ANL DIS-13 06 Modeling AS PSH PDF

ANL/DIS-13/06
Modeling Adjustable Speed Pumped

Storage Hydro Units Employing Doubly-Fed
Induction Machines
Decision and Information Sciences

About Argonne National Laboratory
Argonne is a U.S. Department of Energy laboratory managed by UChicago Argonne, LLC
under contract DE-AC02-06CH11357. The Laboratory’s main facility is outside Chicago,
at 9700 South Cass Avenue, Argonne, Illinois 60439. For information about Argonne
and its pioneering science and technology programs, see www.anl.gov.
Availability of This Report

This report is available, at no cost, at http://www.osti.gov/bridge. It is also available
on paper to the U.S. Department of Energy and its contractors, for a processing fee, from:
U.S. Department of Energy
Office of Scientific and Technical Information
P.O. Box 62
Oak Ridge, TN 37831-0062
phone (865) 576-8401
fax (865) 576-5728
[email protected]
Disclaimer
This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States
Government nor any agency thereof, nor UChicago Argonne, LLC, nor any of their employees or officers, makes any warranty, express
or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus,
product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific
commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply
its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of
document authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof,
Argonne National Laboratory, or UChicago Argonne, LLC.
This report is being disseminated by the Department of Energy. As such, this document was prepared in compliance with Section 515 of
the Treasury and General Government Appropriations Act for Fiscal Year 2001 (Public Law 106-554) and Information Quality Guidelines
issued by the Department of Energy. Although this report does not constitute “influential” information, as that term is defined in DOE’s
Information Quality Guidelines or the Office of Management and Budget’s Information Quality Bulletin for Peer Review, the study was
reviewed both internally and externally prior to publication. For purposes of external review, the study benefited from the advice and
comments of an advisory working group consisting of more than 30 experts from the industry, government, and research institutions.
ANL/DIS-13/06
Modeling Adjustable Speed Pumped

Storage Hydro Units Employing Doubly-Fed
Induction Machines
prepared for
U.S. Department of Energy – Wind and Water Power Technologies Office
prepared by
Vladimir Koritarov and Leah Guzowski
Decision and Information Sciences, Argonne National Laboratory
James Feltes, Yuriy Kazachkov and Bo Gong

Siemens PTI
Bruno Trouille and Peter Donalek

MWH Americas
August 2013
This page intentionally left blank.
Preface
This report is one of several reports developed during the U.S. Department of Energy
(DOE) study on the Modeling and Analysis of Value of Advanced Pumped Storage
Hydropower in the United States. The study is led by Argonne National Laboratory in
collaboration with Siemens PTI, Energy Exemplar, MWH Americas, and the National
Renewable Energy Laboratory. Funding for the study was provided by DOE’s Office of
Energy Efficiency and Renewable Energy (EERE) through a program managed by the
EERE’s Wind and Water Power Technologies Office (WWPTO).
The scope of work for the study has two main components: (1) development of vendor-
neutral dynamic simulation models for advanced pumped storage hydro (PSH)
technologies, and (2) production cost and revenue analyses to assess the value of PSH
in the power system. Throughout the study, the project team was supported and guided
by an Advisory Working Group (AWG) consisting of more than 30 experts from a diverse
group of organizations, including the hydropower industry and equipment manufacturers,
electric power utilities and regional electricity market operators, hydro engineering and
consulting companies, national laboratories, universities and research institutions,
hydropower industry associations, and government and regulatory agencies.
The development of vendor-neutral models was carried out by the Advanced

Technology Modeling Task Force Group (TFG) led by experts from Siemens PTI with
the participation of experts from other project team members. First, the Advanced
Technology Modeling TFG reviewed and prepared a summary of the existing dynamic
models of hydro and PSH plants that are currently in use in the United States. This is
published in the report Review of Existing Hydroelectric Turbine-Governor Simulation
Models. The review served to determine the need for improvements of existing models
and for the development of new ones.
While it was found that existing dynamic models for conventional hydro and PSH plants
allow for accurate representation and modeling of these technologies, it was concluded
that there is a need for the development of dynamic models for two PSH technologies for
which at present there are no existing models available in the United States. Those two
technologies are (1) adjustable speed PSH plants employing doubly-fed induction
machines (DFIMs) and (2) ternary PSH units. The Advanced Technology Modeling TFG
developed vendor-neutral models of these two PSH technologies, which are published in
two reports: (1) Modeling Adjustable Speed Pumped Storage Hydro Units Employing
Doubly-Fed Induction Machines and (2) Modeling Ternary Pumped Storage Units.
Extensive testing of newly developed models was performed using the Siemens PTI’s
standard test cases for the Power System Simulator for Engineering (PSS®E) model as
well as the Western Electricity Coordinating Council’s (WECC’s) modeling cases for the
Western Interconnection that were provided in PSS®E format. The results of model
i
testing are presented in the report Testing Dynamic Simulation Models for Different
Types of Advanced Pumped Storage Hydro Units.
In addition to the project team members and DOE, all these reports have been reviewed
by members of the AWG, and their comments and suggestions have been incorporated
into the final versions of the reports. Parts of these reports will also be included in the
final report for the entire study to illustrate the model development component of the
work.
ii
Acknowledgments
The authors would like to acknowledge the support and guidance provided to the project
team by the staff and contractors of the U.S. Department of Energy Office of Energy
Efficiency and Renewable Energy’s (DOE/EERE’s) Wind and Water Power
Technologies Office (WWPTO), including Michael Reed, Rajesh Dham, Charlton Clark,
Rob Hovsapian, Patrick O’Connor, Richard Gilker, and others. The authors are also
grateful to the members of the Advisory Working Group for their excellent collaboration
and efforts in advising the project team and guiding the study. The Advisory Working
Group included a broad spectrum of global pumped storage hydropower specialists.
Rajesh Dham, Charlton Clark, Rob DOE/EERE – Wind and Water Power Technologies
Hovsapian, Patrick O'Connor, Richard Gilker Office (WWPTO)
DOE – Office of Electricity Delivery and Energy
Rachna Handa
Reliability (OE)
Rahim Amerkhail Federal Energy Regulatory Commission (FERC)
Michael Manwaring, Douglas Divine National Hydropower Association (NHA)
Mark Jones, Elliot Mainzer Bonneville Power Administration (BPA)
Xiaobo Wang California Independent System Operator (CAISO)
Zheng Zhou Midwest Independent System Operator (MISO)
Matt Hunsaker Western Electricity Coordinating Council (WECC)
Tuan Bui California Department of Water Resources (CDWR)
David Harpman Bureau of Reclamation (Reclamation)
Kyle L. Jones U.S. Army Corps of Engineers (USACE)
Scott Flake, Greg Brownell Sacramento Municipal Utility District (SMUD)
Paul Jacobson, Stan Rosinski Electric Power Research Institute (EPRI)
Alan Soneda Pacific Gas and Electric Co. (PG&E)
Osamu Nagura Hitachi Mitsubishi Hydro
Teruyuki Ishizuki Toshiba Corp.
Rick Miller. Rick Jones HDR Engineering Inc. (HDR|DTA)
Jiri Koutnik, Maximilian Manderla Voith Hydro
Christophe Nicolet Power Vision Engineering (PVE)
Peter McLaren Center for Advanced Power Systems (CAPS)
Landis Kannberg Pacific Northwest National Laboratory (PNNL)
Klaus Engels E.ON Wasserkraft GmbH
Kim Johnson Riverbank Power
Steve Aubert, Le Tang ABB Switzerland Ltd
Ali Nourai DNV KEMA
iii
iv
Contents
Preface ........................................................................................................................... i
Acknowledgments...................................................................................................... iii
Section 1 Introduction ........................................................................................... 1-1
Section 2 Theory of the Dynamic Simulation Model of the Doubly-Fed

(Wound Rotor) Induction Machine ..................................................... 2-1
2.1 Steady-State Theory .................................................................................... 2-9
Section 3 Modeling Units Employing a DFIM in PSSE.................................... 3-1

3.1 The Detailed Model ...................................................................................... 3-1
3.2 The Simplified Model .................................................................................... 3-5
Section 4 Modeling Adjustable Speed Drive Hydro Pumped Storage

Units Employing a DFIM ...................................................................... 4-1
4.1 Turbine Mode of Operation ........................................................................... 4-3
4.2 Pumping Mode of Operation ......................................................................... 4-7
4.3 Rotor Speed/Gate Position Relationships ................................................... 4-10
Section 5 Bibliography .......................................................................................... 5-1
v
vi
Introduction
Section
1
Introduction
This report is a part of a series of documents to be issued in the course of a
U.S. Department of Energy (DOE) project titled “Modeling and Analysis of Value of
Advanced Pumped Storage Hydropower in the U.S.” The purpose of this report is to
propose a model structure for an adjustable speed pumped storage hydro (PSH) unit
employing a doubly-fed induction machine (DFIM).
The use of a DFIM with its rotor windings controlled by a power converter is the present
state-of-the-art design for adjustable speed hydro pumped storage units. Although the
theoretical basis for the analysis of DFIMs was developed at least 70 years ago (a paper
by Charles Concordia et al. in 1942 describes the mathematical modeling and dynamic
response of DFIM and refers to even earlier publications), it seems appropriate to
provide the fundamental equations governing their dynamic behavior so that the
following discussions on modeling these devices for stability studies are more
meaningful.
Fortunately, very significant efforts have been previously applied to determine the proper
level of modeling required for DFIMs, albeit not specifically for pumped storage units.
Previous efforts to describe the theoretical and modeling aspects of DFIMs were related
to the development of models for wind turbines employing doubly-fed induction
generators. For a period of more than 10 years starting from the early 2000s, a
significant evolution has been observed in the approach to modeling this type of
electrical machine. While the machines for pumped storage units will be much larger
than those used in wind units, and thus the higher currents and voltages may require
different power electronic devices, the control strategies and overall responses will be
similar.
Section 2 of this report includes a comprehensive review of the basics of the operation of
a DFIM, including the theory of this operation for dynamics and steady state.
Section 3 of the report provides a discussion on modeling units employing a DFIM for
dynamic stability studies. This discussion takes into account the experience gained in
the course of developing Power System Simulator for Engineering (PSS®E) dynamic
simulation models of wind units.
1-1
Introduction
Section 4 provides a discussion on modeling of adjustable speed hydro pumped storage

power units employing a DFIM. While the focus of this report is on how to model these
units in the PSS®E program, the information is provided in a manner that is also
applicable to modeling such devices in other commercial power system stability
simulation programs.
Section 5 provides a bibliography.
1-2
Theory of the Dynamic Simulation Model of the Doubly-Fed (Wound Rotor) Induction Machine
Section
2
Theory of the Dynamic Simulation Model
of the Doubly-Fed (Wound Rotor)
Induction Machine
The basics of DFIM operation are as follows. As shown in Figure 2-1, the rotor of the
machine is connected to the terminals of the machine through a power converter.
Therefore, the power converter can control the voltage, current, and frequency in the
rotor circuit.
Vt
PS+jQS Ptotal+jQtotal
Rotor
Stator
Power Pr+jQr
Converter
Figure 2-1 Configuration of a DFIM
Neglecting losses:
Pr = sPs,
where Ps is the power on the stator in W, Pr is the power on the rotor in W, and s is the
slip defined as
s = (ω1 - ωm)/ ω1,
where ω1 is the supply angular frequency in electrical rad/s, and ωm is the rotor
mechanical angular frequency in electrical redians per second.
2-1
If the generator runs below the synchronous speed (s > 0), then Pr > 0, and the
mechanical power is less than the power on the stator, Pmech < Ps. The rotor absorbs
power, and a fraction of the stator power is absorbed by the rotor circuits.
If the generator runs above the synchronous speed (s < 0), then Pr < 0 and Pmech > Ps.
The rotor generates power, and power is delivered to the grid via the stator and rotor
circuits.
Modern DFIMs employ voltage source power converters utilizing fully controlled
transistors. The rotor side converter can supply d and q components of rotor current that
determine both active and reactive power at the unit’s terminals. The line side converter
cannot only transfer the active power from or to the rotor side converter, but also could
control the reactive power injected or absorbed by the line side converter to or from the
grid. The cycloconverters utilized by the previous generation of variable speed pump
storage units did not have this capability. However, for the known applications, the line
side power converter usually operates with a unity power factor, that is, the reactive
power Qr = 0 Vars.
In a wound rotor induction machine, rotor windings, similar to those of the distributed
windings on the stator, are wound on a cylindrical laminated core with uniformly spaced
slots on the outer periphery. The terminals of the rotor windings are brought out via slip
rings and brushes. The alternating current (AC) terminals of the power converter are
connected to the rotor windings via these slip rings to provide slip control.
Figure 2-2 shows the connections and current conventions of the subject machine.
2-2
Figure 2-2 Connections and Current Conventions Used for the DFIM
In this discussion, capital E, I, and Ψ will be used for stator terminal voltage, current, and
flux linkages, respectively. Lower case e, i, and λ are used for rotor terminal voltage,
current, and flux linkages, respectively. Capital R and I subscripts indicate components
in the real and imaginary axes of the synchronous reference frame, with the imaginary
axis I leading the real axis R by 90 electrical degrees.
The following equations are used to describe the dynamic performance of the wound
rotor single cage induction machine using the motor convention.
The stator voltage equations are:
dΨR
ER = − ra I R − ωb ΨI ( D − 1)
dt
dΨI
EI = − ra I I + ωb ΨR ( D − 2)
dt
Here ωb is the base angular frequency in p.u., normalized with respect to synchronous
speed. It is introduced for generality. For most studies, ωb =1.
2-3
The rotor voltage equations are:
dλ R
eR = + r1i R + (ω m − ω b )λ I ( D − 3)
dt
dλ
e I = I + r1i I − (ω m − ω b )λ R ( D − 4)
dt
Here ωm is the rotor mechanical angular speed in p.u., normalized with respect to
synchronous speed.
Note that both components of the rotor voltage are constant in steady state, similar to the way
the R and I components of the synchronous machine stator voltage are constant in steady
state.
To be certain of this, one may use the routine conversion formula from the three-phase
voltage system to the real and imaginary components with respect to the synchronously
rotating reference frame, namely:
2
e R = [e A cos(δ ) + e B cos(δ − 120) + eC cos(δ + 120)]
3
2
e I = [e A sin(δ ) + e B sin(δ − 120) + eC sin(δ + 120)]
3
Here,
• eA,B,C are rotor physical phase voltages which in steady state are of sω frequency
where s is the rotor mechanical slip, that is,
e A = E cos( sωt )
e B = E cos( sωt − 120)
eC = E cos( sωt + 120)
• δ is the angle between the rotor and the synchronously rotating reference frame, that
is,
δ = sωt electrical degrees
In the course of the dynamic simulation, eR and eI are subject to control. Because of this,
it is important to realize that changing eR and eI actually means changing the magnitude
and frequency of the physical rotor phase voltages.
2-4
The real and imaginary components of the flux linkages, as functions of stator and rotor
currents do not interact; thus they can be written without indices as:
Stator Flux Linkage :

Ψ = − La I + Ψm ( D − 5)
Mutual Flux Linkage :
Ψm = Lm (i − I ) ( D − 6)
Wound Rotor Flux Linkage :
λ = L1i + Ψm ( D − 7)
From the equations above, the rotor currents can be expressed as a function of the stator
current and rotor flux linkage:
λ + Lm I
i = ( D − 8)
Lm + L1
and the stator flux linkages can then be expressed as:
Lm Lm L1
Ψ= λ − I ( La + ) ( D − 9) .
Lm + L1 Lm + L1
The last equation can be rewritten in a form similar to that used in classical synchronous
machine theory by introducing the transient flux linkages:
Lm
λ' = λ ( D − 10)
Lm + L1
and transient (derived) inductance:
1
L' = La + ( D − 11)
1 1
+
Lm L1
Then
Ψ = λ' − IL' ( D − 12)
2-5
The rotor voltage equations (D-3 and D-4) can be rewritten in state space form:
dλ R
= e R − r1i R − λ I pθ ( D − 13)
dt
dλ I
= e I − r1i I + λ R pθ ( D − 14)
dt
where pθ = ωm-ωb is a rotor negative slip with respect to the synchronous reference.
Substituting rotor currents from (Equation D-8) into (Equations D-13 and D-14), the rotor
transient flux linkages component time derivatives can be found as follows:
dλ R
'
2πf b Lm
= e R − λ R − ( L − L' ) I R − ω bT0 λ I ∗ pθ ) ( D − 15)
' ' '
(
dt ω bT0 '
r1
dλ I
'
2πf b Lm
= e I − λ I − ( L − L' ) I I + ω bT0 λ R ∗ pθ ) ( D − 16)
' ' '
(
dt ω bT0 '
r1
where
L
T0 =
'
ω b r1
L = L1 + Lm
The approach to saturation representation is similar to that used in synchronous

machines; that is, employing the open circuit saturation curve and entering a flux level
that is internal to the machine and empirically approximated as the flux behind transient
inductance. The convenience of the transient flux level approach is that, with the use of
explicit integration, saturation effects can be represented without requiring iteration of
the model equations with the network power flow solutions.
R and I components of the saturation factor will be calculated in proportion to the

respective components of the rotor flux linkage.
The differential equations above should be supplemented by equations describing the

inertial dynamics.
Electrical torque:
Tel = −λ I I R + λ R I I
' '
Rotor slip with respect to the base frequency:
2-6
d 2θ d T − Tel − (ω m − ω b ) * DAMP
2
= (ω m − ω b ) = mech ( D − 17)
dt dt 2H
where θ is the rotor angle measured with respect to the magnetic axis of the three-phase
stator windings, Tmech is the mechanical torque, H is the total inertia constant, and DAMP
is a damping factor.
The rotor angle is equal to:
d
ANGLE = ω b * ∫ (ω m − ω b )dt + ANGLE INIT ( D − 18)
dt
The differential and algebraic equations above define the model for the single cage
wound rotor induction machine.
Figure 2-3 1 shows the formulation of the model in the synchronous reference frame.
The equations and block diagram for the double cage wound rotor induction machine
could be derived in a similar manner.
1
de Mello, F.P., “Modeling of Induction Machines,” Siemens PTI Course Notes, 2001.
2-7
Lm
iR
L + +
Lm
e
R1 R 1
ra
L
+ ψR IR +
ER ωo 1 - 1
- L − L' T'o P
λ'R
+ P L' + + -
+ + -
λ'R
λ 'R
λ'
SAT ∆I ω o T'o pθ ω o T'o pθ
X X
| λ '|
λ'I
λ'
- + + +
EI ωo ψI 1 II - 1
L − L' λ'I
P - L' + T'o P
+ + +
ra
Lm 1
eI
R1 L
+
Lm
iI
L +
IR
−λ'I
X D Angleo
+ -
TE - 1 pθ ωo
Angle
2HP + P +
+ + +
λ'R
X TM
(pθ)o
I1 Doubly-Fed (Wound Rotor) Induction Generator Block Diagram
ψ stator flux linkage ω o = 2πf

λ' transient flux linkage T'o =
L
E, I stator voltage, current ω oR1
e, i rotor voltage, current p - Laplace operator
Figure 2-3 Block Diagram for the Single Cage DFIM
2-8
2.1 Steady-State Theory

Steady-state equations can be derived by setting the dλ/dt terms from the voltage equations
to zero:
vds = − Rs ids − ω s λq s
vqs = − Rs iqs + ω s λd s
vdr = Rr idr − sω s λq r
vqr = Rr iqr + sω s λd r
These equations can be rewritten by introducing the following phasors or quantities:
Vs = vds + jvqs

Vr = vdr + jvqr
I = i + ji
s ds qs
Ir = idr + jiqr

λs = λds + jλqs
λ = λ + jλ
r dr qr
Then, the equations for steady-state voltages and flux linkages will be as follows:
Vs = − Rs Is + jϖ s λs

Vr = Rr Ir + jsϖ s λr
λ = − L I + L I
s ss s m r
λr = − Lm Is + Lrr Ir
or
Vs = − Rs Is + j ( − X ss Is + X m Ir )

Vr = Rr Ir + js( − X m Is + X rr Ir )
or
Vs = − Rs Is − jX 1 Is − jX m ( Is − Ir )

Vr = Rr Ir − jsX m ( Is − Ir ) + jsX 2 Ir
2-9
where
X 1 = X ss − X m
X 2 = X rr − X m
or
Vs = − Rs Is − jX 1 Is − jX m ( Is − Ir )

Vr Rr 
= I r − jX m ( Is − Ir ) + jX 2 Ir
s s
The equivalent circuit shown in circuit a of Figure 2-4 fits these equations.
By substituting
Rr 1− s
= Rr (1 + ),
s s
the equivalent circuit can be changed to the form shown in circuit b of Figure 2-4.
a) r2
r1 X1 X2 S
I I
1 2

V

V I + I X12 2
1 1 2
S
b) l−S
r2
r1 X1 X2 r2 S
I I
1 2
 
V
V I + I X12 2
1 1 2
S
Figure 2-4 Equivalent Circuit of a Wound Rotor Induction Machine
In the last two diagrams, V1 = Vs, and V2 = Vr.
It is worth remembering that the notion of the equivalent circuit makes sense if dealing
with phasors of the same fundamental frequency from both stator and rotor sides. Thus
V1 = Vs and V2 = Vr are voltage phasors of the same frequency. To fit with the underlying
physics and to obtain the correct value of the stator current which is of the fundamental
frequency, we have to use not V2 but V2/s, and not r2 but r2/s.
2-10
Assuming Rs = 0, the complex stator and rotor currents can be derived as
X
(1 + jξ )Vs − j m Vr
Rr
Is = −
X ss ( j − ξσ )
X m  X ss 
ξ Vs − Vr
X rr Rr
I2 = − j
X ss ( j − ξσ )
where
2
Xm
σ = 1− µ = 1−
X ss X rr
Rr
ρ2 =
X rr
s
ξ=
ρ2
For the sake of simplicity, the base voltage for the rotor is selected according to the ratio
Vsbase X
= m.
Vrbase Rr
Then the expression for the stator current phasor can be rewritten as
(1 + jξ )Vs − jVr
Is = − ,
X ss ( j − ξσ )
and electrical torque as
Tel = Re(Vs Is ) .
Assuming the stator voltage phasor as reference, then
Vs = Vs ,
that is, the phasor of the stator voltage is oriented along the real axis, R. The position of
the phasor of the rotor voltage will then be displaced by the angle θ+90o; that is,
j (θ +π / 2 )
Vr = Vr e .
2-11
Then electrical torque can be found as
VsVr Vs
2
ξµ
Tel = sin(θ − α ) + ,
X ss 1 + (ξσ ) 2 X ss 1 + (ξσ ) 2
where α = atan(ξσ) is a function of the slip s.
The first term of Tel is a synchronous torque which at ξ = 0, that is, at s = 0, is equal to
synchronous torque of the synchronous machine. The second term of Tel is the
asynchronous torque identical to the electrical torque of the induction machine with a
squirrel cage rotor.
Figure 2-5 parts 5a, 5b, and 5c show Tel versus θ characteristics for V1/V2 = 1.5, 1.0, and
0.5, respectively; σ = 0.2; and different values of ξ. One can see that, by virtue of the
asynchronous component, a wound rotor induction machine can operate as a generator
(Tel >0) for θ < 0 and as a motor for θ > 0.
2-12
Angular Characteristics of Electrical Torque
2.5
Elec. Torque, p.u.

2 ξ=−1.0
1.5
1 ξ=−0.5
0.5
0 ξ=0.0
-200 -100 -0.5 0 100 200
ξ=0.5
-1
-1.5 ξ=1.0
-2
-2.5
Angle, degrees
V2=1.5V1
Figure 5, a
Figure 2-5 Electrical Torque Tel versus Phase Shift Angle θ for Ratios of Stator and Rotor
Voltages V1/V2 = 1.5, 1.0, and 0.5, Respectively, with σ = 0.2 and Different Values of ξ
2-13
2-14
Modeling Units Employing a DFIM in PSSE
Section
3
Modeling Units Employing a DFIM in
PSSE
3.1 The Detailed Model
On the basis of the theory described in Section 2, the PSS®E dynamic simulation model
DFIGDC was developed to represent the double cage wound rotor induction machine.
Figure 3-1 shows the model data sheet for the DFIGDC model.
Since the DFIGDC model includes the stator flux linkage dynamics, it fairly well
represents the fundamental frequency or rotor slip frequency components and
exponential components triggered by faults. The rotor controls are modeled by using a
simplified, generic representation.
The characteristics of the model are demonstrated by its response to a series of tests.
A 2-MW 50-Hz DFIM was modeled in these tests.
The power versus rotor speed characteristic of the DFIM used in these tests is shown in
Figure 3-2. As seen in this figure, at rated power, the rated speed is 20% above
synchronous speed. The DFIM was dispatched at 57% of its rated power for the tests.
Therefore, the respective rotor speed was 6.67% above the synchronous speed. To
demonstrate the characteristics of the machine (as compared to the characteristics of
the controls), the rotor controls were disabled; that is, the rotor voltage d and q
components were kept constant.
The plots in Figure 3-3 depict the machine terminal voltage and rotor speed response to
a remote fault. It can be seen that voltage is maintained close to its initial value, and the
speed of the unit exhibits a well-damped dynamic response.
Stator and rotor currents are shown in Figure 3-4. Note the presence of a high-frequency
component as well as components at the same frequency that is seen in the plot of the
rotor speed, that is, the typical inter-machine rotor angle oscillation frequency.
Figure 3-5 shows the same stator and rotor currents zoomed in for the one-second
period starting from the fault inception. The higher frequency component now can be
clearly seen. The frequency of this component is related to the rotor slip (in this case
3-1
Siemens PTI Nonstandard Model Data Sheet

DFIGDC
DOUBLY-FED (WOUND ROTOR) INDUCTION MACHINE WITH THE POWER CONVERTER ROTOR
CONTROL
This model is located at system bus #_______IBUS
Machine #_______I
This model uses CONs starting with #_______J
and STATEs starting with #_______K
and VARs starting with #_______L
and ICONs starting with #_______M
CONs # Value Description STATEs # Description

CONs of the Main Program STATEs of the Main Program
J Ra, Stator resistance, pu K Psi_d, Stator Flux Linkage, pu
J+1 La, Stator Inductance, pu K+1 Psi_q, Stator Flux Linkage, pu
J+2 Lm, Mutual Inductance, pu K+2 Lambda_d_prim, rotor transient
J+3 R1, Rotor Resistance, pu flux linkage, pu
J+4 L1, Rotor Inductance, pu K+3 Lambda_q_prim, rotor transient
J+5 H, total drive train inertia, sec. flux linkage, pu
J+6 D, Damping Factor, pu K+4 Rotor Speed Deviation, pu
J+7 E1, Saturation Parameter, pu K+5 Rotor Angle deviation, degrees
J+8 S(E1), Saturation Parameter, STATEs of the Control Program
pu K+6 Q-axis PI
J+9 E2, Saturation Parameter, pu K+7 D-axis Integrator
J+10 S(E2), Saturation Parameter, K+8 D-axis PI
pu
CONs of the Control Program VARs # Description
J+11 KP2, P-gain of PI Control, pu VARs of the Main Program
J+12 KI2, I-gain of PI Control, pu L Ed, Rotor Voltage, pu on MBASE
J+13 KP3, Gain of Voltage Control, L+1 Eq, Rotor Voltage, pu on MBASE
pu L+2 Rotor Speed, pu
J+14 KI3, gain of Voltage Control L+3 Rotor Slip, pu
Integral, pu L+4 Id, Stator Current, pu on MBASE
J+15 IRMAX, Rotor Current Limit, L+5 Iq, Stator Current, pu on MBASE
pu L+6 RTR_P, Rotor Real Power, pu
J+16 VRMAX, Rotor Voltage Limit, on MBASE
pu L+7 RTR_Id, Rotor Current, pu on
J+17 SMAX, Total Converter Power MBASE
Limit, pu L+8 RTR_Iq, Rotor Current, pu on
CONs of the Power Curve Program MBASE
J+18 Init. Power Limit L+9 Initial Machine Internal Angle
J+19 Init. Speed Limit (rads)
J+20 to Coordinates of the Power vs L+10 Initial Slip, pu
J+49 Rotor Speed Characteristic L+11 Initial Mechanical Torque, pu
VARs of the Control Program
L+12 TSP Proportional Factor
ICONs # Description
ICONs of the Main program
M Memory
IBUS, ‘USRMDL’, I, ‘DFIGDC’, 1, 1, 1, 50, 9, 13, 0, CONs from (J) to (J+49) /
Notes:
1. The wind power to speed curve is approximated by 15 pairs of data points consisting of the rotor speed in p.u.
with respect to the synchronous speed and the total real power of the machine in p.u. of MBASE.
Figure 3-1 Model Data Sheet for the DFIGDC Model of the Doubly-Fed Induction Machine
3-2
Power vs Rotor Speed Characteristic
1.5
Power, p.u. on
Rated Power
1
0.5
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
-0.5
Rotor Speed, p.u. on Synchronous Speed
Figure 3-2 Steady-State Power versus Rotor Speed

Characteristic of the DFIM Used in the Tests
53 Hz = synchronous speed of 50-Hz frequency times 1.067, the overspeed determined

by the power versus rotor speed characteristic of Figure 3-2. Note that the higher
frequency component is also well damped.
The model allows simulation of a single cage machine by setting the rotor inductance L1
to a very high value, for example, 99999 per unit. It is noteworthy that, although possible,
it is not very likely that one would encounter a DFIM with two rotor windings.
Figure 3-3 Response of DFIM Terminal Voltage and Rotor Speed to a Remote Fault
3-3
Figure 3-4 Response of DFIM Stator and Rotor Currents to a Remote Fault
Figure 3-5 Response of DFIM Stator and Rotor Currents Zoomed in to Show Response
in the One-Second Period Following the Remote Fault
3-4
3.2 The Simplified Model

The full order model described above is not appropriate for stability studies because it
requires a significant reduction of the integration time step. Moreover, it is well known
that taking into account the stator flux linkage dynamics is not necessary for stability
studies. However, the rotor flux linkage dynamics must be represented in the electrical
machine model used for rotor angle and voltage stability studies.
The PSS®E dynamic model library contains several models of induction generators; for
example, the CIMTR3 and CIMTR4 models. These models include both the rotor flux
dynamics and the inertial dynamics, but not stator flux dynamics. The plots in Figure 3-6
show the dynamic response of the CIMTR3 and DFIGDC models to a system
disturbance. Three key variables are shown: machine terminal voltage (Vterm), active
power (Pelec), and reactive power (Qelec). This plot clearly demonstrates that the
CIMTR3 model provides good accuracy in representing the average component of these
variables, which is sufficient for rotor angle and voltage stability studies.
Figure 3-6 Comparison of the Response of the DFIGDC Model

and the Standard CIMTR3 Induction Motor Model
The vector control provided by a power converter connected to the rotor of the DFIM
processes commands in terms of d (R) and q (I) components of rotor current responsible
for controlling the machine voltage across the magnetizing branch and electromagnetic
torque. This control is done in a very fast manner by means of high-frequency PWM
electronics. The rotor current follows these commands with negligible delay. As currents
of the rotor and stator are closely related, the current injected by the DFIM into the grid
3-5
may be determined also based on the commands issued by these controls. This
fundamental consideration allows us to dramatically change the approach to modeling of
the DFIM controlled by a power converter connected to its rotor winding. The model
should have a fair representation of the controls whose outputs are applied to a
combined machine/converter model, but it does not need to model either the very fast
PWM electronics or include a detailed model of the machine.
This philosophy has been successfully employed for modeling DFIM wind turbines. It
can be illustrated with the control structure used in the generic WT3 wind turbine model. 2
It is worth mentioning that the implementation of this model in the two most widely used
commercial stability simulation software packages, PSS®E and Positive Sequence Load
Flow (PSLF), is identical.
Figure 3-7 shows the signal flow diagram of the DFIM-based wind turbine model and
demonstrates how the modules of the different elements interact with each other. The
aerodynamic mechanical power Pmech is calculated based on the turbine blade’s pitch
angle. The wind turbine model calculates the DFIM rotor speed deviation by using the
difference between the prime mover Pmech and the active power Pgen returned by the
generator/converter module. The generator/converter module gets inputs from the
converter control module in the form of an active current command and the voltage
across the magnetizing branch command.
Figure 3-7 Signal Flow Diagram of the DFIM-Based Wind Turbine Model
The active power (electromagnetic torque) control block diagram is shown in Figure 3-8.
The DFIM rotor speed is an input to this model. The rotor speed reference is provided by
a lookup table f(Pgen) with active power as the input. This lookup table actually
2
Ellis, A., E. Muljadi, J. Sanchez-Gasca, and Y. Kazachkov, “Generic Models for Simulation of Wind
Power Plants in Bulk System Planning Studies,” IEEE paper 2011GM1462, IEEE PES Annual
Meeting, Detroit, MI, July 2011.
3-6
reproduces the so called “Power-Rotor speed” characteristic that is unique for each type
of unit. The rotor speed error is processed through a PI controller and a first order delay
that has the power order as output. This output divided by the machine terminal voltage
produces the active current command, Ip cmd.
ω
(shaft speed)
Anti-windup on
Power Limits Pmax & dPmax/dt
Ipmax
Pgen ωref +
ωerr Pord
1 1 . Ip cmd
f ( Pgen ) Σ Kptrq+ Kitrq / s X
1+ sTpc .
1 + Tsps To
Generator /
Converter
Pmin & -dPmax/dt Model
To Pitch
Control Model To Pitch Vterm
Control Model
Figure 3-8 Active Power (Electromagnetic Torque) Control Block Diagram
The reactive power control block diagram is shown in Figure 3-9. The reactive power
command Qcmd may come from an external model that may be responsible for the
remote bus voltage control or power factor control. The reactive power error, the
mismatch between the reactive power command and the machine generated reactive
power, is the input for the integral control whose output is the terminal voltage reference.
The terminal voltage error is integrated, and the output of the integral controller is the
internal voltage command Eq cmd.
Figure 3-9 Reactive Power Control Block Diagram
Both Ipcmd and Eqcmd from the uncoupled active and reactive power controls are input to
the generator/converter model whose simplified example is shown in Figure 3-10. Two
first-order lags with small time constants simulate the electromagnetic dynamics of the
machine. This model also includes the mechanical dynamics of the rotor.
3-7
Eq" IYinj Isorc

Eq"cmd 1 -1
(efd) 1+ 0.02s X"
From s0
exwtge
T
IPcmd IP Low Voltage IXinj

1
Active Current
(ladifd) 1+ 0.02s
Regulation
From s1
exwtge
Vterm /θ
jX"
Figure 3-10 Generator/Converter Model
3-8
Modeling Adjustable Speed Drive Hydro Pumped Storage Units Employing a DFIM
Section
4
Modeling Adjustable Speed Drive Hydro
Pumped Storage Units Employing a
DFIM
The modeling of wind turbines employing a DFIM was described in Section 3 to explain
the pertinent mathematics and the lessons learned in the long evolution of these models
into their presently well-accepted versions. While there are obviously differences
between the modeling requirements of those wind units and of PSH units employing a
DFIM, there are also many similarities. For example, the required level of modeling of
the machine was thoroughly investigated for the wind turbines, and this knowledge can
be applied to the modeling of these pumped storage units. General control strategies
can also be adapted. For example, the wind turbine’s aerodynamic mechanical power
Pmech is calculated based on the turbine blade’s pitch angle; this can be thought of as
analogous to the gate position control used in a hydro application. Also similar to the
approaches taken in the wind application, some control functionalities cannot be taken
into account in the positive sequence fundamental frequency phasor approach typical for
the stability studies; thus, these functionalities also will be ignored in the PSH modeling.
An example of this is the well-known crowbar control used for protecting the converter
equipment.
The recommended model structure for a PSH unit employing a DFIM is given in this
section. The design of this model draws heavily on the control structures described in a
paper by T. Kuwabara, A. Shibuya, and H. Furuta titled “Design and Dynamic Response
Characteristics of 400 MW Adjustable Speed Pumped Storage Unit for Ohkawachi
Power Station” (IEEE Transactions on Energy Conversion, Vol. 11, No. 2, June 1996).
We found the approach to modeling—which is based on the experience of these
Japanese authors that was gained at their Ohkawachi PHS plant — to be the most
practical and useful of the many references reviewed. It is also consistent with our
understanding and the modeling practices of commercial stability programs. However,
there are many other references and ongoing development efforts, thus the selection of
this control structure is not to be construed as indicating that other potential control
structures are impossible, since they may have advantages for particular installations or
technologies.
The control structure selected reflects a thorough review of the literature. Two primary
approaches to control are described in the reviewed literature, based on the two primary
variables to be controlled.
4-1
In a conventional pumped storage plant, synchronous machines are employed. The

input to the governor controls is speed, and the gate position is controlled to adjust
power. The machine speed is not controlled; since the machine is a synchronous
machine, speed is locked to the system frequency.
In a pumped storage plant employing a DFIM, there are two controllable variables: gate
position and speed, as the speed of the machine is no longer locked to system
frequency. In the steady state, the controls select the optimum relationship between gate
position and speed to get the desired power. However, there are three basic control
approaches:
1. The electrical power is controlled by the power converter, and the rotating speed
is controlled by the turbine governor adjusting the gate position.
2. The rotating speed is controlled by the power converter, and the electrical power
is controlled by the turbine governor adjusting gate position.
3. A combination of these two approaches.
Since the power converter can be adjusted very quickly (tenths of a second) compared
to gate position (seconds), control strategy 1 above is labeled as Fast Power Control
and control strategy 2 as Fast Speed Control. In a general sense, it can be shown that
both approaches work, although the transient response of the units to events occurring
on the system is very different.
The control investigations that are related to the design of the Okhawachi power plant in
Japan showed the response employing both control strategies. 3 The authors show that
Fast Power Control is a superior approach compared to Fast Speed Control. They
conclude that a DFIM pumped storage unit with Fast Power Control would exhibit a
transient behavior superior to a unit with a synchronous machine, while a DFIM unit with
Fast Speed Control would exhibit a transient behavior inferior to a unit with a
synchronous machine.
The two control strategies are also compared in another reference. 4 The authors
conclude that the Fast Speed Control strategy is detrimental to grid stability. This is
illustrated by the simulation of an event calling for a reduction in power (in this case, a
step reduction in the active power set point). It is shown that the speed is quickly
controlled to the new desired setting, but the power injected by the unit into the grid
actually goes in the opposite direction and increases initially and decreases only after
several seconds. The authors state that the transient behavior resulting from the
variation is not acceptable. They then show that the Fast Power Control strategy is
capable of following changes of the power set point very quickly, about 84 times quicker
than using the alternate strategy, and it results in very acceptable transient behavior.
3
Nagura, O., and M. Yoshida, “Transient Behavior Analysis of Adjustable Speed Pumped Storage
System,” Hydro Vision International 2011, Session# 414; Sacramento, CA; July 20, 2011.
4
Pannatier, Y., et al., “Transient Behavior of Variable Speed Pumped-Turbine Units,” IAHR, 24th
Symposium on Hydraulic machines and Systems, October 2008.
4-2
The findings of the fifth reference 5 are based on the experience gained at the Goldisthal
pumped storage plant in Germany. Similar to the previous references, both control
strategies were investigated. The Fast Power Control strategy was shown to be superior
and was selected as the preferred strategy.
There are other PSH plants, such as the Yagisawa and Shiobara plants in Japan, where
the third control strategy is employed, using a combination of control strategy 1 (Fast
Power Control) and strategy 2 (Fast Speed Control). 6 The output power of these units is
normally controlled by adjusting the guide vane opening, and the converter controls
adjust to maintain the optimal rotating speed. This is essentially strategy 2. However, in
the case of a large disturbance casing a significant frequency deviation, the controls
respond very quickly due to an additional compensation signal applied to both the
converter control and the gate controls. Thus for large disturbances, the response of
strategy 3 will be similar to strategy 1, because both of these control strategies adjust
the electrical output not only by mechanical input (gate control) but also by using the
inertia of the rotating mass (through power control by the converter). The stated
advantage of this approach is that it meets both electrical and mechanical requirements
in that it reduces the movement of the guide vane and the adjustment of the rotating
speed of the units, while still maintaining the ability to make a quick response under
conditions where it is needed.
In the modeling effort described below, the Fast Power Control strategy is employed.
4.1 Turbine Mode of Operation

Figure 4-1 shows the control diagram suggested for the turbine (generating) mode of
operation. This figure shows the general control scheme and the connection between
the many components of the control structure. Each of these components will be further
explained below.
5
Kopf, E., et al., “Optimized Control Strategies for Variable Speed Machines,” 22nd IAHR Symposium
on Hydraulic Machinery and Systems, June 2004.
6
Communications with Teruyuki Ishizuki of Toshiba Corporation, Japan.
4-3
Threshold Setter for

Pulling-Back Control
Rotating Speed Speed

Pulling-Back Control Detector
Auto Voltage
Regulator Auto
Current
+ + Regulator Power
Auto Power Power Converter
+ - Regulator Converter To Generator
Motor
Actual Power From Generator
Feedback Motor
F.G. for - Speed

Optimum Speed + Governor
+ To Wicket
F. G. for Feed-Forward Hydraulic
Wicket Gate Signal Power Amplifiers Gates of
+ Pump Turbine
+ + Frequency- - Frequency Change
Pset
Responsive Governor of Transmission Line
Power -
Command
Signal Frequency Regulation
Figure 4-1 Control Scheme of Adjustable Speed Machine in Generator Mode
The converter control is responsible for controlling active power and voltage.
The power command is combined with the command from the frequency control as
shown in Figure 4-2. The total power command is Pset.
Bus 1 1+sTn
Frequency Deadband
1+sTp 1+Tnp
-
+ 1 +
1+sTr
Fref
1+sTff rsTr Pset
+
-
R Power
Command
Figure 4-2 Frequency Responsive Governor Control
The total power command Pset is summed up with the “Rotating Speed Pulling-back
Control.” The purpose of the latter is to make sure that the machine’s speed never goes
beyond the predetermined lower and upper limits that are determined by the size of the
converter.
The power error is processed by a PI regulator. Then the active current command, along
with the reactive current command from the voltage regulator, is fed to the power
converter control.
4-4
The turbine governor is controlling the machine shaft speed. The total power command
Pset enters the “Speed Optimizer” whose output is the speed reference. The speed error
is the input for the hydro turbine speed governor.
Figure 4-3 shows the signal flow diagram of the model for turbine operation.
Vreg bus
Vterm
IP (P)
Command
Converter Eq(Q) Generator/
Control Command Converter
Model Model PGen, QGen
PGen, QGen
Shaft Speed
Pmech
Power + Turbine-Governor
Command
Pset
Flow
+
Speed and Gate Speed and Gate
Frequency Bus Optimizer References
Governor Frequency
Figure 4-3 Connectivity Diagram for the Turbine Mode of Operation
Specific controllers are shown below. Figure 4-4 shows the active power controller that
supplies the power reference to the generator/converter module.
Over/under-speed
limitation
(Shaft Speed)
Anti-Windup on
Power Limits Pmax &dPmax /dt
Pord Ipmax
+ Ipcmd
Pset ∑ Kptrq +Kitrq1/s 1
÷
- 1+sTpo
1
Pgen Pmin & − dPmax /dt
1+sT
Vterm
Figure 4-4 Active Power Controller for the Turbine Mode of Operation
Figure 4-5 shows the model for the governor, turbine, and penstock dynamics. While the
model of the physics of the turbine and penstock are similar to that of conventional hydro
units, this model addresses the extra degree of freedom in control of the adjustable
speed unit. The mechanical power can be adjusted by either a change in turbine speed
or in a change in gate position or in a combination of both. The speed and gate
optimizers select the proper coordination of speed and gate position to maximize
efficiency. Note that the optimization process itself is not modeled; however, the optimal
4-5
mechanical power/gate/speed relationships supplied by the manufacturer are

represented through simplified functional characteristics. The model also includes
provision for simulating hydraulic coupling effects for units sharing the same penstock.
Ho Pset
Gate Reference
Speed Reference Speed Optimizer
Speed and
Gate Optimizer
KGR
1+sTGR
Gmax2
Pilot and
rgate K1gov Distribution
s Valves
Vop Gmax1
Gmin2
+ - + +
+ Kg 1
∑ ∑ Kpgov ∑ ∑
+ + 1+sTp s
- + -
Speed Kpgov Vol Gmin1
Deadband
1+sTdgov
Governor Gate Position
Speed
Dturb
q1 X
h1
-
Δh1
Pg
÷ X
-
∑ ∑ X At
+ ∑ Trate Pmech
h1 +
+ q1 -
Ho qnl
Δh2 1 q2
TW s
Δh3 q3
Δh4 q4
Figure 4-5 Governor, Turbine, and Penstock Dynamics Model for

the Generating Mode
Figure 4-6 shows the generator/converter model. The power reference from the active
power controller and the voltage reference from the reactive power controller (described
below) are used to determine the desired current command; and thus, the desired
generator active and reactive power. The generator/converter model also includes the
rotor inertial dynamics as shown in Figure 4-6.
4-6
E -
1 q -1 I cmd
Eqcmd
1+0.02s X"
1 I
Ipcmd P
1+0.02s
Coordinate
Conversion
Figure 4-6 The Generator/Converter Model
T mech + 1 Rotor Slip

∑
2Hs
-
T gen
Figure 4-7 Inertial Dynamics of the Generator/Converter Model
Figure 4-8 shows the reactive power controls. The voltage of the regulated bus is
controlled to the voltage reference by the PI controller. The PI controller determines a
voltage reference which is passed to the generator/converter controls.
Vreg ref
Kiv /s Eqmax
+ + To
Vreg bus 1 - Kpv Generator/
1+sTr 1+sTv Converter
+ Eqcmd Model
Kpv
Eqmin
1+sTv
Figure 4-8 Reactive Power Controller for the Turbine Mode of Operation
4.2 Pumping Mode of Operation

The following control diagrams are suggested for modeling the pumping mode of
operation. The control scheme is similar to that used for the generating mode but has
some significant differences. Figure 4-9 shows the general control strategy. Figure 4-10
shows the control structure and the linkages between its many components. The
components that are different from those described above for the generating mode will
be further explained below.
4-7
Speed
Detector
-
F. G. for Speed Governor
Optmimum Speed + (PID Type)
Auto Voltage
Regulator Auto
Current
+ + Regulator
Auto Power Power Converter
+ - Regulator To Generator
Motor
Actual Power From Generator
Feedback Motor
F. G. for Optimum Hydraulic To Wicket

Gate Operation Power Amplifiers Gates of
Pump Turbine
Pset
+ + Frequency- + Frequency Change
Responsive Governor of Transmission Line
Power -
Command
Signal Frequency Regulation
Figure 4-9 Control Scheme of Adjustable Speed Machine in Pump Mode
As in the generating mode, the converter control is responsible for controlling both active
power and voltage. The major differences from the generating model are in the active
power controller and the modeling of the physics of the gate, pump, and penstock
relationships.
Figure 4-11 shows the active power controller in the pump mode. The desired speed at
power Pset is determined from the speed optimizer and compared to the actual speed. A
PI controller determines the desired power, which is then compared to the power
absorbed by the unit, and then another PI controller determines the active current
command. The active current command, along with the reactive current command from
the voltage regulator, is fed to the power converter control.
The total power command Pset is also used by the gate optimizer to determine the
desired gate position. Figure 4-12 shows the model of the gate, pump, and penstock
dynamics.
The generator/converter controls and the reactive power controls are identical to those
documented above for the turbine mode of operation.
4-8
Vreg bus
Vterm
IP (P)
Command
Converter Eq(Q) Machine/
Control Command Converter
Model Model PGen, QGen
PGen, QGen
Speed_ref Pmech
Shaft Speed
Pump Model
Pset Head Flow

+ Gate_ref
+ Speed and Gate
Frequency Bus Optimizer
Governor Frequency
Figure 4-10 Connectivity Diagram of the Model for the Pump Operation
Pgen
1
1+sTp Pmax &dPmax /dt
Anti-Windup on
(Shaft Speed) Power Limits Pard Ipmax
Ipcmd
Speed and ω +
- Kp2 +Ki2/s
1
÷ To
ωerr + 1+sTpo
Pset ref ∑ Kp1+Ki1/s Machine/
Gate
- Converter
Optimizer + Pmin & − dPmax /dt
Vterm Model
Figure 4-11 Active Power Controller for the Pump Mode of Operation
4-9
Ho
Vop Gmax1
Gate
Speed and +
Pset Kg 1
Gate Deadband
Optimizer 1+sTp s
-
Vol Gmin1
Speed
Dturb
q1 X
h1
Δh1
Pg
÷ X
-
∑ ∑ X At ∑ Trate Pmech
h1 +
+ q1 -
Ho qnl
Δh2 1 q2
TW s
Δh3 q3
Δh4 q4
Figure 4-12 Model for Gate, Pump, and Penstock
4.3 Rotor Speed/Gate Position Relationships

The basic physics of the pump/turbine and penstock for an adjustable speed machine
are similar to the physics of conventional PSH units. However, a conventional unit
operates at synchronous speed; and thus, the only controllable quantity is gate position.
In the generating mode, the power output of the unit is controlled by adjusting the gate
position. Of course, throttling the flow through the closing of the gate has an impact on
efficiency. In the pumping mode, the impact on efficiency would be even greater, as
closing the gate would result in an increase of the effort required to pump a given
amount of water. Thus, conventional units generally operate at a wide-open gate in the
pumping mode at a fixed pump load.
With an adjustable speed pumped storage unit, both speed and gate position can be
controlled. Of course, this results in an additional need to determine the optimal settings
for these two interrelated quantities. Manufacturers have detailed models and computer
programs that calculate the efficiency in generating or pumping mode. These
calculations result in a “hill” diagram that shows efficiency for a range of gate, speed,
and flow (or gate, speed, and power) at a given static head. These diagrams can be
used to determine the gate and speed which give the best efficiency for a given static
head and flow rate (or power at a given static head).
The model does not need to contain these complex optimization calculations. The speed
and gate optimizers of the model will be used to select the proper coordination of speed
and gate position to maximize efficiency. While the optimization process itself is not
4-10
modeled, the optimal mechanical power/gate/speed relationships supplied by the

manufacturer are represented through simplified functional characteristics.
Some publications suggest examples of optimal characteristics obtained as a result of

calculations or tests. Harbort et al. (1998) shows a performance graph that provides the
optimal speed as a function of the MW output and the water head. A possible range of
rotor speed determined by the size of the power converter is marked as ±15%. The
performance graph shows that for a given power, the optimal speed is almost indifferent
to the water head for the three values of static head shown (a variation of more than
10%).
Based on the performance graph for optimal turbine speed, it is also possible to obtain
optimal gate opening function (Harbort et al, 1998). This function shows the gate
opening’s dependence on power and water head for the optimal speed.
Note that the model of the PSH unit described above will not be project specific. A big
challenge of this project is to develop these functions as generic lookup tables that can
be used for simulating the performance of any commercially available hydro unit.
4-11
4-12
Bibliography
Section
5
Bibliography
Numerous publications are available on different aspects of variable speed hydro pumped
storage units. The most relevant ones for the purpose of this study are provided here.
1. Kuwabara, T., A. Shibuya, and H. Furuta, “Design and Dynamic Response

Characteristics of 400 MW Adjustable Speed Pumped Storage Unit for Ohkawachi
Power Station,” IEEE Transactions on Energy Conversion, Vol. 11, No. 2, June 1996.
2. Kita, E., A. Bando, and T. Kuwabara, “400 MW Adjustable Speed Pumped Storage
Hydraulic Power Plant,” Hitachi Review, Vol. 44, No. 1, 1995.
3. Harbort, T., G. Lein, and E. Goede, “Power Frequency Control of a Pump Turbine as
an Example for the Operation with Adjustable Speed of Hydraulic Machines,”
University of Stuttgart, 1998.
4. Pannatier, Y., C. Nicolet, B. Kawkabani, J.-J. Simond, and P. Allenbach, “Dynamic

Behavior of a Two Variable Speed Pump-turbine Power Plant,” Proceedings of the
18th International Conference on Electrical Machines, 2008.
5. Fujii, T., K. Mine, and O. Nagura, “Contribution of Adjustable Speed Pumped Storage
System for Electrical Grid Stability,” Presented at Hydro 2010, Session 23 – Pumped
Storage, Role in the Grid; September 29, 2010; International Journal on Hydropower
& Dams, Wallington, Surry, UK SM6 6AN, 2010.
6. Are Suul, Jon, “Variable Speed Pumped Storage Hydropower Plants for Integration of
Wind Power in Isolated Power Systems,” Norwegian University of Science and
technology, 2009.
7. Gao, H.M., and C. Wang, “A Detailed Pumped Storage Station Model for Power
System Analysis,” 1-4244-0493-2/06 IEEE, 2006.
8. Kopf, E., S. Brausewetter, M. Giese, and F. Moser, “Optimized Control Strategies for
Variable Speed Machines,” 22nd IAHR Symposium on Hydraulic Machinery and
Systems, June 29–July 2, Stockholm, 2004.
9. Grotenburg, K., F. Koch, I. Erlich, and U. Bachmann, “Modeling and Dynamic

Simulation of Variable Speed Pump Storage Units Incorporated into the German
Electric Power System,” Graz, EPE, 2001.
5-1
Bibliography
10. Erlich, I., and U. Bachmann, “Dynamic Behavior of Variable Speed Pump Storage
Units in the German Electric Power System,” 15th Triennial World Congress,
Barcelona, Spain, 2002.
11. Liang, J., and R.G. Harley, “Pumped Storage Hydro-plant Models for System
Transient and Long–term Dynamic Studies,” 978-1-4244-8357-0/10, IEEE, ???.
12. Kerkman, R.J., T.A. Lipo, W.G. Newman, and J.E. Thirkell, “An Inquiry into Adjustable
Speed Operation of a Pumped Hydro Plant,” Parts 1 and 2, IEEE Transactions on
Power Apparatus and Systems, Vol. PAS-99, No. 5, Sept./Oct., 1980.
13. Nagura, O., and M. Yoshida, “Transient Behavior Analysis of Adjustable Speed
Pumped Storage System,” Hydro Vision International 2011, Session# 414;
Sacramento, CA, July 20, 2011.
14. Akagi, H., “Large Scale Converters and Rectifier/inverter Based Systems,”
Proceedings of the IEEE, Vol. 89, No. 6, June 2001.
15. Turel, V., F. Prime, and T. Machino, “Implementation of a Variable Speed Unit in the
AVCE PSPP,” Hydro 2008 Conference, Ljubljana, Slovenia, October 6–8, 2008.
16. Donalek, P., H. Clark, R. Nakata, and J. Stein, “Application of Adjustable Speed
Doubly Fed Machines in Pumped Storage and Conventional Hydroelectric Plants,”
Proceedings of the American Power Conference, Chicago, IL, Vol. 55-I, pp. 245–290,
April 1993.
17. Donalek P., and H. Clark, Application of Adjustable Speed Machines in Conventional
and Pumped Storage Hydro Projects, EPRI, Palo Alto, CA 94304, November 1995.
18. Pannatier, Y., C. Nicolet, B. Kawkabani, J.-L. Deniau, A. Schwery, F. Avellan,

and J.-J. Simond, “Transient Behavior of Variable Speed Pumped-Turbine Units,”
24th IAHR Symposium on Hydraulic Machinery and Systems, October 27–31,
Foz Do Iguassu, 2008.
19. Hodder, A., “Double-fed Asynchronous Motor-generator Equipped with a 3-level

VSI Cascade,” Ph.D. Thesis TH2939, École Polytechnique Federale de
Lausanne (EPFL), 2004.
20. Kopf, E., S. Brausewetter, M. Giese, and F. Moser, “Optimized Control Strategies
for Variable Speed Machines,” 22nd IAHR Symposium on Hydraulic Machinery
and Systems, Stockholm, Sweden, June 29–July 2, 2004.
21. Nicolet, C., Y. Vaillant, B. Kawkabani, P. Allenbach, J.-J. Simond, and F. Avellan,
“Pumped Storage Units to Stabilize Mixed Islanded Power Network—Transient
Analysis,” Hydro 2008 Conference, Ljubljana, Slovenia, October 6–8, 2008.
5-2
Decision and Information Sciences
Argonne National Laboratory
9700 South Cass Avenue, Bldg. 221
Argonne, IL 60439-4844
www.anl.gov
Argonne National Laboratory is a U.S. Department of Energy

laboratory managed by UChicago Argonne, LLC

ANL DIS-13 06 Modeling AS PSH PDF

Uploaded by

Copyright:

Available Formats

ANL DIS-13 06 Modeling AS PSH PDF

Uploaded by

Document Information

Original Title

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

ANL DIS-13 06 Modeling AS PSH PDF

Uploaded by

Copyright:

Available Formats

ANL/DIS-13/06

Modeling Adjustable Speed Pumped

Decision and Information Sciences

Availability of This Report

Modeling Adjustable Speed Pumped

James Feltes, Yuriy Kazachkov and Bo Gong

Bruno Trouille and Peter Donalek

The development of vendor-neutral models was carried out by the Advanced

Section 1 Introduction ........................................................................................... 1-1

Section 2 Theory of the Dynamic Simulation Model of the Doubly-Fed

Section 3 Modeling Units Employing a DFIM in PSSE.................................... 3-1

Section 4 Modeling Adjustable Speed Drive Hydro Pumped Storage

Section 5 Bibliography .......................................................................................... 5-1

Section 4 provides a discussion on modeling of adjustable speed hydro pumped storage

Section 5 provides a bibliography.

Figure 2-1 Configuration of a DFIM

s = (ω1 - ωm)/ ω1,

The stator voltage equations are:

The rotor voltage equations are:

δ = sωt electrical degrees

Stator Flux Linkage :

and the stator flux linkages can then be expressed as:

and transient (derived) inductance:

Ψ = λ' − IL' ( D − 12)

The approach to saturation representation is similar to that used in synchronous

R and I components of the saturation factor will be calculated in proportion to the

The differential equations above should be supplemented by equations describing the

Rotor slip with respect to the base frequency:

The rotor angle is equal to:

I1 Doubly-Fed (Wound Rotor) Induction Generator Block Diagram

ψ stator flux linkage ω o = 2πf

Figure 2-3 Block Diagram for the Single Cage DFIM

2.1 Steady-State Theory

These equations can be rewritten by introducing the following phasors or quantities:

Vs = vds + jvqs

Ir = idr + jiqr

Vs = − Rs Is + jϖ s λs

λr = − Lm Is + Lrr Ir

Vs = − Rs Is + j ( − X ss Is + X m Ir )

Vs = − Rs Is − jX 1 Is − jX m ( Is − Ir )

Vs = − Rs Is − jX 1 Is − jX m ( Is − Ir )

Figure 2-4 Equivalent Circuit of a Wound Rotor Induction Machine

In the last two diagrams, V1 = Vs, and V2 = Vr.

Assuming Rs = 0, the complex stator and rotor currents can be derived as

and electrical torque as

Tel = Re(Vs Is ) .

Assuming the stator voltage phasor as reference, then

Then electrical torque can be found as

where α = atan(ξσ) is a function of the slip s.

Angular Characteristics of Electrical Torque

Elec. Torque, p.u.

This page intentionally left blank.

Siemens PTI Nonstandard Model Data Sheet

CONs # Value Description STATEs # Description

IBUS, ‘USRMDL’, I, ‘DFIGDC’, 1, 1, 1, 50, 9, 13, 0, CONs from (J) to (J+49) /

Power vs Rotor Speed Characteristic

Figure 3-2 Steady-State Power versus Rotor Speed

53 Hz = synchronous speed of 50-Hz frequency times 1.067, the overspeed determined

3.2 The Simplified Model