Transient Stability Analysis of the IEEE 14-Bus Test System Using Dynamic

Computation for Power Systems (DCPS)

N. Hashim, N. Hamzah, M.F. Abdul Latip A.A. Sallehhudin

Faculty of Electrical Engineering Advanced Power Solutions Sdn. Bhd.
UniversitiTeknologi MARA (UiTM) 40675 Shah Alam, Selangor, Malaysia
40450 Shah Alam, Selangor, Malaysia [email protected]
[email protected]

Abstract Transient Stability Analysis (TSA) is a major The aim of this investigation is to analyze the transient
analysis in the operation of power systems, due to the stability of the system by analyzing the characteristics of
increasing stress on power system networks. One of the main the machine states, including machine speed, rotor angle,
goals of this analysis is to gather critical information, such as output electrical power and terminal voltage with respect
critical clearing time (tCCT,) of the circuit breakers for faults to fault clearing time after the three-phase fault occurs in
in the system. tCCT is defined as the maximum time between the system. Section II provides a brief overview of the
the fault initiation and its clearing, such that the power Dynamic Computation for Power Systems (DCPS)
system is transiently stable. This paper presents a transient software package. Section III describes the Differential
stability analysis of the IEEE 14 bus test system using
Algebraic Equations (DAEs) of the power system
Dynamic Computation for Power Systems (DCPS) software
package. This C++ based software package has the ability to
dynamic device, including the synchronous machine,
handle systems up to 1000 buses and 250 generators, turbine-governor and exciter. Section IV describes an
providing an alternative to expensive commercial software overview of transient stability analysis. The results of the
packages. To analyze the effect of the distance of the fault case study using the IEEE 14-bus test system are
location and critical clearing time on the system stability, a presented in section V. Section VI concludes the paper.
three-phase fault has been applied at five different locations
in the system. The stability of the system has been observed II. DYNAMIC COMPUTATION FOR POWER SYSTEMS
based on the simulation graphs of terminal voltage, (DCPS)
machines rotor angle, machines speed and output electrical The Dynamic Computation for Power Systems
power. The simulation results showed that tCCT decreases as program was created by the late of Dr Sallehhudin Yusof,
the fault location becomes closer to the main generator. the former managing director of Advanced Power
Solutions Sdn Bhd, for the purpose of research. There are
KeywordsTransient Stability Analysis; DCPS; Improved
three simulation programs embedded in DCPS: (i) LF for
Euler Method; Critical Clearing Time
load flow, (ii) TS for transient stability and (iii) CG for
I. INTRODUCTION coherency grouping. As a pre-requisite to transient
stability analysis, the load flow program must be run to
The two major areas in stability studies are steady- provide the steady-state operation points. There are three
state stability and transient stability. Steady-state stability load flow calculation methods employed in DCPS,
refers to the ability of the electrical power system to including (i) Gauss- Seidel, (ii) Decoupled Newton, and
regain synchronism after encountering slow and small (iii) Fast Decoupled. Each method has its own approach to
disturbances, such as gradual power. Transient stability solving the non-linear algebraic equations. Although each
analysis of a power system refers to the system's ability to method points to equal objectives for load flow solution,
remain in synchronism when subjected to a large the answers obtained from each method is likely to be
disturbance, such as a three-phase fault and the sudden different. The speed of convergence of the three methods
outage of a transmission line or the sudden addition or is extremely important in achieving cost effective
removal of the loads [1, 2]. Studies regarding these types simulations. From experience in using the DCPS, the
of stability are helpful in determining crucial metrics such Decoupled Newton method gives the most satisfactory
as critical clearing time (tCCT) of the circuit breakers and results [7].
the voltage level of the power system. In the study of
transient stability, the critical clearing time is one III. THE IMPROVED EULER METHOD FOR SOLVING
significant factor for maintaining the transient stability of DIFFERENTIAL ALGEBRAIC EQUATIONS (DAES)
the power system [3]. The definition of tCCT is related to
transient stability. tCCT is defined as the maximum time An electrical power system consists of multiple
between the fault initiation and its clearing, such that the individual dynamic devices connected together to form a
power system is transiently stable. The clearing time is the large and complex dynamic system. The three most
time duration from the instant the disturbance occurred important dynamic devices are synchronous machine,
until all three poles of the circuit breaker are completely exciter and turbine-governor. The relationship between
open. The critical clearing time is obtained by increasing these devices is illustrated in Figure 1.
the fault time interval until the system loses its stability.
Increasing the tCCT can reduce the protection system rating
and its cost and can increase the reliability of the
protection system [4].
 Figure 1. Relationship Between Machine States & Controller States

Before TSA can be performed, the physical model of

these devices needs to be formulated into Differential
Algebraic Equations (DAEs). First-order DAEs are in the
form of:

dy/dt f y, t (1)

where y(t) is a time-varying function subject to the initial Figure 2. Graphical Depiction of Improved Euler Method; (a) Predictor
condition: y(0) = y0. This initial value problem must be and (b) Corrector
solved numerically, since there is no closed solution. The
objective of solving (1) is to determine y0, y1, y2, ,yi at A. Synchronous Machine Model
time t0, t1, t2, ti. The solution must take into In DCPS, there are three synchronous machine
consideration these three aspects, including accuracy, models that can be used for transient stability analysis,
stability, and efficiency. The solution methods for DAEs including:
can be divided into implicit and explicit numerical a) gen0 a classical machine model.
integration techniques. Under the implicit technique, the
b) gen2 a transient level machine model.
Trapezoidal method is the most popular choice. Under the
c) gen5 a detailed machine model
explicit technique, both Runge-Kutta and Euler Methods
have been used extensively. In this paper, the Improved In this paper, only a detailed synchronous machine model
Euler Method, also known as Predictor-Corrector Method, (gen5) has been used in transient stability analysis. This
is chosen to represent the explicit numerical integration model, also known as round rotor machine model, is
technique that was developed in DCPS. The Improved shown in Figure 3. The gen5 model presumes an electric
Euler Method is depicted graphically in Figure 2. The transmission network with a positive sequence source
Improved Euler Method is a Runge-Kutta based method voltage where instantaneous amplitude and phase are
for approximating the solution of the initial value problem known and current is to be determined. For clarity, the
[5]. In this method an initial estimate value is calculated magnetic saturation and the stator resistance are excluded
by using the following Predictor Formula: from the model. The stator flux is derived from stator
currents and the exciter voltage [8]. The parameters of this
yP y hf y , t (2) model are shown in Appendix section. The equations
which describe this model are:
The time-step, h is given by (t = ti+1 - ti), which can be
constant or variable. Using the result in (2), the final d/dt 2f (4)
estimate value can be determined by applying the d/dt PMECH PE 2f D / M (5)
following Corrector Formula:
dE /dt EFD E X X I /T (6)
yC y 0.5h f y , t f yP , t (3)
dE /dt E X X I /T (7)
dE /dt E E X X I /T (8)

dE /dt E E X X I /T (9)

Where is the rotor angle; is the angular velocity;

PMECH is the input mechanical power; PE is the output
electrical power; D is the damping constant; M is the
machine inertia constant; f is the system frequency; X is
the direct-axis synchronous reactance; X is the direct- 1 1
axis transient reactance; X is the direct-axis sub-transient R
1 + sT1
reactance; T is the open circuit direct-axis transient time
constant; T is the open circuit direct-axis sub-transient
time constant; X is the quadrature-axis synchronous
reactance; X is the quadrature-axis transient reactance;
X is the quadrature-axis sub-transient reactance; T is Figure 4. Turbine-Governor Model (Type Gov10) in DCPS
the open circuit quadrature-axis transient time constant;
T is the open circuit quadrature-axis sub-transient time C. Exciter Model
constant; and X is the leakage reactance. The exciter is a key component of a synchronous
generator control system, as it maintains the output
X 'd' X terminal voltage at a constant level. It provides direct
X 'd X current to the field winding of the synchronous machine.
The amount of excitation required is a function of the
1 1 X 'd X 'd' +
E fd + + + d generator load. As the generator load increases, the
Td' 0S Td''0S X 'd X
amount of excitation increases. Figure 5 shows a simple
X 'd' X excitation system used in DCPS. The differential
X ad I fd
'
X X '' equations relating the input and output variables of this
d d
( X 'd X ) 2 exciter model is given as:

+ dVC /dt VT VC /TR (13)

X d X 'd + Id

X q X 'q Iq

X 'q X 'q'
( X 'q X q ) 2 1
X 'q' X
1 + sTR K1
' ''
1 1 X X
q q

Tq' 0S Tq''0S '

X X q
q

X 'q' X Figure 5. Exciter Model (Type Exc10) in DCPS

X 'q X
IV. TRANSIENT STABILITY ANALYSIS
Figure 3. Detailed Synchronous Machine Model (Type Gen5) in DCPS As a pre-requisite to TSA, load flow calculations need to
be solved to determine the steady-state conditions of the
B. Turbine-Governor Model network. Then, the values from these solutions will be
The function of a turbine-governor system is to used to calculate the initial conditions for dynamic models
monitor the speed of the the rotor being driven by the namely synchronous generators, turbine-governors system
turbine. The governor provides the appropriate signal to and exciters system. The purpose of transient stability
fuel valve controllers either to open or close the valve, analysis is to evaluate the ability of the power system to
depending on the deviation of the rotor speed with respect withstand large disturbances and to survive transition to a
to the synchronous speed [7]. Figure 4 shows a simple normal operating condition. These disturbances can
turbine-governor control system embedded in DCPS. The include a three-phase fault on a transmission line, loss of a
input mechanical power PMECH to the generator is generator or loss of a major load [2]. In the study of
controlled by this turbine-governor system in order to transient stability, one of the most significant factors for
maintain the output electrical power, PE at an acceptable maintaining the transient stability of the power system is
the critical clearing time (tCCT). There are many methods
level. A speed error signal Terr is calculated by comparing
to determine the tCCT, including the Lyapunov direct
the recorded speed at the shaft to a desired value. Terr is
methods, transient energy function methods, equal area
used to determine the new gate/valve position. The speed
criterion, artificial intelligence and conventional time
change as an input is given by = (0 - )/0. The
domain simulation. Time domain simulation can be
equations that describe the control transfer function
considered as the most accurate solution, and in theory it
relating the input and output variables of the Simple
can handle unlimited detailed models of generator, load
Turbine-Governor are given below:
and other system controllers. However, the solution of this
simulation is very complicated, because it involves
dPMECH /dt PMECH /T (10) multiple nonlinear differential-algebraic equations
(DAEs). In this paper, the Improved Euler Method, also
known as Predictor-Corrector Method, is chosen to
represent the explicit numerical integration technique that
was developed in DCPS. The overall process of
performing TSA in DCPS is shown in Figure 6.

Figure 8. IEEE 14-Bus Test System

V. RESULTS AND DISCUSSION

The TSA was carried out on IEEE 14-bus test system
as shown in Figure 8. The network data used for this work
Figure 6. Flowchart of Transient Stability Analysis in DCPS is obtained from [9]. Dynamic data for generators,
turbine-governors and exciters used in this work are given
in the Appendix. These data values were kept constant
In this simulation, a three-phase fault has been injected during the simulation. The main generator for this system
to the system at t = 1s and removed from the system at the is located at bus 1, and produces the largest real power of
clearing time, tCT. In practice, fault clearance occurs approximately 232.4 MW. To study the effect of fault
through the removal of the affected line or branch from distance from the main generator and critical clearing
the power system network by protective relay action and time, a three-phase fault has been applied at five
opening of circuit breakers. After fault clearance, if the difference buses namely bus 1(swing bus), bus 2, bus 3,
system is still stable, then the system terminal voltage is bus 4 and bus 5. The distance of each bus from the main
expected to recover. The generator's rotor angle would generator is described by the reactance value as shown in
settle back to its steady state condition. To determine tCCT, Table I.
tCT was increased gradually using a step time of 0.01s
until the system appears to be unstable as determined by Figure 9-12 shows the simulation graphs of terminal
observing terminal voltage as a reference point. The tCCT voltage, machines rotor angle, machines speed and
is obtained by calculating the midpoint between the fault output electric power in the case where a three-phase fault
time when the system starts to be unstable and the time occurs at bus 1 when the fault clearing time is less than
where the system was last known in a stable state. The critical clearing time (tCT < tCCT). As can be seen,
procedure for determining the tCCT when fault occurred at following the removal of fault from the system, the
bus 1 is shown in Figure 7 [7]. terminal voltage (Figure 9) is recovered and the machines
rotor angle (Figure 10) is settling back to its steady state
after experiencing damped transient oscillations. For this
particular operating condition, the system is said to be
transiently stable following a three-phase fault. When the
clearing time exceeds the critical clearing time (tCT >
tCCT), the machines rotor angle will go out of step, as can
be seen in Figure 14, and in this case the system is in
unstable condition. The system model is unable to sustain
the fault and loses its synchronism. Based on this analysis,
the critical clearing time for this case is 0.145 sec. The
same analysis has been done for other buses and the
summary of this analysis is shown in Table I. The results
Figure 7. Critical Clearing Time Determination in DCPS
in the table clearly show that tCCT decreases as the fault
location becomes closer to the main generator. The
distance of the fault location from main generator was
described by the values of reactance in p.u.
 Pre-fault Fault-on Post-fault Pre-fault Fault-on Post-fault

tct tct

Figure 9. Plots of terminal voltage for all buses (tct=0.14). (Stable) Figure 13. Plots of terminal voltage for all buses (tct=0.15). (Unstable)

Figure 10. Plots of machines rotor angle (tct=0.14). (Stable) Figure 14. Plots of machines rotor angle (tct=0.15). (Unstable)

Figure 11. Plots of machines speed (tct=0.14). (Stable) Figure 15. Plots of machines speed (tct=0.15). (Unstable)

Figure 12. Plots of machines output electric power (tct=0.14). (Stable) Figure 16. Plots of machines output electric power (tct=0.15). (Unstable)

TABLE I. SUMMARY OF ANALYSIS RESULTS

Rank Faulted Bus Shortest Reactance from Bus 1 Clearing Time (s) Analysis Result Critical Clearing Time (s)
0.14 Stable
1 1 (Swing Bus) - 0.145
0.15 Unstable
0.38 Stable
2 2 0.0592 p.u 0.385
0.39 Unstable
0.58 Stable
3 5 0.2230 p.u 0.585
0.59 Unstable
0.34 Stable
4 4 0.2355 p.u 0.595
0.35 Unstable
0.63 Stable
5 3 0.2572 p.u 0.635
0.64 Unstable
 VI. CONCLUSION [7] N. Hashim, N.R. Hamzah, P. Mohd Arsad, R. Baharom, N.F. Nik
Ismail, N. Aminudin, D. Johari and A.A. Sallehhudin "Modeling of
In this paper, the TSA of the IEEE 14-bus test system Power System Dynamic Devices Incorporated in Dynamic
has been successfully analyzed using DCPS software Computation for Power Systems (DCPS) for Transient Stability
Analysis," 2011 IEEE International Electric Machines & Drives
package. For the comparative analysis, five different Conference (IEMDC), pp. 647 - 652, May 15-18, 2011
locations of the three-phase fault was chosen to study the [8] A.M Mohamad, N. Hashim, N. Hamzah, N.F. Nik Ismail, M.F.
effect of the fault distance from the main generator and Abdul Latip "Transient stability analysis on Sarawak's Grid using
critical clearing time. As can be seen from Table I, the Power System Simulator for Engineering (PSS/E)," 2011 IEEE
fastest tCCT is 0.145s when the fault occurred exactly at the Symposium on Industrial Electronics and Applications (ISIEA),
main generator bus, while the longest tCCT is 0.635s when pp. 521 - 526, Sept. 25-28, 2011
the fault occurred on bus 3, located far from the main [9] R. Christie, UW Power System Test Case Archive, Available:
http://www.ee.washington.edu/research/pstca/
generator. The tCCT decreases as the fault location
becomes closer to the main generator. An interpretation of
this finding is that faults that occur closer to the main
generator must be cleared more quickly than faults that APPENDIX
occurred some distance away. To ensure the stability of
TABLE II. DYNAMIC DATA FOR TRANSIENT STABILITY ANALYSIS IN
power system due to fault occurrence, proper protection DCPS
system settings must be made. Information from this
analysis and similar analysis based on the configuration of Parameters Generator Turbine-Governor Exciter
the target network can be used in identifying the correct
M 5 - -
settings.
f 50 hz - -
ACKNOWLEDGMENT X 1.9 - -
The financial assistance of the Faculty of Electrical X 0.3 - -
Engineering, Universiti Teknologi MARA is greatly
X 0.3 - -
appreciated. Also, special thanks to Mr Ahmad Adib
4.57 - -
Sallehhudin for permission to use DCPS software. T
0.042 - -
T
REFERENCES
X 1.6 - -
[1] P. Kundur, Power system stability and control, New York: Mc-
Graw-Hill, 1994. X 0.7 - -
[2] P. Kundur, J. Paserba, V. Ajjarapu, G. Andersson, A. Bose, C. X 0.293 - -
Canizares, N. Hatziargyriou, D. Hill, A. Stankovic, C. Taylor, T.
Van Cutsem, and V. Vittal, "Definition and classification of power T 0.5 - -
system stability IEEE/CIGRE joint task force on stability terms and T 0.3 - -
definitions," IEEE Transactions on Power Systems, vol. 19, no. 3,
pp. 13871401, Aug. 2004. X 0.25 - -
[3] IEEE Committee Report, "Proposed terms and definitions for VMAX - 1.0 -
power system stability,"IEEE Transactions on Power Apparatus
and Systems, vol. PAS-101, pp. 18941898, 1982. VMIN - -1.0 -
[4] A. M. Hemeida, Improvement of voltage stability and critical T - 30.0 -
clearing time for multi-machine power systems using static var R - 0.05 -
compensator, ICGST-ACSE, vol. 9, no. 2, pp. 4147, December
2009. EFDMAX - - 1.0
[5] S.C Chapra and R.P Canale, Numerical methods for engineers with EFDMIN - - -1.0
software and programming applications, 4th. ed., McGraw Hill,
2002. K - - 30.0
[6] I. Xyngi, A. Ishchenko, M. Popov, and L. van der Sluis, "Transient TR - - 0.01
Stability Analysis of a Distribution Network With Distributed
Generators," IEEE Transactions on Power Systems, vol. 24, no. 2,
pp. 1102 - 1104, May 2009.