00000 1.0000 0.00000 K+1 SCRX 2 GEN 1
1
Figure 26-14. The Eigenvalues and Eigenvectors of the Linear System

Used in the Frequency Response Calculations
While calculating the frequency response, FRRS allows the user to see one input/output
pair. For the pair the real part, imaginary part, magnitude and angle are printed.
As a prerequisite activity FRRS requires that the matrix eigenvalues and eigenvectors
have been calculated (activity EIGN has been run) and the matrices have been diagonal-
ized (activity DIAG has been run). If they have not, an appropriate message will be
printed. When invoked, activity FRRS instructs the user to specify the output device for
the desired input/output pair to scan. The activity then instructs the user to
ENTER OUTPUT FILE NAME:
The user responds with the name of the output file to store all input/output pair informa-
tion from the subsequent frequency response calculations. If no filename is specified, the
activity is aborted.
The activity then instructs the user to
ENTER OUTPUT NO., INPUT NO. FOR PRINTING (-1 TO EXIT)
The user must then enter valid input and output numbers, zeroes, or -1 to exit the activity.
The default values for the input and output are zero (no printing). The activity then
instructs the user to
ENTER MIN OMEGA, MAX OMEGA, INC FACTOR (-1 TO EXIT)
26-27
Plotting Results Program Application Guide: Volume II
After the user has entered the values, the activity then applies its calculations until the
maximum frequency input has been reached. At that point the activity instructs the user to
ENTER NEW MAX OMEGA, INCREMENT FACTOR
If no new increment factor is specified, the previous one input is assumed. The new fre-
quency specified must be greater than the current value of frequency. A carriage return
should be entered to exit the activity. Activity FRRS only responds to the interrupt control
code AB. It forces a pause by setting the maximum frequency to the current value of fre-
quency, overriding the value entered. In addition, if the activity was invoked in the
response file mode (via IDEV, response file) the response file is closed and subsequent
user commands are taken from the user's terminal. Figure 26-15 shows the dialog for our
sample system.
26.10 Plotting Results

Activities BODE and POLR (Polar) create plots of the frequency response predictions
obtained using FRRS. The dialog is straightforward and similar to that of auxiliary pro-
gram PSSPLT. Figure 26-16 shows the dialog to obtain a Bode plot for the exciter output,
field voltage (EFD), at bus #2 when the reference voltage, Vref, is the input. Figure 26-17 is
the resulting plot. Note that this linear system incorporates only the states of the SCRX
model for the machine at bus #2. The data for this model, given in Figure 26-3, should
result in breaks at the following frequencies:
 = 0.1, which corresponds to 1/TB
 = 1.0, which corresponds to 1/TA
 = 2.0, which corresponds to 1/Te
This is consistent with the Bode magnitude plot in Figure 26-17.
26-28
Program Application Guide: Volume II Plotting Results
ACTIVITY? FRRS

1

EXAMFROUT
ENTER OUTPUT NO., INPUT NO. FOR PRINTING (-1 TO EXIT):
5 3 Ask for output 5, input 3
ENTER MIN OMEGA, MAX OMEGA, INC FACTOR (-1 TO EXIT):
.001, 10000, 1.1
OUTPUT 5 EFD1GEN 1
INPUT 3 VRF1GEN 1
OMEGA REAL IMAG MAGNITUDE ANGLE
1 0.0010 0.1126E+01 0.3271E-01 0.1126E+01 0.1664E+01
2 0.0011 0.1126E+01 0.3598E-01 0.1126E+01 0.1831E+01
3 0.0012 0.1126E+01 0.3958E-01 0.1126E+01 0.2014E+01
4 0.0013 0.1126E+01 0.4354E-01 0.1127E+01 0.2215E+01
5 0.0015 0.1126E+01 0.4790E-01 0.1127E+01 0.2437E+01
6 0.0016 0.1126E+01 0.5269E-01 0.1127E+01 0.2680E+01
7 0.0018 0.1126E+01 0.5796E-01 0.1127E+01 0.2947E+01
8 0.0019 0.1126E+01 0.6376E-01 0.1128E+01 0.3241E+01
9 0.0021 0.1126E+01 0.7014E-01 0.1128E+01 0.3564E+01
10 0.0024 0.1126E+01 0.7715E-01 0.1129E+01 0.3919E+01
11 0.0026 0.1126E+01 0.8487E-01 0.1129E+01 0.4309E+01
12 0.0029 0.1126E+01 0.9336E-01 0.1130E+01 0.4738E+01
13 0.0031 0.1127E+01 0.1027E+00 0.1131E+01 0.5208E+01
14 0.0035 0.1127E+01 0.1130E+00 0.1133E+01 0.5724E+01
15 0.0038 0.1127E+01 0.1242E+00 0.1134E+01 0.6290E+01
16 0.0042 0.1128E+01 0.1367E+00 0.1136E+01 0.6911E+01
17 0.0046 0.1128E+01 0.1503E+00 0.1138E+01 0.7591E+01
18 0.0051 0.1129E+01 0.1653E+00 0.1141E+01 0.8335E+01
19 0.0056 0.1129E+01 0.1819E+00 0.1144E+01 0.9149E+01
20 0.0061 0.1130E+01 0.2000E+00 0.1148E+01 0.1004E+02
. . . . . .
. . . . . .
. . . . . .
. . . . . .
290 6637.7700 0.1300E-03 -0.4520E-01 0.4520E-01 -0.8984E+02

291 7301.5474 0.1075E-03 -0.4109E-01 0.4109E-01 -0.8985E+02
292 8031.7021 0.8882E-04 -0.3735E-01 0.3735E-01 -0.8986E+02
293 8834.8721 0.7340E-04 -0.3396E-01 0.3396E-01 -0.8988E+02
294 8945.3086 0.7160E-04 -0.3354E-01 0.3354E-01 -0.8988E+02
295 9839.8398 0.5917E-04 -0.3049E-01 0.3049E-01 -0.8989E+02
296 10331.8320 0.5367E-04 -0.2904E-01 0.2904E-01 -0.8989E+02
ENTER NEW MAX OMEGA,INCREMENT FACTOR:
No higher frequencies
Figure 26-15. Frequency Response Calculation
26-29
Plotting Results Program Application Guide: Volume II
ACTIVITY? BODE

0 = NONE 1 = VERSATEC
2 = HP 7221A 3 = TEKTRONIX 4010
4 = TEKTRONIX 4014 5 = TEKTRONIX 4014 W/EGM
10 = CALCOMP 12 = KMW V.P.
25 = Invalid X Window DISPLAY 27 = HP 7470A
28 = HP 7475A 31 = QMS LASERGRAFIX
33 = TEKTRONIX 4111 36 = REGIS
37 = TEKTRONIX 41XX FILE 38 = HP-GL FILE
39 = Invalid X Window DISPLAY 40 = TEKTRONIX 4010/4014 FILE
41 = POSTSCRIPT 99 = INDE. PLOT FILE
ENTER DESIRED PLOTTING DEVICE:
31
EXAMFROUT
BODE PLOT
Field voltage versus reference voltage at bus 2
ENTER MINIMUM, MAXIMUM FREQUENCY:
.001 10000.
ENTER LINE TYPE

0 = SYMBOLS, 1 = PATTERNED LINES, 2 = BOTH:
2
ENTER OUTPUT, INPUT NUMBERS (-1 TO EXIT):
5 3
EFD1GEN 1 /VRF1GEN 1
ENTER 1 FOR LOG MAGNITUDE
2 FOR NATURAL MAGNITUDE
3 FOR ANGLE :
1
ENTER MINIMUM, MAXIMUM PLOTTING RANGE:
.01 1000.
5 3
EFD1GEN 1 /VRF1GEN 1
ENTER 1 FOR LOG MAGNITUDE
2 FOR NATURAL MAGNITUDE
3 FOR ANGLE :
3
ENTER MINIMUM, MAXIMUM PLOTTING RANGE:
-180, 180
-1
5/3
ENTER:
0 = RETURN
2 = NEW PLOT ON THE QMS LASERGRAFIX
3 = REPEAT PLOT ON NEW DEVICE:
0
ENTER NUMBER OF COPIES (0 TO 5), DEVICE NAME FOR QMS LASERGRAFIX:

1
Job 00290 entered on queue QMS_1200.
Figure 26-16. Dialog to Create Bode Plot
26-30
Program Application Guide: Volume II Plotting Results
Figure 26-17. Bode Plot of Output 5 Varying with Input 3
26-31
Small Perturbation Simulations Program Application Guide: Volume II
Figure 26-18 and 26-19 present the frequency response results in polar plots. Note that
when entering the maximum and minimum plotting range and using a logarithmic scale,
the user should enter positive values that are integral powers of 10. For example, .001 or
10-3 is valid while .002 or 10-2.6989 is not valid.
26.11 Small Perturbation Simulations

Execution of activities LSTR and LRUN allow the user to run a simulation using the
matrices. If the matrices were initially obtained from a standard PSS®E setup, use of these
activities would give the user a quicker simulation as compared to simulating the entire
system with network to check if the system is dynamically stable. These activities can also
be used to confirm the results obtained from calculating the eigenvalues and frequency
response of the system.
The initialization activity LSTR requires the matrices first to be read into memory via
MCAS. The activity, when invoked, first initializes all state variables to zero.
Activity LSTR then instructs the user:
DELT=0.01 SECONDS, ENTER NEW DELT:
A response of carriage return will leave the value unchanged. The activity then sets the
value of simulation TIME to minus two time steps. This provides for a period of simula-
tion under steady-state conditions prior to initiating a disturbance. Activity LSTR then
instructs the user to:
ENTER INPUT #, PERTURBATION (P.U.):
The user should then enter the desired input to perturb (default is zero) and a perturbation
amount (no default value is placed). The program will repeat the question until a carriage
return, zero, or negative value has been entered or all inputs have been perturbed. These
disturbances are automatically applied at time equals zero during activity LRUN.
Activity LSTR then instructs the user to:
The user responds with the name of the Simulation Channel Output File to be used in the
subsequent dynamic simulation run. If no filename is specified in response to the above
instruction, the writing of the output channel variable values to a file is suppressed during
the simulation run.
The time simulation activity LRUN, sequences through time solving the system's differen-
tial equations at each time step. Upon selecting activity LRUN, the user is instructed:
AT TIME=0.000
ENTER TPAUSE, NPRT, NPLT
26-32
Program Application Guide: Volume II Small Perturbation Simulations
Figure 26-18. Natural Scale Polar Plot of Output 5 Varying with Input 3
26-33
Small Perturbation Simulations Program Application Guide: Volume II
13:57
POWER
TECHNOLOGIES,
INC.R FILE: gp
FRI JAN 12, 1996

CHNL# 2: [ETRM 1 [ ] MC 1]
1.1000 0.90000
CHNL# 1: [EFD 1 [ ] MC 1]
2.8000 0.80000
10.000
9.0000
8.0000
7.0000
6.0000
TIME(SECONDS)
5.0000
4.0000
3.0000
2.0000
1.0000
0.0
Figure 26-19. Logarithmic Scale Polar Plot of Output 5 Varying with Input 3
26-34
Program Application Guide: Volume II Small Perturbation Simulations
The user responds with three data items having the following significance.
TPAUSE Activity LRUN steps through time simulating the system, starting at
the present value of simulation TIME, until TIME > TPAUSE.
NPRT The values of the output channel variables are written to the output
device every NPRT time steps. They are printed for the present value
of TIME, every NPRT time steps thereafter, and for TIME = TPAUSE.
If NPRT is entered as zero, the tabulation of output channel values is
bypassed.
NPLT The values of the output channel variables are written to the Simulation
Channel Output File every NPLT time steps. They are written to the
file for the present value of TIME, every NPLT time steps thereafter,
and for TIME = TPAUSE. If NPLT is entered as zero (or one), channel
values are written to the output file at the completion of each time step.
NPLT has significance only if a channel output file has been specified
either in activity LSTR.
Following the specification of the above parameters, if the value specified for TPAUSE is
less than the current value of simulation TIME, activity LRUN is terminated. Otherwise,
if a Simulation Channel Output File has been opened, the message is printed at the dialog
output device:
CHANNEL OUTPUT FILE IS aaaaaa
The simulation, starting at the present value of simulation TIME, then commences.
Activity LRUN responds to the following interrupt control codes:
AB Force a pause by setting TPAUSE to the current value of simulation

TIME. This overrides the value of TPAUSE, which was specified at the
time activity LRUN was invoked. In addition, if LSYSAN is operating
in its response file mode (i.e., activity IDEV,file had previously been
selected), the response file is closed and subsequent user commands
are taken from the user's terminal.
CH Print the output channel values at each time step. The value of NPRT
that was entered at the time activity LRUN was invoked is overridden
while this option is in effect. When this option is cleared, the original
value of NPRT is restored and the next printing of channels occurs
NPRT time steps from the TIME of the clearing of the option.
TI Print the value of simulation TIME at the user's terminal.
Activity LSTR sets the output control variables NPRT and NPLT to one. On the first sub-
sequent execution of activity LRUN, these values are the defaults for these variables. On
following executions of activity LRUN, these variables default to the values specified on
the previous execution of activity RUN.
Whenever the value of simulation TIME reaches the time specified for TPAUSE, the time
derivative calculations are applied with a flag set to inform the simulation that this is a t-
26-35
Stopping the Program Program Application Guide: Volume II
calculation. For any of the channel output options that are active, the appropriate output is
generated regardless of whether the present time step is a print or plot time step as deter-
mined by the variables NPRT and NPLT. When activity LRUN is re-entered, the time step
is repeated as a t+ calculation. Appropriate channel output is run at the completion of the
time step as determined by the new specification of the output control variables. This
applies both for the case in which activity LRUN has advanced TIME to the value speci-
fied for TPAUSE and when the user forces a pause with the AB interrupt control code.
When using the AB interrupt control code to force a pause in the simulation, care must be
taken when LSYSAN is being executed from a command file with LSYSAN commands
and responses taken from a response file. As described above, activity LRUN closes the
PSS®E response file.
The PSS®E auxiliary plotting program PSSPLT can be used to plot the results placed in
the output file.
26.12 Stopping the Program

The LSYSAN termination activity STOP, is the normal exit from the program. When
selected, all files opened are closed and the user is returned to the operating system level.
26.13 Timing Statistics

The timing statistics activity TIME allows the user to obtain execution time statistics dur-
ing a work session. When invoked with the suffix INIT, or on the first selection of activity
TIME following entry into LSYSAN, the timers are initialized and the following message
is printed:
TIMER INITIALIZED
On subsequent executions of activity TIME when no suffix is specified, activity TIME

prints a summary of elapsed, CPU, and disc channel times, in seconds, since the previous
execution of activity TIME, and cumulative times from the point at which the timers were
last initialized. This tabulation is of the form:
ELAPSE CPU DISK
SINCE LAST "TIME" XX.XXX X.XXX X.XXX
CUMULATIVE XX.XXX X.XXX X.XXX
Activity TIME is not sensitive to any interrupt control code options.
26.14 The Null Activity

The null activity TEXT doesn't really do anything. However, it does provide a mechanism
by which the user may insert descriptive comments in a Response File at any point at
which the next activity to be run is specified.
In response to the question ACTIVITY?, the user may enter the activity name TEXT fol-
lowed by any descriptive text that is suitable; for example:
TEXT *** HAVE YOUR MESSAGE PRINTED HERE!
26-36
Program Application Guide: Volume II Report Output Devices
Activity TEXT is not sensitive to any interrupt control code options.
26.15 Report Output Devices

The report output device selection activity OPEN allows the user to preselect the destina-
tion for output reports generated by the LSYSAN output reporting activities.
The report output device closing activity CLOS terminates output to the previously
OPENed selection and returns to the operating mode in which each reporting activity
requests the user to select the destination for its report.
When initiated, activity ECHO first CLOSes the previously OPENed output device if any,
and then instructs the user to:
The user then selects the device to be used for output reports. If a file is selected,
activity OPEN asks the user to:
If the Versatec or line printer is selected, the user is able to have up to six copies printed:
ENTER NUMBER OF COPIES:
If an alternate spool device is selected, the user is instructed to:

ENTER SPOOL DEVICE NAME:
and
ENTER NUMBER OF COPIES:
After the report output device is selected and activity OPEN is terminated, the above
ENTER OUTPUT DEVICE instruction from subsequently selected PSS®E reporting activities
is suppressed, and the output is automatically sent to the device selected in activity OPEN.
When the selection is a file or high speed printing device, output reports are stacked in the
order in which they are generated.
When a hard copy device is selected in activity OPEN, the output is not printed until
activity OPEN, CLOS or STOP is selected.
The device specified in activity OPEN applies to all reporting activities.
Activity OPEN is not sensitive to any interrupt control code options.
Activity CLOS is not sensitive to any interrupt control code options.
26.16 Activity ECHO

The dialog echoing activity ECHO enables the writing of all subsequent user dialog input
to a designated echo file.
When activity ECHO is initiated, if an echo file had previously been opened with an ear-
lier execution of activity ECHO, it is closed and the message:
26-37
Activity HELP Program Application Guide: Volume II
CLOSING ECHO FILE old-echo-filename
is printed. If activity ECHO was selected with no suffix specified, it is terminated and the
echoing of subsequent user responses is disabled.
When invoked with the activity command ECHO,filename, the designated file is opened
as an echo file and activity ECHO prints the message:
OPENED ECHO FILE filename
All subsequent dialog input is then written to the file <filename>. If some file system
related error occurs in opening the echo file, an appropriate error message is printed and
activity ECHO is terminated with the echoing of subsequent user responses disabled.
Activity ECHO is not sensitive to any interrupt control code options.
The file built by activity ECHO is in the form of a LSYSAN Response File, which may be
specified to activity IDEV to exactly reproduce a sequence of PSS®E activity commands
and responses to instructions. Files written by activity ECHO are source files that may be
modified with the text editor to tailor the original set of user responses to the application at
hand.
26.17 Activity HELP

The help activity HELP prints a summary of the activity names and a brief description of
their function.
26.18 Directory Listing

The directory listing activity CATA tabulates an alphabetical listing of the names of files
contained in the current directory.
If no suffix is specified in invoking activity CATA, a listing of the names of all files in the
directory is printed.
Otherwise, the suffix is treated as a partial filename specification, and the names of all
files in the directory that appropriately contain this character string as a part of their name
are listed. The activity suffix may contain imbedded asterisks () that are treated as wild
characters matching zero, one or more characters. For example, selecting activity CATA
with the activity command CATA,AB will produce a tabulation of all filenames beginning
with AB; the activity command CATA,AB results in a tabulation of all files that contain
the consecutive characters AB anywhere in the filename.
The actual implementation of activity CATA differs among the various host computers of
PSS®E. There may be certain host computers on which it is not implemented.
Activity CATA is not sensitive to any interrupt control code options.
26.19 Dialog Input Selection

The dialog input device selection activity IDEV is used to change the source from which
LSYSAN accepts the user's portion of its conversational dialog to either the user's termi-
nal or a Response File in the user's directory.
26-38
Program Application Guide: Volume II Changing Elements of the Matrices
If run with no suffix specified, activity IDEV sets the dialog input device to be the user's
terminal.
When invoked with the activity command IDEV,filename, subsequent dialog input will be
taken from the specified file. If the file <filename> does not exist or some other file sys-
tem related error occurs, an appropriate message is printed and terminal input is assumed.
Otherwise, the message:
DIALOG INPUT FROM FILE filename
is printed, activity IDEV is terminated, and LSYSAN looks to the specified file for the
next activity command.
The Response File must contain an exact image of the user responses that would normally
be typed by the user in terminal input mode. Therefore, the manual creation of a Response
File, which is done with the text editor, requires an intimate familiarity with the LSYSAN
dialog.
The recommended method of setting up a Response File, therefore, is to first run the
sequence of activities to be implemented in terminal input mode with an ECHO file
enabled (see Section 26.16).
The final activity command in a Response File should be either STOP (if the Response
File completely defines the sequence of activities to be run) or IDEV (to return control
back to the user's terminal). It is also possible for the last command in the Response File to
be the activity command IDEV,file2; that is, Response Files may be chained. However, it
is not possible to nest Response Files; if the command IDEV,file2 occurs in the middle of
the Response File <file1>, it is not possible to get back to <file1> at its next command
upon the completion of the sequence of activities contained in <file2>.
26.20 Changing Elements of the Matrices

Any element of any of the four matrices A, B, H or F can be changed with
activity MCHG. The user may wish to change elements in order to investigate the effects
of opening a feedback loop or changing a time constant. After any element is changed, the
program assumes that it is just as though the matrices have just been read.
26-39
26-40
Chapter 27
Extended Simulations
Chapter 27 - Extended Simulations
27.1 Introduction
PSS®E was initially designed to model transients over a period of few to several seconds
following disturbances. During this time frame, the important effects are inertial motions
of turbine generators as affected by the characteristics of generators, excitation systems,
loads, static var sources, dc converters and, to a lesser extent, turbine-governors. These
phenomena have been broadly labeled Power System Stability by power system engi-
neers. The bandwidth of the effects being modeled is limited to about 10 Hz at the high
end with typical integration time steps of 0.00833 (1/2 cycle) for 60-Hz systems and 0.01
(1/2 cycle) for 50-Hz systems.
Higher frequency effects require modeling of the electrical network with differential equa-
tions using much smaller time steps, the domain of electromagnetic transients programs.
At the low-frequency end, the dynamic system models being represented are valid down
to 0 Hz or steady state. Thus within the bounds of modeling assumptions (loads modeled
without allowance for longer term constant power control effects, no tap changer action in
substation transformers, no switching of reactive sources, and no AGC or boiler effects)
the dynamic simulation is valid within a time scale resolution of about 20 ms to infinity.
As the time spectrum of power system dynamic effects is extended beyond several sec-
onds following a set of disturbances, additional effects come into play, such as the ten-
dency of loads to exhibit constant power characteristics through tap changer and/or load
control devices, automatic switching of reactors or shunt capacitors, prime mover power
changes through primary speed control and/or AGC (automatic generation control). Simu-
lation of such longer term effects requires additional modeling of load restoration mecha-
nisms and prime mover characteristics including boiler effects, exhaust temperature
control effects on gas turbines, etc.
For many cases where the system survives the initial disturbance, the higher frequency
effects (rotor angle swings) subside after a few seconds and then the transition to a new
post contingency state occurs over minutes. While this process can be solved by extending
the stability run through minutes with 1/2 cycle time step explicit integration by adding
relevant load restoration and prime mover models, the computation time can be excessive.
In the past, skilled investigators have dealt with the longer term effects through intelligent
27-1
Extended Simulations PSS®E 33.4
Explicit and Implicit Integration Algorithms Program Application Guide: Volume II
use of power flow solutions (pre- and postcontingency) along with 1/2 cycle time step
dynamic simulation.
Enhancements to the numerical techniques of solution of the dynamic and algebraic equa-
tions of the power system process have made it possible to switch back and forth between
a small time-step solution yielding good resolution of high-frequency effects and larger
time-step solutions that inherently filter out high-frequency effects while preserving fidel-
ity of longer term effects.
The extended term dynamic simulation option of PSS®E was created to analyze system
behavior over the period of many seconds to minutes following disturbances where fre-
quency durations and prime mover response can be important as can the process of resto-
ration of loads, action of excitation limiters, tap changing transformers, switching of
capacitors and reactors, etc. It has required the development of new models such as tur-
bine-governors with boiler effects, load restoration mechanisms, tap changing, etc.
In addition, to make it practical to run simulations using small time steps during periods
where high frequency effects are important as well as larger time steps, to cut down run
time, during the extended term period dominated by low-frequency phenomena, all exist-
ing models in the PSS®E model library were modified to accommodate the change from
explicit to implicit integration.
This section describes the features of the extended term version of PSS®E that enables the
user to efficiently investigate power system dynamic effects beyond the time scale of a
few seconds used in typical stability runs. The program includes the option of locking syn-
chronous sources to a common global system frequency.
27.2 Explicit and Implicit Integration Algorithms
27.2.1 Explicit Integration Algorithm

The standard PSS®E simulation activities STRT/RUN use an explicit integration technique:
the Modified Euler integration algorithm. Integration algorithms can be conveniently
expressed in terms of block diagrams using the time delay operator Z-1 (Z-form)1. The
modified Euler algorithm is shown in Figure 27-1 where T is the time step.
Input TZ – 1  3 – Z – 1  Output
------------------------------------
Train of Pulses 2  1 – Z –1  Train of Pulses
Figure 27-1. Modified Euler Integrator Z-Form Function
1 F.P. de Mello, "Z Transform Techniques for the Analysis of Systems," AIEE Conference Paper
CP-60-1291, 1960, Fall General Meeting, Chicago, Illinois.
27-2
PSS®E 33.4 Extended Simulations
Program Application Guide: Volume II Explicit and Implicit Integration Algorithms
The response of discrete integrator output pulses to a discrete input pulse representative of
the derivative will be a train of pulses delayed by one time step as shown in Figure 27-2.
1
T
Unit Pulse or Discrete
0 Approximation of
Unit Impulse at t = 0
1.5
1.0 Response of Modified
Euler Integration
0 Algorithm to Unit Pulse
–T 0 T 2T 3T 4T 5T
Figure 27-2. Time-Domain Relationships
Illustrating the Response of the Modified Euler Integration Algorithm to a Unit-Pulse
The advantage of this simple algorithm is that outputs of integrators (states) are not instan-
taneous functions of their inputs (derivatives of states). Hence, the computation of the
state space equations
X· = AX + BU (1.1)
is a straightforward noniterative procedure with the vector of states, X, remaining fixed

during the computation of the vector of derivatives of states, X· . The disadvantage of this
explicit integration algorithm is that numerical stability dictates use of time steps smaller
than the smallest time constant or reciprocal of the largest root of the linearized system of
equations describing the process. This can be readily shown with the example of a first-
order continuous system descriptive of a time constant, .
1
X· = ---  Y – X  (1.2)

The solution of this equation, using discrete time steps and the modified Euler integration
algorithm, will exhibit numerical instability if time step T > . Figure 27-3 shows the effect
of time-step size in the simulation of a typical high, initial response excitation system con-
trolling generator terminal voltage on open circuit using the explicit or modified Euler
integration algorithm. In this case, the onset of numerical instability as the time step is
increased to 0.03 seconds is noticeable.
27-3
PSS/E PROGRAM APPLICATION GUIDE EXAMPLE
15:56
BASE CASE INCLUDING SEQUENCE DATA
POWER
TECHNOLOGIES,
INC.R
THU MAR 07, 1996

CHNL# 5: [EFD 201 [HYDRO 500.00] MC 1]
5.0000 FILE: /tfl/example/out3 0.0
CHNL# 5: [EFD 201 [HYDRO 500.00] MC 1]
5.0000 FILE: /tfl/example/out1 0.0
2.5000
2.2500
2.0000
Time Step = 0.3 sec
1.7500
1.5000
TIME(SECONDS)
1.2500
1.0000
Time Step = 0.008333 sec
0.75000
Generator Field Voltage
0.50000
0.25000
0.0
Figure 27-3. Simulation of an Excitation System Controlling Generator Terminal Voltage

on Open Circuit Using the Modified Euler Integration Algorithm,
Showing Numerical Instability for a Time Step of 0.03 sec
27-4
27.2.2 Implicit or Trapezoidal Algorithm

The Z-form expression for the trapezoidal integration algorithm is shown in Figure 27-4,
and the response of the integrator in terms of a train of pulses to a single input pulse is
This algorithm, which is inherently more accurate, has an instantaneous response of 1/2 to
an input pulse of unit area (pulse height 1/T). The solution of (1.2) would have to be
expressed as shown in Figure 27-6, which shows a closed-form solution of the output train
of pulses as a function of the input train. The block Z-1 is a one time-step delay operator.
For a linear system, the user can derive a closed-form solution of output as function of
input and previous history (output of Z-1 block).
In the general case of nonlinear differential equations, however, the instantaneous propa-
gation of input to output and some nonlinear feedback of output to input would require
iterations, i.e., implicit integration. The trapezoidal integration is advantageous because it
avoids numerical instability. The effect of larger and larger time steps is loss of fidelity of
high-frequency transients; the system essentially yields its steady-state response as the
time step is increased to infinity.
Figure 27-7 shows the response of the excitation control system previously described but
now using different time step sizes with the trapezoidal integration scheme.
Unlike the case in Figure 27-3 where numerical instability arises with T of 0.03 sec or
larger, stable response is obtained with T as large as 0.2 sec. The response with the larger
time step demonstrates the same steady-state and low-frequency characteristics.
The explicit integration algorithm allows the computation to follow, in a straight forward
manner, the state space formulation of the dynamics equation
X· = AX + BU
whereas the use of implicit or trapezoidal integration requires a more complex closed
form, or iterative solution of the resulting algebraic equations at each time step. In the fol-
lowing sections the use of the explicit integration method will be referred to as the state
space formulation, while the use of the implicit or trapezoidal algorithm will be referred to
as the Z-form formulation.
Input T  1 + Z –1  Output
--- -----------------------
Train of Pulses 2  1 – Z –1  Train of Pulses
Figure 27-4. Trapezoidal Integrator Z-Form Function
27-5
1
T
Unit Pulse or Discrete
0 Approximation of Unit
Impulse at t = 0
1.0 Response of
0.5 Trapezoidal Integration
0 Algorithm to Unit Pulse
-T 0 T 2T 3T 4T 5T
Figure 27-5. Time-Domain Relationships

Illustrating the Response of the Trapezoidal Integration Algorithm to a Unit Pulse
1
+ ---------------
Input, I  2 Output, O
1 + -----
+ T
Z-1
+ – 2
 1 – -----
T
Figure 27-6. Z-Form Block Diagram Giving Discrete Solution

of O(s)/I(s) = 1/1 + s
27-6
Generator Field Voltage
Time Step = 0.03 sec (- - -)

Figure 27-7. Simulation of an Excitation System Controlling Generator Terminal Voltage

on Open Circuit Using the Trapezoidal Integration Algorithm
with Different Time Step Sizes
27-7
Modes of Extended Term Simulation Program Application Guide: Volume II
27.3 Modes of Extended Term Simulation

Extended term simulation has three modes of operation, selected automatically and based
on the size of the time step and user-defined thresholds.
• For small time steps, such as the typical one-half cycle time step, the simulation
response will be essentially identical to that of simulations using the state space
method.
• For larger time steps, typically 0.05 to 0.1 sec (3 to 6 cycles for a 60-Hz system), all
states of the model and, hence, all swing modes are preserved. Damping of the lower
frequency swing modes is preserved while higher frequency swing modes tend to be
filtered out. Numerical stability is maintained.
• For even larger time steps, 0.15 to 0.2 sec, the program switches to a mode where the
rotor angle of each generator is adjusted at each time step to yield electrical torque
equal to mechanical torque less the unit’s share of the connected system’s global
accelerating torque. This may be used when all units in an island follow essentially the
same frequency.
At the user’s discretion, the time step can be changed at any point in the simulation to
adjust the solution mode, i.e., to disclose high-frequency characteristics for a few seconds
following a disturbance and then revert back to a large time-step mode. If modeling results
in an island with generators all essentially infinite (very large inertias, H), then the con-
stant island frequency mode should not be used. This mode assumes all units are at the
same speed. With infinite inertias, all machines, by default, are at the same frequency
(speed). To increase the time step, the user should also increase the constant island fre-
quency threshold, DLTEXT (see Section Section 27.6).
Most of the models distributed with PSS®E can be run in all modes of the extended term.
Those models not available are listed in Section 14.22.1 of the PSS®E Program Operation
Manual. Models CIMTR3 and CIMTR4 have a restriction: they cannot be the only
machine in an island when using the constant island frequency mode.
User-written models are allowed in the extended term simulation. These models that have
STATE variables and calculate derivatives must follow additional procedures established
for extended simulation models (documented later in this section). All user-written models
must follow the additional procedures to be used in extended term simulations.
27.4 Additional Models for Extended Term Simulations
27.4.1 Online Tap Changer, OLTC1

The online tap changing model, OLTC1, allows the modeling of transformer tap adjust-
ments to help control system voltage: it has two main components. The first is the voltage
sensor, which compares the input voltage to the preselected setting (voltage level) and a
tolerance, or spread, in voltage level (bandwidth). If the voltage input to the sensor is out
of the control band, the control will operate after the time delay has been exceeded. Thus,
the output of the regulator will be either raised or lowered until the voltage feedback into
the sensor is again within the control band. Because the time delay is generally magni-
tudes greater than the voltage sensor transducer time constant, the transducer time con-
27-8
Program Application Guide: Volume II Additional Models for Extended Term Simulations
stant is not modeled. OLTC1 operates only on transformers controlling voltage in the
power flow. If the power flow was solved without adjusting taps and the voltages are out-
side the control band, OLTC1 will print a message during activity STRT stating that the
timer is started. A positive tap step is also required in the power flow.
The second major component of OLTC1 is the time-delay circuit. It permits the regulator
to ignore brief, self-correcting voltage variations. The time delay enables the transformer
to correct only those voltage variations that exist for longer than a preset time. The time
delay for subsequent tap changes can be specified independently of the first time delay.
OLTC1 incorporates the most common type of timer, an integrator that adds the total
amount of time that the voltage is outside the preselected control band and subtracts the
time it is inside the control band. When the voltage is outside more than inside by an
amount greater than the time-delay setting, TD, the time-delay sends a tap signal to the tap
changer motor. The tap changer motor operates TC seconds later. Using a timer setting of
TD = 30 sec, Figure 27-8 shows a voltage that is out-of-range for 13 sec and returns within
the bandwidth for 5 sec. Therefore, a tap signal is not sent until another 22 sec (30 - (13 -
5)) elapses with the voltage below the minimum value. OLTC1 also allows for a subse-
quent delay for sending additional signals to the tap changer motor. The user can enter a
delay, TSD, to get additional delay. Delay TSD must be greater than TC to be effective. A
common practice in distribution transformer control is line-drop compensation. OLTC1
recognizes when a remote bus voltage is being controlled, as defined by the power flow
data. The user represents the feeder circuit, possibly including a dummy bus, to represent
line-drop compensation. Typical data for this model would be 25.0 sec for TD, 10 sec for
TC, and 15 sec for TSD.
VMAX
5
VMIN
kV or MW
First Tap
Signal Sent
13 22
10 20 30 40 50
Time (sec)
Figure 27-8. Example of Integration Timer for the OLTC1 and OLPS1 Models
27-9
Additional Models for Extended Term Simulations Program Application Guide: Volume II
27-10
27.4.2 Online Phase Shifter, OLPS1

The OLPS1 model represents the automatic movement of taps on phase-shifting transform-
ers to control power flow. OLPS1 is composed of the same two components as OLTC1:
the delay circuit and the sensor/integrator, but the controlling input in the latter component
is megawatts. It also uses the parameters specified in the power flow. Most phase shifters
in operation have variable angle steps. For simplicity of data input and simulation pur-
poses, a uniform step is assumed in the dynamics simulation. The user should enter a num-
ber of positions from which a uniform angle step will be calculated into the power flow
data even though the power flow calculation will use a continuous adjustment. The step
size will, however, be saved and reported with the case.
27.4.3 The dc Line Tap Changer, DCTC1

The DCTC1 model represents the automatic movement of taps associated with two termi-
nal dc lines, but does not represent the tap control logic of any specific dc line. The model
is intended to bring the same tap movement philosophy as used in the power flow; for
example:
• Adjust rectifier taps to hold ALFA between the minimum and maximum values entered
in the power flow, and
• Adjust inverter taps to hold GAMMA between the minimum and maximum values
entered in the power flow, or
• Adjust inverter taps to hold the dc voltage near the scheduled voltage
The model recognizes remote commutation buses as specified in the power flow and will
adjust the remote transformers as specified in the dc line data. The sensor/integrator of this
model is identical to that used by OLTC1 except that converter angles, ALFA and
GAMMA, are inputs. When in constant GAMMA control, the model checks to see if
movement of the tap will bring the dc voltage closer to the desired dc voltage (VSCHED).
If tap movement does correct voltage, the integrator will start integrating.
Unlike the power flow tapping logic, this model will only move one tap step at a time,
similar to OLTC1. The model is ignored if MDC is set to zero; i.e., the line is manually
blocked. During automatic blocking and bypassing performed by the dc lines, such as
CDC6, the model is active. This model cannot be used with CDC1.
27.4.4 Maximum Excitation Limiters, MAXEX1 and MAXEX2

Maximum excitation limiters are designed to protect the generator field of an ac machine
with automatic excitation control from overheating due to prolonged overexcitation,
which can be caused either by failure of a component of the voltage regulator or an abnor-
mal system condition. Model MAXEX1 acts through the regulator to correct an overexcita-
tion problem. This model assumes an inverse time characteristic as shown in Figure 27-9;
i.e., operating time is a function of the magnitude of the overvoltage.
MAXEX2 allows limiting of either field voltage or field current; for the sake of simplicity,
this text uses field voltage but field current would be limited in an equivalent manner.
After the unit has timed-out, the signal sent to the regulator is a function of field voltage
deviation from a desired value, as shown in Figure 27-10 for MAXEX1 and MAXEX2.
27-11
Additional Models for Extended Term Simulations Program Application Guide: Volume II
(EFD1, TIME1)
Time (sec)
(EFD2, TIME2)
(EFD3, TIME3)
EFD (Per Unit of Rated)
Figure 27-9. Inverse Time Characteristics of MAXEX1 and MAXEX2
EFDDES * EFDRATED
+
0.
– KMX VOEL
EFD 
VLOW To Regulator
EFDDES or IFDDES
0.
+
EFD – KMX VOEL


or IFD S To Regulator
VLOW
Figure 27-10. Block Diagrams of MAXEX1 (Top) and MAXEX2 (Bottom)
After timeout, the model does not reset (i.e., it is assumed that the operator must reset the
device). The following assumptions are also made by the model.
1. Below EFD1, the device is inactive.
2. Above EFD3, the time to operate is constant and equal to TIME3.
3. If EFD goes below EFD1 at any time before the device has timed-out, the timer resets.
27-12
Most General Electric and Westinghouse excitation systems have either an optional
device or an integral part of the excitation limiter that trips the exciter and/or the generator.
These devices are provided for regulator failure, and are not modeled by MAXEX1 or
MAXEX2. The user must manually trip the exciter or generator to model a regulator fail-
ure. Typical values for the inverse characteristic would be the following:
EFDx
X TIMEx (sec)
(pu of rated)
1 120 1.1
2 40 1.2
3 15 1.5
The characteristics of the MAXEX1 model require a low gain for the KMX constant
because the model output feeds directly into the voltage regulator input of the exciters. For
high-gain exciters a value of 0.01 is suggested. If the gain is too high, the network may not
converge. To overcome the droop of the exciter and interaction with terminal voltage, a
value of 0.9 is suggested for EFDDES.
The MAXEX2 model is a more accurate representation of the physics of actual maximum
excitation limiters than the MAXEX1 model. MAXEX2 can also be used to approximate
the actions of operators attempting to relieve field overload by reducing voltage setpoints.
The MAXEX2 model uses an integral controller and hence controls the field voltage/cur-
rent to the desired value with the drop inherent in MAXEX1. Typical values would be
0.002 for KMX and 1.0 for EFDDES.
27.4.5 EXTL Type Models

The most accurate representation of loads would require the full modeling of distribution
circuits representing all tap changing transformers, voltage regulators, static and switched
capacitors. The model would also contain detailed characteristics of the load, such as the
percentage of the load represented by specific load types, such as motors or lighting, and
the characteristics of these load types with respect to time and voltage. Such full modeling
is not generally practical when the user is already starting from a very large base-case
power flow. The additional data required would not only double the size of the model and
thus the computational requirement, but would also be burdensome on the user trying to
collect the data. For these reasons, PTI has developed the family of EXTL type models
(EXTLBL, EXTLOW, EXTLZN, EXTLAR, EXTLAL). This model is designed to simulate the
general effects of loads being reset to a constant MW and constant Mvar in the steady state
without specifically modeling the equipment (taps, capacitors, etc.) responsible for this
resetting. This model works on the constant current and constant impedance loads created
by activity CONL, returning them at a user defined-rate to constant MW and Mvar.
27-13
TGOV5 Program Application Guide: Volume II
27.5 TGOV5
27.5.1 Introduction
TGOV5 is a model of a steam turbine and boiler that represents governor action, main,
reheat and low-pressure effects, including boiler effects. The boiler controls will handle
practically any mode of control including coordinated, base, variable pressure, and con-
ventional. The control mode is selected by the proper choice of constants.
27.5.2 Governor Model

The governor model is similar to the IEEEG1 model, and can be used for tandem and
cross-compound units. The valve has rate limits, as well as minimum and maximum lim-
its. Steam flow is proportional to the product of throttle pressure and valve area rather than
just proportional to valve position as in the standard governor models. Proper selection of
the time constants and gains allows the modeling of the reheater and intermediate and
low-pressure turbine effects. The variables of the governor that interface with the other
portions of the model are Po (load reference of the governor), PT (throttle pressure) and ms
(steam flow).
27.5.3 Fuel and Boiler Models

Drum pressure, PD, is proportional to the integral of steam generation less steam flow out
of the boiler. Throttle pressure, PT, is equal to drum pressure less a pressure drop across
superheaters and steam leads. This pressure drop varies as square of steam flow and also
with density of steam. The pressure drop coefficient is shown to be a function of boiler
pressure.
Steam generation is controlled by the inputs to the boiler (fuel and air) that are driven by a
three-mode controller. Provision is made to introduce anticipation from steam flow or
MW demand signal. The three-mode controller has nonwindup limits: KI can be set to
zero to model the fuel and air control on manual. To model two mode control, set TR and
TR1 to zero. The time delay is modeled by a fourth-order approximation, which allows the
modeling of long time delays without an inordinately long delay table.
27.5.4 Pressure Set Point and Load Reference Control

A MW demand is modified by the frequency bias and pressure in the coordinated control
front end. This bias demand signal is matched by the load reference motor position with
appropriate rate limits. The possibility exists that the load demand signal will act on the
pressure set point. The gain of pressure error modifying the MW demand signal can be set
as a constant or proportional to the product of pressure set point Psp and load reference Po.
The MW demand signal is stored in VAR(L+2). External logic to change the MW
demand, due to AGC pulsing, for example, could be added to the CONEC subroutine.
The initialization of the model is dependent upon the gains, K13 and K14, and may require
that the CON C3 be changed from its original value. VAR(L+6) is used to store the
27-14
Program Application Guide: Volume II TGOV5
changed value of C3. If K13 is zero, then C3 is set equal to Psp and the MW demand signal
is set equal to any of the following:
• Electrical power, per unit on machine base (KMW = 1.0).
• Po (KL = 1.0).
• 0 KMW and KL= 0.0).
Setting K14 to zero locks the load reference at its original value. If K13 and K14 are both
zero, the MW demand signal is used only if K11 is nonzero. If K13 and K14 are both non-
zero, the MW demand is set as above, and C3 is calculated to get the proper relationship
between the pressure set point and MW demand. On certain control strategies, C2 may
also need to be calculated. In mode 1, VAR(L+5) is set either to the value of CON(J+31)
or a calculated value depending on the sign of the CON’s value. If C2 is negative,
VAR(L+5) is set to electrical power, per unit on machine base to allow the pressure error
bias signal to be proportional to machine initial loading.
27.5.5 Model Constants

The following is an explanation of the model constants:
K The inverse of the governor speed droop.
T1 and T2 The governor controller lag and lead time constants (sec).
T3 The valve servomotor time constant for the control valves
(sec).
Uo The control valve open rate limit (per unit/sec).
Uc The control valve close rate limit (per unit/sec).
VMAX The maximum valve area (per unit).
VMIN The minimum valve area (per unit).
T4 The steam flow time constant (sec).
T5, T6, and T7 The first and second reheater time constants, and the cross-
over time constant. They may be set to zero if all steps are
not necessary: i.e., no second reheat stage.
K1, K3, K5, and K7 The fractions of the HP unit’s mechanical power developed
by the various turbine stages. The sum of these constants
should be one for a non cross-compound unit.
K2, K4, K6, and K8 Similar fractions of the LP unit’s mechanical power. These
fractions should be zero for a noncross-compound unit. For
a cross-compound unit, the sum of K1 through K8 should
equal one.
K9 The adjustment to the pressure drop coefficient as a func-
tion of drum pressure.
K10 The gain of anticipation signal from main stream flow.
27-15
K11 The gain of anticipation signal from load demand.

K12 The gain for pressure error bias.
K13 The gain between MW demand and pressure set point.
K14 Inverse of load reference servomotor time constant (= 0.0 if
load reference does not change).
RMAX The load reference positive rate of change limit (per
unit/sec).
RMIN The load reference negative rate of change limit (per
unit/sec).
LMAX The maximum load reference.
LMIN The minimum load reference.
C1 The pressure drop coefficient.
C2 The gain for the pressure error bias.
C3 The adjustment to the pressure set point.
B The frequency bias for load reference control.
CB The boiler storage time constant (sec).
KI The controller integral gain.
TI The controller proportional lead time constant (sec).
TR The controller rate lead time constant (sec).
TR1 The inherent lag associated with lead TR (usually about
TR/10) (sec).
CMAX The maximum controller output.
CMIN The minimum controller output.
TD The time delay in the fuel supply system (sec).
TF The fuel and air system time constant (sec).
TW The water wall time constant (sec).
Psp The initial throttle pressure set point.
TMW The MW transducer time constant (sec).
KL The feedback gain from the load reference (0.0 or 1.0).
KMW The gain of the MW transducer (0.0 or 1.0).
Pe The deadband in the pressure error signal for load reference
control (per unit pressure).
27.5.6 Typical Data

There are four basic boiler/turbine control strategies:
27-16
Program Application Guide: Volume II TGOV5
1. Conventional Control. Most conventional drum-type steam units are operated in the
boiler follow mode, where changes in generation are initiated by turbine control valves
and the boiler controls respond with necessary control action upon sensing the
changes in steam flow and deviations in pressure. In this mode the turbine has access
to the stored energy in the boiler and load changes within reasonable magnitudes
occur with fairly rapid response.
2. Turbine Follow. The turbine follow mode involves use of the turbine control valves to
regulate boiler pressure. This method can be done with practically no time delay so that
boiler pressure suffers virtually no transient deviations, and stored energy in the boiler
is not used. Steam flow through the turbine and, therefore, turbine power follows
closely the amount of steam generation, i.e., the input to the boiler. The response of
turbine power is considerably slower than conventional control.
3. Coordinated Optimal. The coordinated optimal mode recognizes the advantages and
disadvantages of the conventional and turbine follow modes and the need for varying
degrees of compromise between the desire for fast response to load changes and the
desire for boiler safety and good quality of control of steam conditions. A new demand
signal is modified by a frequency deviation bias matching the unit’s governor droop
characteristic to develop the desired MW. Comparison with the unit’s actual output
develops the MW error. The desired MW signal as well as the MW error signal are sent
to the boiler controls. Turbine-speed changer position is directed to reduce a combina-
tion of MW error and pressure error to zero while the boiler controls are directed to
reduce the pressure error biased by the MW error to zero. Depending on the cross-
coupling strength between pressure and MW loops, the load response can be adjusted
to any degree between that of the conventional and turbine follow modes.
4. Variable Pressure. In the variable pressure control mode, the pressure set point is pro-
portional to MW demand. The pressure error between set point and actual throttle
pressure drives steam generation. No coupling occurs between the demand signal and
the turbine control valve position.
The TGOV5 model can simulate each of the control strategies by judicious setting of the
model gains. Actual plant controls could be variants of the standard control philosophies
or combinations of them. Typical data is given for each control mode for both coal- and
oil/gas-fired units in Tables 1 and 2, respectively. The applicability of this data to any spe-
cific implementation of control strategy on a specific unit requires sound engineering
analysis and judgement. Typical data for several governing systems is given in the IEEE
Committee Report "Dynamic Models for Steam and Hydro Turbines in Power Systems",
IEEE Transactions on Power Apparatus and Systems, Vol. PAS-92, pp. 1904-1915, 1973.
27-17
Table 1: TGOV5 Typical Data for Coal-Fired Units*
Conventional
Conventional
Coordinated
Coordinated
Pressure
Pressure
Variable
Variable
Optimal
Optimal
Turbine
Turbine
Follow
Follow
CON CON
K 20 K14 5 1.0 0
T1 0 RMAX 0.0005 0.1
T2 0 RMIN -0.0005 -0.1
T3 0.15 LMAX 0.9
Uo 0.4 LMIN 0.2
Uc -0.4 C1 0.2 0.87
VMAX 0.9 C2** 0 -1 15
VMIN 0.2 C3 1 0
T4 0.4 B 0 20 20 20
K1 0.3 CB 200
K2 0 KI 0.02 0
T5 9 TI 90
K3 0.4 TR 60
K4 0 TR1 6
T6 0.5 CMAX 1.1 1.1
K5 0.3 CMIN 0.5 0 0.1
K6 0 TD 60
T7 0 TF 25
K7 0 TW 7
K8 0 Psp 1.0 0.9
K9 0 0.67 TMW 106 1.0
K10 0 KL 0.0
K11 0 1.0 KMW 1.0
K12 0 1.0 PE 0.0
K13 0 1.0
*If specific data is not shown, use conventional control data.
** When C2 is negative, the C2 VAR is set to Pelec in per unit on MBASE.
27-18
Program Application Guide: Volume II Running Extended Term Simulations
Table 2: TGOV5 Typical Data For Oil- or Gas-Fired Units*
Conventional
Conventional
Coordinated
Coordinated
Pressure
Pressure
Variable
Variable
Optimal
Optimal
Turbine
Turbine
Follow
Follow
CON CON
K 20 K14 5 1.0 0
T1 0 RMAX 0.0005 0.1
T2 0 RMIN -0.0005 -0.1
T3 0.15 LMAX 0.9
Uo 0.4 LMIN 0.2
Uc -0.4 C1 0.2 0.87
VMAX 0.9 C2** 0 -1 15
VMIN 0.2 C3 1.0 0
T4 0.4 B 0 20 20 20
K1 0.3 CB 150
K2 0 KI 0.04 0
T5 9 TI 70
K3 0.4 TR 30
K4 0 TR1 3
T6 0.5 CMAX 1.1 1.1
K5 0.3 CMIN 0.5 0 0.5
K6 0 TD 0
T7 0 TF 5
K7 0 TW 7
K8 0 Psp 1.0 0.9
K9 0 0.67 TMW 106 1.0
K10 0 KL 0.0
K11 0 1.0 KMW 1.0
K 12 0 1.0 PE 0.0
K13 0 1.0
*If specific data is not shown, use conventional control data.
** When C2 is negative, the C2 VAR is set to Pelec in per unit on MBASE.
27.6 Running Extended Term Simulations

The extended term initialization takes place during activity MSTR, which does not run all
the data checking run by activity STRT. The user should first clean up all data with the
standard dynamic simulation activities and make sure there are no units out-of-limits. It is
also prudent to insure, with the standard RUN activity, that steady state can be run for sev-
27-19
Running Extended Term Simulations Program Application Guide: Volume II
eral seconds without system movement. Furthermore, it is beneficial to run at least one
standard simulation for 10 sec with activities STRT/RUN before making simulations with
MSTR/MRUN to make sure the data itself does not cause dynamic instabilities.
Comparison at switching times when using the same time step, DELT, between RUN and
MRUN will show most models in MRUN to react slightly faster. This is because of the
instantaneous pass-through of a fraction of the input changes to the output in the integra-
tion algorithm. This slight difference does not, however, change results for system simula-
tions because the two methods converge quickly to the same output. Repetition of
simulations using both techniques with the same time step will show essentially identical
results.
Though designed for running simulations when local intermachine oscillations have
damped out, a user may wish, on occasion, to increase the time step to decrease computer
running time for other types of simulations. Care should be taken in doing this because the
algorithm used does add damping to higher frequency modes as the time step increases.
This is especially true when the time step, DELT, is made equal to or greater than the first
user-set threshold, DLTBKW. Figure 27-11 demonstrates the amount of artificial damping
that may be added. The three simulations show the result of the following runs:
1. 1/2 cycle time step for the entire simulation (using either RUN or MRUN).
2. 1/2 cycle time step for 2 sec and then changing to 2-1/2 cycles.
3. 1/2 cycle time step for 2 sec and then changing to 3 cycles.
For these simulations, DLTBKW was set to 3 cycles.

A second user-set threshold, DLTEXT, defines when the model goes to its uniform island
frequency assumptions. In this mode, the rotor angle of each generator is adjusted to yield
electrical torque equal to mechanical torque less the unit’s prorated share of an island
acceleration torque. Shown algebraically,
Mi
T ei = T mi – ---------- T ai
Mi
A global island frequency is calculated as
1
f =   T mi – T ei  ---------- dt
Mi
and fed to all models affected such as motors, governors, etc. In this running mode, local
intermachine oscillations are totally ignored as shown dramatically in Figure 27-12. A
machine with poorly damped high-frequency oscillations was purposely chosen. Increas-
ing the time step to 0.25 sec (above the user set DLTEXT value of 0.20 sec) at 5 sec
removes the local intermachine oscillations while the general trend remains.
Another phenomenon to be noted: When plotting results after using a time step larger than
DLTEXT, the values placed in channels for angles have different meanings. Absolute
angles (i.e., the user is not using RELANG) are automatically relativized to an island
weighted average angle. Relative angles (i.e., the user is using RELANG) are calculated
based on the automatically relativized absolute angles.
All major switching events, including line tripping, bus disconnection and generator trip-
ping, should occur only when running with a small DELT. This recommendation makes
27-20
good engineering sense, because the user will want to see any local problems that these
disturbances cause. Switching back to the small DELT is not a problem as the models have
been written to accommodate it. The user cannot always anticipate a relay operation so it
is recommended when using relays to take snapshots during the run. A bus disconnection
or line tripping that causes a new island when running with a time step greater than
DLTEXT must be avoided because it can lead to erroneous results. PSS®E does not update
island information at every switching. Tap movement and small load-shedding can be tol-
erated even with large time steps.
27-21
Running Extended Term Simulations Program Application Guide: Volume II
TEST SYSTEM
NW/SW INTERTIE SCHEDULE - 3280 MW AC / 1495 MW DC
15:40
THU OCT 03, 1991
CHNL# 259: [V-MALIN]
1.2000 FILE:outmid9 0.95000
1.2000 FILE:outmid1 0.95000
1.2000 FILE:outmid 0.95000
15.000
13.500
1-1/2-Cycle Time Step
12.000
3-Cycle Time Step
10.500
9.0000
7.5000
TIME
6.0000
1/2-Cycle Time Step
4.5000
3.0000
1.5000
0.0
Figure 27-11. Three Simulations of Artificial Damping
27-22
DAMPING INTRODUCED BY INCREASING DELT
14:47
FRI SEP 20, 1991
SPEED
CHNL# 43: [SPDHCRAW U8]
0.0 FILE: outnew3 -0.0010
25.000
22.500
20.000
Delta = 0.1 sec
17.500
15.000
1/2-Cycle Time Step
Delta = 0.25 sec
12.500
TIME
10.000
7.5000
5.0000
2.5000
0.0
Figure 27-12. Three Simulations of Artificial Damping Introduced by Increasing DELT
27-23
The dc Line Auxiliary Signals Program Application Guide: Volume II
When switching to a large time step, which causes the uniform island frequency mode, the
user should increase the maximum number of iterations allowed. This increase of itera-
tions is required at the t+ calculation of the network because the speed of each unit is
adjusted to the island frequency. This adjustment causes a movement in machine angles
and thus power readjustment in the network.
27.7 The dc Line Auxiliary Signals

Changes in dc line auxiliary signal output are not automatically updated in the extended
term mode during network solution. The extended term user must use a computer editor to
copy the subroutine call to the auxiliary signal model from CONEC to CONET, along
with the appropriate statement that sets the dc line input equal to the auxiliary signal out-
put. The first letter of the auxiliary signal subroutine name must then be changed to a T.
For example, suppose the user has the following in subroutine CONEC:
CALL PAUX1 (1, 196, 71, 9)
VAR(31) = VAR(9)
CALL CDC4 (1, 206, 73, 31, 3)
Take the first two statements and place them before the TDC4 call in CONET and modify
the PAUX1 call to be TAUX1. In addition, an IF(MIDTRM) should be placed before these
two statements and a FIN after them so that the two statements will not be executed for
state space simulations. The resultant CONET would be:
IF(MIDTRM)
CALL TAUX1 (1, 196, 71, 9)
VAR(31) = VAR(9)
FIN
CALL TDC4 (1, 206, 73, 31, 3)
After making the above changes, CONET should be recompiled and CLOAD4 executed.
27.8 Converting a State Space Model to Extended Term
27.8.1 Background, Prerequisites, and Requirements

Conversion of a state space explicit integration approach to the trapezoidal (implicit) inte-
gration approach is straightforward in the case of models without feedback loops. How-
ever, the proper modeling of limits, the representation of nonlinear effects, and the logic
for bumpless switching of time steps requires careful coding.
The trapezoidal approach, referred to as the Z-form of the models, uses all the same arrays
as the state space modeling. These arrays are listed in Table 19-1 of the PSS®E Program
Operation Manual. The model writer must become familiar with one new array,
STORMT. This array, which is present in the PSS®E COMMON memory, is dimensioned
twice as large as the STATE array. For each STATE in a model, two STORMT locations
have been reserved; therefore, in each model this array is initially indexed by twice the
STATE number minus one. Its contents include the bigger value of inputs at TPAUSE– for
27-24
Program Application Guide: Volume II Converting a State Space Model to Extended Term
averaging with TPAUSE+ inputs and it is also available for other items the user feels
appropriate. Other items stored in the STORMT array:
• Variables used to test convergence in cases where a direct solution of the model
cannot be obtained and a loop is required.
• Filters used on current injections of certain models.
This STORMT array shares space with the STATE and DSTATE arrays. Therefore, the
user cannot use the STATE and DSTATE arrays in the extended term modeling.
There are four new variables that the extended term model writer must use. These vari-
ables are all present in the PSS®E COMMON memory. The first is the logical MIDTRM.
When this is set to true, the Z-form representation of the model is being used. The logical
DLTCHG is set to true to indicate that a change in DELT has occurred. The old value of
DELT is stored in DLTOLD. The last new variable is the logical INSITR. This variable
indicates whether the user model should update the STORE arrays (Z-form integration) or
the user model is being called during the power flow iterative process and should only cal-
culate the model’s outputs.
As in the state space representation, certain flags are set by the program before calling the
models. Some of the flags take on different meanings in the Z-form representation and
only the differences are mentioned here. The significance of the flags MODE and
KPAUSE are as follows:
MODE = 1 The model must initialize the array STORE with the values to
be used by the delay operator of the Z-form representation.
MODE = 2 The model must use the appropriate inputs, compute the solu-
tion of the results for the entire model, and place the appropri-
ate values in the STORE array. In this MODE, the present
value of the model output (VOTHSG, VUEL, VOEL, EFD,
PMECH, etc.) must be set.
MODE = 3 The model must calculate new values of STORE only when
DLTCHG is true. The model does not calculate new values of
the output quantities.
KPAUSE = 0 Models are being called at a normal time step.
KPAUSE = 1 Models are being called for the value of simulation time equal
to TPAUSE–. Inputs must be saved in the STORMT array.
KPAUSE = 2 Models are being called at a time step immediately following a
pause. Inputs must be averaged with values saved in the
STORMT array.
User models in the Z-form representation are also called during the power flow solution to
set outputs such as VOTHSG, VUEL, VOEL, EFD, and PMECH, etc. The KPAUSE flag
should be checked to decide whether inputs should be averaged. This part of the model
code is similar to MODE 2 except that the STORE array is not updated.
27-25
Converting a State Space Model to Extended Term Program Application Guide: Volume II
27.8.2 Writing the Basic Model

The first step in adding the Z-form (trapezoidal) integration to a state space model is to
redraw the block diagram with the trapezoidal Z-form function replacing the integrators.
The Z-form expression for the trapezoidal integration algorithm is shown in Figure 27-13.
The second step identifies all inputs, and the third step allocates locations in the STORE
array for variables being integrated and locations in the STORMT array for inputs.
–1
1 T 1 + Z 
--- -----------------------
S 2  1 – Z –1 
State Space Integrator
Trapezoidal Integrator Z-Form Function
Figure 27-13. Z-Form Expression of Trapezoidal Integration Algorithm
Many of the model blocks have standard forms. Figures 27-14 through 27-17 show exam-
ples of these standard state space blocks converted into a Z-form representation and the
code needed to model them. There will be times when the model writer will need to derive
the Z-form representation and the associated code for nonstandard model logic. As stated
earlier, in the case of models without feedback loops, conversion to Z-form is straight for-
ward. The conversion process is probably best demonstrated by example.
The simple exciter written in state space form in Chapter 21 of the PSS®E Program Oper-
ation Manual will be converted first (Figure 27-18).
Later a model with a feedback loop will be converted to demonstrate complexities that
may be introduced. Figure 27-19 shows the block diagram redrawn with the Z-form trape-
zoidal function replacing the integrator.
The model inputs are:
ECOMP(I)=EC which is stored in STORMT(K2)at t–
VREF(I)+VOTHSG(I)=VOTHSG+VREF which is stored in STORMT(K2+1)at t–
where K2=2K-1 and K is the STATE index. CON assignments are the same as in state
space, where:
TR in CON(J)
K in CON(J+1)
Te in CON(J+2)
Two STORE array entries are needed (as two STATE entries were needed in state space).
The STORE entries do not define the same quantities as the state variables. In this simple
model, the STATE variables are the outputs of the blocks; they are not stored. The STORE
variables are internal quantities. The STORE allocation is:
The voltage sensing block uses STORE(K).
The voltage amplifier block uses STORE(K+1).
From Figure 27-15, for the amplifier block, the MODE 1 code needed is:
27-26
STORE=Y(2T/t)
HI
 +
X K  Y
2
+ LO
HI
Z-1
K
X Y
S
+ +

LO
State Space Z-Form Representation, Where t = DELT
Mode 1:
STORE = Y
Mode 2:
Y = K
------- X + STORE
2
CONDITIONAL
(Y.GE.HI)
Y=HI
STORE=HI (If limits are not pres-
FIN ent, omit the condi-
(Y.LE.LO) tional and use only the
Y=LO statement following the
STORE=LO
FIN
(OTHERWISE)
FIN K
STORE = Y + X ------
Mode 3 (only when DLTCHG): 2
K
old
Y = X --------------- + STORE old
2
K new
STORE new = Y – X ------------------
2
Figure 27-14. Z-Form Code for Integrator
27-27
HI
+ 1
----------------
X K  Y
1 + 2T ------
HI 
+ LO
Z-1
K
X 1 + sT Y
+ – 2T
 1 – ------
LO 
Mode 1:
2T
STORE = Y  -------
 
Mode 2:
 
1
Y =  XK + STORE   ----------------
 2T
 1 + -------

CONDITIONAL
(Y.GE.HI)
Y=HI
STORE=HI (If limits are not pres-
FIN ent, omit the condi-
(Y.LE.LO)  2T
------
Y=LO   tional and use only the
STORE=LO
FIN
 2T
(OTHERWISE) ------
 
FIN 2T
STORE = XK – Y  1 – ------
 
Mode 3 (only when DLTCHG):
 
 1 
Y =  XK + STORE  --------------------
old  2T

1 + -----------
  old
 2T 
STORE = Y  1 + ------------- – XK
new   new
Figure 27-15. Z-Form Code for Lag with Gain
27-28
1
2T + -------------------
X 1 + --------1-  2T Y
 1 + --------2-

+
Z-1
1 + T1 s
-----------------
- 2T + – 2T
X
1 + T2 s
Y 1 – --------1-  1 – --------2-
 
Mode 1:
2
STORE = Y  ---  T 2 – T 1 
 
Mode 2:
 
2T 1  1 
Y = X  1 + ---------- + STORE  -------------------
  2T 2
 1 + ----------

2T 2T
STORE = X  1 – ---------1- – Y  1 – ---------2-
     

 
 2T 1   
1
 --------------------
Y = X  1 + ----------
- + STORE
  old old  2T 2 
 1 + ---------- -
  old
 2T 2   2T 1 
STORE new = Y  1 + -----------
- – X  1 + -----------
-
  new   new
Figure 27-16. Z-Form for Lead/Lag or Lag/Lead
27-29
1
2K + ---------------------
X -------   1 + ----- 2T Y
 
-
- 
Z-1
Ks + +
X --------------- Y 2T
1 + Ts  1 – ------

Mode 1:
2K
STORE = X  -------
  
Mode 2:
 
2K 1
Y = X ------- – STORE  ----------------
  2T
 1 + -------

2K 2T
STORE = X ------- + Y  1 – -------
  

 
2K- – STORE  1 
Y = X ----------
old  --------------------
 old 2T-
 1 + ----------
  old
2K  2T 
STORE new = X ------------- – Y  1 + -------------
  
new new
Figure 27-17. Z-Form Code for Rate Feedback
27-30
Vref
+
1 – K
EC ------------------ S EFD
1 + sT R Esensed Verror 1 + sTe E 0
+
Other Signals
VOTHSG
Figure 27-18. Sample Exciter Model
Voltage Sensing Block Amplifier Block

Vref
+ 1
1 ------------------
+ ------------------- Esensed – Verror + E
EC  2T R  K  2T EFD
1 + ---------- 1 + --------e- 0
  
+ +
-1
Z-1 Other
Signals
VOTHSG   2T
+ – 2T R  1 – --------e-
 1 – ---------
- 

Figure 27-19. Z-Form Block Diagram of Sample Exciter
where Y is the block output, T is the time constant, and  is the time step DELT. The For-
tran code using the locations assigned above would be:
STORE(K+1)=EFD(I)*(2.*CON(J+2)/DELT)
Similarly, for the voltage sensing block, the Fortran code would be
STORE(K)=ECOMP(I)*(2.*CON(J)/DELT)
As in state space, statements must also be added to initialize references.

The first items addressed in MODE 2 code should be inputs. The model must handle the
storage of inputs at a TPAUSE- and the averaging of inputs at TPAUSE+ using the flag
KPAUSE. The following example code (based again on Figure 27-15) would be typical for
MODE 2:
WHEN (KPAUSE.EQ.1)
STORMT(K2)=ECOMP(I)
STORMT(K2+1)=VREF(I)+VOTHSG(I)
FIN
ELSE
27-31
C HANDLE INPUTS FIRST

S1=ECOMP(I)
S2=VREF(I)+VOTHSG(I)
IF (KPAUSE.EQ.2)
S1=(S1+STORMT(K2))*0.5
S2=(S2+STORMT(K2+1))*0.5
FIN
C VOLTAGE SENSING BLOCK
OUT1=(S1+STORE(K))/(1.+2.*CON(J)/DELT)
STORE(K)=S1-OUT1*(1.-2.*CON(J)/DELT)
C AMPLIFIER BLOCK
OUT2=(CON(J+1)*(S2-OUT1)+STORE(K+1))/(1.+2.*CON(J+2)/DELT)
STORE(K+1)=CON(J+1)*(S2-OUT1)-OUT2*(1.-2.*CON(J+2)/DELT)
C LIMIT CLIP
IF (OUT2.LT.0.) OUT2=0.
EFD(I)=OUT2
FIN
MODE 2 has responsibility for setting EFD. As stated in Section 27.8.1, EFD must also be
set during the power flow iterative process. The code between ELSE and the final FIN of
MODE 2 would have to be repeated there with the exception of the two statements setting
the STORE values. MODE 3 is responsible for resetting the STORE array for changes in
DELT. Based on Figure 27-15, the code would be as follows:
C AVERAGE INPUTS
S1=(ECOMP(I)+STORMT(K2))*0.5
S2=(VREF(I)+VOTHSG(I)+STORMT(K2+1))*0.5
OUT1=(S1+STORE(K))/(1.+2.*CON(J)/DLTOLD)
STORE(K)=OUT1*(1.+2.*CON(J)/DELT)-S1
C AMPLIFIER BLOCK
OUT2=(CON(J+1)*(S2-OUT1)+STORE(K+1))/(1.+2.*CON(J+2)/DLTOLD)
STORE(K+1)=OUT2*(1.+2.*CON(J+2)/DELT)-CON(J+1)*(S2-OUT1)
The entire extended term model code is shown in Figure 27-20.
27-32
WHEN (MIDTRM)
C
. CONDITIONAL
C
. . (INSITR)
C
. . . S1=ECOMP(I)
. . . S2=VREF(I)+VOTHSG(I)
. . . IF (KPAUSE.EQ.2)
. . . . S1=(S1+STORMT(K2))*0.5
. . . . S2=(S2+STORMT(K2+1))*0.5
. . . ...FIN
. . . OUT1=(S1+STORE(K))/(1.+2.*CON(J)/DELT)
C AMPLIFIER BLOCK
. . . OUT2=(CON(J+1)*(S2-OUT1)+STORE(K+1))/(1.+2.*CON(J+2)/DELT)
. . . IF (OUT2.LT.0.) OUT2=0.
. . . EFD(I)=OUT2
. . ...FIN
C
. . (MODE.EQ.1)
C
C AMPLIFIER BLOCK
. . . STORE(K+1)=EFD(I)*(2.*CON(J+2)/DELT)
. . . STORE(K)=ECOMP(I)*(2.*CON(J)/DELT)
. . . VREF(I)=ECOMP(I)+EFD(I)/CON(J+1)
. . ...FIN
C
. . (MODE.EQ.2)
C
. . . WHEN (KPAUSE.EQ.1)
. . . . STORMT(K2)=ECOMP(I)
. . . . STORMT(K2+1)=VREF(I)+VOTHSG(I)
. . . ...FIN
. . . ELSE
C HANDLE INPUTS FIRST
. . . . S1=ECOMP(I)
. . . . S2=VREF(I)+VOTHSG(I)
. . . . IF (KPAUSE.EQ.2)
. . . . . S1=(S1+STORMT(K2))*0.5
. . . . . S2=(S2+STORMT(K2+1))*0.5
. . . . ...FIN
. . . . OUT1=(S1+STORE(K))/(1.+2.*CON(J)/DELT)
. . . . STORE(K)=S1-OUT1*(1.-2.*CON(J)/DELT)
C AMPLIFIER BLOCK
. . . . OUT2=(CON(J+1)*(S2-OUT1)+STORE(K+1))/(1.+2.*CON(J+2)/DELT)
. . . . STORE(K+1)=CON(J+1)*(S2-OUT1)-OUT2*(1.-2.*CON(J+2)/DELT)
C LIMIT CLIP
. . . . IF (OUT2.LT.0.) OUT2=0.
. . . . EFD(I)=OUT2
. . . ...FIN
. . ...FIN
C
. . (MODE.EQ.3)
C
C AVERAGE INPUTS
. . . S1=(ECOMP(I)+STORMT(K2))*0.5
. . . S2=(VREF(I)+VOTHSG(I)+STORMT(K2+1))*0.5
. . . OUT1=(S1+STORE(K))/(1.+2.*CON(J)/DLTOLD)
. . . STORE(K)=OUT1*(1.+2.*CON(J)/DELT)-S1
C AMPLIFIER BLOCK
. . . OUT2=(CON(J+1)*(S2-OUT1)+STORE(K+1))/(1.+2.*CON(J+2)/DLTOLD)
. . . STORE(K+1)=OUT2*(1.+2.*CON(J+2)/DELT)-CON(J+1)*(S2-OUT1)
. . ...FIN
. ...FIN
...FIN
Figure 27-20. Extended Term Code for Sample Exciter Model
27-33
27.8.3 Models With Feedback and Limits

The simplified model of Figure 27-18 is now assumed to have nonwindup limits and a
feedback. The new model in state space form is shown in Figure 27-21.
As with the previous example, the first step would be to draw the Z-form block diagram as
Vref EFDMAX
+
1 – + K
EC ------------------   EFD
1 + sT R Esensed Ve 1 + sTe
+ –
Other Signals EFDMIN
VOTHSG
KFs
------------------
1 + TF s
Figure 27-21. State Space Block Diagram of Exciter with Feedback and Nonwindup Limits
Voltage Sensing Block Amplifier Block

Vref
+ EFDMAX
1 1
+ ------------------- – + + ------------------
EC  2T R   K  2T EFD
1 + --------- - Ve Vin 1 + --------e-
+  + –  
EFDMIN
Z-1 Other Z-1
Signals
VOTHSG   2T
+ – 2T R  1 – --------e-
 1 – ---------
- 

1 +
Vfb ------------------- 2K F
Note  = DELT 2T F  ----------
1 + --------- -
- 

Z-1
2T F + +
1 – ---------
- 

Feedback Block
Figure 27-22. Z-Form Block Diagram of Exciter with Feedback and Nonwindup Limits
27-34
No additional inputs have been added; therefore, step 2 is the same as the simplified
model. Step 3 involves the assignment of the new delay operator required for the feedback
block to STORE (K+2). The additional CON assignments are:
EFDMAX in CON(J+3)
EFDMIN in CON(J+4)
KF in CON(J+5)
TF in CON(J+6)
The new feedback path requires an additional statement in MODE 1 initialization. Based
on Figure 27-17 it is:
STORE(K+2)=EFD(I)*2.*CON(J+5)/DELT
When a loop is present, the code required cannot simply be copied from the typical block
examples. This is because the integration technique is a closed form solution; that is, out-
put changes instantaneously with input changes. With a loop, the input to the forward path
is a function of the output of the feedback, which is itself a function of the input. This
simultaneous transmission around the loop requires a direct solution of the entire loop
equation. The model writer is required to (a) choose a point in the loop to solve for and (b)
to go back and make sure that the limits are satisfied.
In the example, the voltage sensing block is not in the loop so the code for it and the input
code is unchanged from the previous simplified example. Ignoring limits and using the
output equations shown in Figures 27-14 through 27-17 gives three equations describing the
feedback loop:
V in = V e – V fb (1.3)
EFD =  K  V in + STORE  K + 1     1. + 2T e    (1.4)
2K F
V fb =  EFD  ---------- – STORE  K + 2    1. + 2T F    (1.5)
  
Substituting ((1.3)) into ((1.4)) results in:

EFD =  K   V e – V fb  + STORE  K + 1     1. + 2T e    (1.6)
Substituting ((1.5)) into ((1.6)) and solving for EFD results in:
2T F
K   V e + STORE  K + 2   + STORE  K + 1    1 + --------- -
  
EFD = ----------------------------------------------------------------------------------------------------------------------------------------------
TF 2T 2KK
 1 + 2 ------    1 + --------e- + -------------f
     
Limits would then be applied. The user should then, starting with the calculated and lim-
ited output, go around the loop using the standard block equations and calculate block out-
puts and, if in mode 2, set the STOREs. In general, feedback loops have negative signs so
that any limits hit in the forward path reduce the output, thereby reducing the feedback
that, in turn, would (if it could) cause a higher model output. Thus, usually, limits are self-
27-35
sustaining in negative feedback systems. The model writer must verify that all limits in
series, either directly or via feedback, are satisfied. The MODE 2 code for the amplifier
and feedback blocks is shown in Figure 27-23.
C SOLVE FOR EFD AND APPLY LIMITS
EFD(I)=CON(J+1)*(VE+STORE(K+2))+STORE(K+1)*(1.+2.*CON(J+6)/DELT)
EFD(I)=EFD(I)/((1.+2.*CON(J+6)/DELT)*(1.+2.*CON(J+2)/DELT)+2.*CON(J+1)*CON(J+5)/DELT)
C CHECK LIMITS
HIGH=EFD(I).GE.CON(J+3)
WHEN(HIGH)
EFD(I)=CON(J+3)
FIN
ELSE
LOW=EFD(I).LE.CON(J+4)
IF(LOW)EFD(I)=CON(J+4)
FIN
C FEEDBACK LOOP
VFB=(EFD(I)*2.*CON(J+5)/DELT-STORE(K+2))/(1.+2.*CON(J+6)/DELT)
STORE(K+2)=EFD(I)*2.*CON(J+5)/DELT+VFB*(1.-2.*CON(J+6)/DELT)
C AMPLIFIER BLOCK
VIN=VE-VFB
CONDITIONAL
(HIGH)STORE(K+1)=CON(J+3)*2.*CON(J+2)/DELT
(LOW)STORE(K+1)=CON(J+4)*2.*CON(J+2)/DELT
(OTHERWISE)STORE(K+1)=CON(J+1)*VIN-EFD(I)*(1.-2.*CON(J+2)/DELT)
FIN
FIN
Figure 27-23. Mode 2 of Feedback Exciter
27-36

PAGV2

Uploaded by

Copyright:

Available Formats

PAGV2

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PAGV2

Uploaded by

Copyright:

Available Formats

PSS®E 33.

Siemens Industry, Inc.

Dynamic Simulation Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1

Dynamic Simulation Setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-1

13.2.18 Initial Condition Snapshot .........................................................................13-36

Generator Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-1

Static var Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1

Excitation System and Controller Models . . . . . . . . . . . . . . . . . . . . . . . . . . 16-1

16.2 The dc Exciters ........................................................................................................16-1

16.7.7 Model PTIST1 ...........................................................................................16-35

Speed Governor System Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-1

17.17 Model GGOV1 .......................................................................................................17-35

Modeling dc Transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-1

18.11.11Special Control Functions .......................................................................18-81

Transmission Line and Other Relay Models . . . . . . . . . . . . . . . . . . . . . . . . 19-1

Load Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1

Generic Wind Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-1

Photovoltaic (PV) System Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1

Elementary Blocks for Handling Transfer Functions in Dynamic Models 23-1

23.6 Wash-Out Block ....................................................................................................23-10

Other Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-1

24.10.3 DCLINE Module ........................................................................................24-40

Data Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-1

Linear Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26-1

26.10 Plotting Results .....................................................................................................26-28

Extended Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-1

12-1 Differential Equations Describing Behavior of Simple Planetary System . . . . . . . . . .12-2

15-9 Dynamic Simulation Setup Main Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15-16

17-7 Elastic versus Inelastic Hydro Governor Models Comparison . . . . . . . . . . . . . . . . 17-19

18-30 Example Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18-57

19-18 Relay Model RXR1 Logic Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-31

23-34 FORTRAN Code for Initialization of Lead-Lag Block (MODE=1) . . . . . . . . . . . . . . .23-12

23-67 Second Order Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-24

25-11 Sample Data for Excitation System Data Verification . . . . . . . . . . . . . . . . . . . . . . .25-17

27-5 Time-Domain Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27-6

12.1 Basic Dynamic Simulation

2. Determination of a set of values of constant and variable parameters describing, in

The overall process is best illustrated by an example.

Figure 12-1. Differential Equations Describing Behavior of Simple Planetary System

Figure 12-2. Graphical Illustration of Simulation Process

Calculate time derivatives,

Calculate outputs and

Update values of x, y, vx, vy using (1.2):

Figure 12-3. Simple Simulation of Planetary Motion Modeled as in Figure 12-1

The program shown in Figure 12-5 consists of two sections:

Figure 12-5. Rewritten Dynamic Simulation Program

12.2 Simulation Model Library

Figure 12-6. Basic Program Structure of PSS®E Dynamic Simulation Section

12.3 Numerical Integration Stability

to the very simple differential equation:

and substituting this into (1.3), with the result:

which shows that:

Figure 12-7. Comparison of Numerical Integrations of Equation (1.4)

12.4 Classification of Variables

12.4.1 General Classification

Figure 12-8. Differential Equations for Generator Connected to Infinite Bus

Calculate igen from:

Pgen = Real (E  × igen)

12.4.2 PSS®E Parameter and Variable Classification

Table 1: Dynamic Simulation Arrays