1524 IEEE Transactions on Power Systems, Vol. 9, No.

3, August 1994
Transient Stability Index from
Conventional Time Domain Simulation

C.K. Tang C.E. Graham M. El-Kady R.T.H. Alden

Member Member Senior Member Senior Member

Ontario Hydro McMaster University

ABSTRACT - This paper presents a practical and efficient under study, but also enable them t o speed up the limits
method t o incorporate energy analysis into conventional time derivation process by making use of stability margin
domain transient stability simulation for producing a first sensitivity information [2].
swing stability index. The proposed method is a variation of
the PEBS approach, utilizing a numerical Transient Energy The capability of producing a stability index has been one of
Function and a pseudo-fault-on trajectory starting from the the major incentives for developing the so-called direct
dominant potential energy peak point of the postfault system methods. The other incentive is the fast computational speed
trajectory. The transient stability index not only provides that these methods may achieve. Over the last t w o decades,
additional insight into the stability studies, but also has the significant progress has been made t o improve the practicality
potential of speeding up the stability limit derivation process of these methods for transient stability analysis of power
by reducing the number of program runs. systems with simplified models. These simplified models are
usually justified for online applications t o achieve fast
Keywords: Transient stability, transient energy margin, computational speed, at the expense of accuracy. Ontario
dynamic security assessment Hydro is currently involved in an EPRl project t o develop the
direct Transient Energy Function (TEF) method for online
1 .O INTRODUCTION transient stability calculation [3]. While the modelling
capability and computational efficiency of direct methods will
Transient stability analysis is an important aspect in the continue t o improve in the future, some of their techniques
planning and operation of the Ontario Hydro Bulk Electricity for computing stability indices can be easily incorporated into
System (BES). Each year engineers perform thousands of conventional time domain simulation.
transient stability simulations t o derive operating security
limits offline. These limits are entered into the Energy Several researchers attempted in the past t o incorporate
Management System (EMS) a t the system control center for transient energy analysis into the conventional time domain
online dynamic security monitoring. simulation. Fouad e t al developed the output analysis
technique [41 t o derive generator stress indices based on
The transient stability analysis tool used by system planners individual TEFs. In [51, Athay et al examined extensively the
and engineers a t Ontario Hydro is based on the step-by-step energy components along simulated system trajectories. In
time domain simulation method. This method has a superb 161. Maria e t al derived a stability index from time domain
power system modelling capability which is essential t o the simulation by performing line searches along linear angle
analysis of modern power systems equipped with complex paths on the potential energy surface associated w i t h
control aids and special protection schemes. In Ontario Hydro, classical system representation.
fast response and high gain static excitation systems [ l l are
employed t o enhance system transient stability, in T o incorporate stability index calculation into conventional
conjunction with special protection schemes such as time domain simulation, the method proposed in this paper
generation rejection and protective relays of various types. follows the basic approach of [61. The proposed method
These devices together with the appropriate load models have removes the classical model restriction in detecting the
t o be accurately represented in transient stability simulations. Potential Energy Boundary Surface (PEBSI crossing, and
replaces the line searches, which are less accurate and at
inspite of its superb modelling capability, the time domain times problematic, with a pseudo-fault-on trajectory starting
method has a shortcoming in that it only produces a yes-or-no from the potential energy peak point of the postfault system
answer. Consequently, stability limit derivation based on this trajectory. The stability index is calculated by evaluating the
tool is essentially a trial-and-error approach resulting in a large numerical TEF along the pseudo-fault-on trajectory. The
number of stability runs. If a stability index can be derived practicality of the proposed method has been successfully
from the time domain simulation, it will not only help demonstrated on t w o Ontario Hydro test systems w i t h
engineers t o gain additional insight into the stability problems detailed generator and excitation system models.

2.0 MATHEMATICAL FORMULATION

93 SM 407-9 PWRS A paper recommended and approved
by the IEEE Power System Engineering Committee of
the IEEE Power Engineering Society for presentation Given an n-generator system, the swing equations are:
at the IEEE/PES 1993 Summer Meeting, Vancouver, B.C., $,=Ui (2.1)
Canada, July 18-22, 1993. Manuscript submitted Aug. Mibi=Pmi-Pei (2.2)
27, 1992; made available for printing Uay 3, 1993.

PRINTED IN USA

0 1993 IEEE
0885-8950194/$04.00

Authorized licensed use limited to: ANDERSON SOSA. Downloaded on February 13,2025 at 17:48:54 UTC from IEEE Xplore. Restrictions apply.
where di, wi, P
,,, and Pni are the rotor angle, rotor speed,
mechanical power input, and electrical output of Machine i
respectively.

Equations (2.1 ) and ( 2 . 2 )together with other differential and

algebraic equations can be put into the general form:

(2.17)
where 3 and y are the state and algebraic variables N
. Nb
respectively. Equation (2.3) describes the dynamic models
while Eq. (2.4) describes the network and generator interface
Ma=CMi,Mb=CMi
1 1
equations. A conventional time domain simulation program
solves these t w o sets of equations through step-by-step
integration in the time domain. 3.0 TRANSIENT ENERGY MARGIN

The center of inertia (Col) frame of reference is defined as: Direct methods of transient stability analysis seek t o
characterize initial conditions of the postfault trajectories, i.e.,
system conditions a t fault clearing, that converge back t o the
postfault SEP. The set of such initial conditions comprises the
n region of attraction, or the stability region. Hence in direct
methods, the transient stability problem is one of determining
n whether the system conditions at fault clearing lie inside the
MT=CMi
i-1
stability region.

The characterization of the stability region boundary for

After transforming the rotor angles and speeds into the CO1 classical systems has been discussed in detail in [81. In
frame, the swing equations become: applications, it is impractical and unnecessary t o compute the
entire stability region for any given system. Very often the
region of stability is approximated by { x I V(x) < Vcr ),
where Vcr is defined as the critical energy of the postfault
system. All direct methods solve the transient stabilitv
problem by computing the stability index, usually called the
Transient Energy Margin (or simply Energy Margin) which is
the excess of critical energy over transient energy a t fault
clearing conditions. A positive energy margin indicates that
the system is stable while its magnitude indicates the margin.

The TEF defined for the postfault system is constructed as There are t w o general approaches t o compute Vcr for
the sum of the first integrals: classical systems, the controlling UEP approach [3,51 and the
PEBS approach [5,91. In the former approach, Vcr is taken as
the potential energy of the postfault system a t the controlling
UEP. In the latter approach, a sustained fault trajectory is
simulated and along this trajectory, the first potential energy
peak is identified which is then used t o define Vcr.

Efforts have been made t o extend these t w o approaches t o

include detailed power system models. Fouad et al I101
extended the controlling UEP approach by replacing the
= vpe (2.12)
controlling UEP with a relevant Peak Point, based on the
observation that time derivatives of the generator control
state variables do not vanish a t PEBS crossing for marginally
where is usually the post fault Stable Equilibrium Point unstable systems. Athay et al [51 extended the sustained
(SEP). Note that Eq. (2.12) applies t o all power system fault approach and concluded that the integral in Eq. 2.1 2 is
models. path dependent, implying that using linear trajectories for
generator internal voltages and rotor angles t o calculate
The kinetic energy (KE) term in Eq. (2.12) needs t o be energy margin could lead t o errors. In [ I 11. Sauer et al
corrected [71 t o properly account for that portion of kinetic extended the sustained fault approach further by making
energy that contributes t o system separation. correction t o the fast state variables along the faulted
trajectory. It appears that more research efforts are desired
The corrected kinetic energy is defined in equations 2.13 - t o enhance these methods and provide a theoretical
2.1 7,where N, is the number of the advanced machines and foundation for the characterization of the stability boundary
N, is the number of the remaining machines. for detailed power system models.

Authorized licensed use limited to: ANDERSON SOSA. Downloaded on February 13,2025 at 17:48:54 UTC from IEEE Xplore. Restrictions apply.
1526
In adding energy margin calculation into conventional time this fault in such a w a y that the system is just marginally
domain simulation, one could suggest that the calculation stable. The transient energy injected by the fault reinsertion
process has t o be computationally efficient and can be easily then defines the energy margin of system originally simulated
incorporated into any time domain simulation program. in the first stage. This intuitive approach, however, is
Understandably, the accuracy of the energy margin produced impractical since it involves critical fault clearing time. A n
from time domain simulation can be relaxed since it is not alternative approach is t o re-insert a permanent fault at the
being used t o determine the absolute stability of the study instant of dominant PE peak so that the system is given a
system. The absolute stability of the study system is second "push" t o move up the potential energy well until it
accurately determined b y simply examining the simulated time finally escapes from the well.
responses of various quantities such as rotor angles. Instead,
the energy margin so produced is used t o develop a "feel" for Numerous simulations reveal that the sustained fault
the stability margin so that the number of stability simulations approach, w h e n applied t o detailed system models, produces
can be reduced in deriving transient stability limits. The PEBS very pessimistic energy margins due t o the fact that
approach is ideal for this purpose since its algorithm is generator internal fluxes are severely suppressed by the
computationally efficient and robust, although its accuracy is permanent fault. After numerous trials, reinsertion of a
usually not as good. pseudo-sustained (intermittent) fault is recommended which
provides a good compromise between computational efforts
The method proposed in this paper for adding energy margin and accuracy. This pseudo-sustained fault (pseudo-fault-on)
calculation into conventional time domain simulation follows has the capability of slowing d o w n the generator internal flux
the basic approach of 161. The method proposed in [61 first decay before the system trajectory reaches the PEBS. It is
identifies the dominant Potential Energy (PE) peak along the observed that the energy margin so produced is still on the
postfault system trajectory. Once this is done, a line search conservative side, especially for the cases which are very
is performed along the direction defined by the postfault SEP stable.
and the dominant PE peak point until this line intersects the
PEBS. In performing the line search, all detailed generator
models are converted t o their equivalent classical models. The 4.0 PROPOSED ALGORITHM
energy margin for stable cases is then defined as the potential
energy difference between the dominant PE peak point the The proposed two-stage method is summarized as follows:
PEBS intersection point (Eq. 3.1 1.
4.1 First Stane

1. Perform conventional time domain simulation as usual by

specifying the fault location, fault type, fault impedance, the
(3.1) switching sequence, and the simulation period. For first swing
stability, a simulation period of up t o 4 seconds is usually
adequate. Shown in Fig. 4.1 are t w o system trajectories
projected into the space of rotor angles, Trajectory S-A-E for
The method proposed in this paper replaces the line searches, an unstable case and Trajectory S-A-6-F for a stable case. A t
which are less accurate, with a psuedo-fault-on simulation t o each time step, also compute the transient KE.
determine energy margin for stable cases. Similar t o the
approach of [61, the proposed method is divided into t w o 2. For the post disturbance period, determine the time a t
stages. Stage 1 is t o perform the time domain simulation t o which the transient kinetic energy reaches a global minimum.
establish the postfault system trajectory. After this stage, the This is also the time the system trajectory reaches its
absolute stability of the system can be determined. If the dominant PE peak (Point B in Fig. 4.1 ).
system is unstable, its trajectory when projected into the
rotor angle space will cross the PEBS. A reasonable stability 3. Save the system conditions a t the dominant PE peak, t o be
index for unstable cases is defined as the corrected transient used later in the second stage.
Kinetic Energy (KE) that remains in the system a t PEBS
crossing [61. This is the portion of the disturbance injected 4. If the system is unstable, the system trajectory will cross
transient KE that the postfault system fails t o convert into the PEBS as shown by Trajectory S-A-E in Fig. 4.1. Usually
potential energy. the time domain simulation is terminated by a subroutine
which detects large rotor angles or large speed deviations.
If the system is stable, its postfault trajectory will not cross After the time domain simulation is stopped, calculate EM as
the PEBS within the simulation period. The dominant PE peak the corrected transient kinetic energy a t PEBS crossing (Point
can be identified for the postfault stable trajectory. A t this E in Fig. 4.1). Point E is taken as the point on the postfault
peak, all disturbance injected transient KE has been converted system trajectory which has the minimum transient KE.
into potential energy and yet the postfault system might still
be able t o convert some more. Energy margin for stable cases 5. If the system is stable, evaluate EM by proceeding with the
is then viewed as a calculation in stage 2 of how much second stage.
"more" transient KE that the postfault system can convert
without going unstable. 4.2 Second Staae

Intuitively, one can perform a second stage in which a fault 1. Restore system conditions a t the dominant PE peak point
is re-inserted a t the instant of the dominant PE peak and clear and initialize EM as zero a t this point.

Authorized licensed use limited to: ANDERSON SOSA. Downloaded on February 13,2025 at 17:48:54 UTC from IEEE Xplore. Restrictions apply.
 1527
2. Apply a fault a t t h e same fault location as specified in the
first stage for one time step. The step size is the same as
that used in the previous stage.
Y

3. Clear the fault and compute the potential energy difference

I r
(Eq. 4.1) for this time step using the trapezoidal rule.
Increment EM by this amount. Note that the quantities f,' and
f,'A' are computed w h e n the fault is off.
J
Ekm

4. With the fault cleared, advance one time step and compute
the PE difference for this time step. Increment EM again by
this amount.

5. Repeat Steps 2 t o 4 until the EM reaches a peak. This

(P.U.)
be9 ~L+~zH;-+ 1 4 81, vs

condition is detected by AV, = 0 , which terminates the Fig. 5.2: Excitation System
second stage.
The critical clearing time was determined for a 3-@fault at
the generator high tension bus, cleared by tripping one of the
t w o transmission lines. Time domain simulations were per-
formed for different clearing times. To compare the perfor-
mance, t w o energy margins were computed for each fault
clearing time; one for the proposed method, and one for the
method of [6] using equivalent classical models and line
searches. See Table 5.1. Energy margins for the proposed
'\ UEP method are plotted in Fig. 5.3.
Z
\
\ Table 5.1 Energy Margins for SMlB System
\
\
\ .\%e8 Fault CI E.M. (pu) Stability E.M. (pu)
Time Proposed Method
\
\ (SI Method [61

82 0.100 2.521 Stable 4.91 8

0.104 2.151 Stable 4.297
Fig. 4.1 Projected System Trajectories 0.108 1.776 Stable 3.692
0.112 1.397 Stable 3.065
5.0 TEST RESULTS 0.1 16 1.021 Stable 2.426
0.1 20 0.652 Stable 1.590
5.1 Sinnle Machine Infinite Bus System 0.124 0.302 Stable 0.840
0.128 0.006 Stable 0.055
The first test system is a single machine infinite bus system 0.132 -0.498 Unstable -0.498
as shown in Fig. 5.1. The machine represents a big thermal 0.136 -1.070 Unstable -1.070
station delivering power t o a remote infinite bus over t w o 0.140 -1.660 Unstable -1.660
transmission lines. The machine is formed by lumping four 0.144 -2.267 Unstable -2.267
thermal units and is loaded a t 1700 MW. It is modelled as a
round rotor generator and is equipped with a static exciter The EM'S for the stable cases were all computed in Stage 2
and a power system stabilizer, as shown in Fig. 5.2. of the proposed method. For this single machine infinite bus
Generator saturation is also represented. system, the initial rotor angle w a s about 65 degrees, and the
rotor angle a t PEBS crossing as computed in Stage 2 was
about 149 degrees. As the fault clearing time increases, the
dominant PE peak gets closer t o the PEBS so that the pseudo-
fault-on trajectory takes less time t o cross the PEBS. For the
fault clearing time of 0.1 00 seconds, the pseudo-fault-on
trajectory required 0.164 seconds of time domain simulation
t o go from the dominant PE peak t o the PEBS. For the fault
clearing time of 0.128 seconds, it only took 0.021 seconds
t o get t o the PEBS. Note that the system first swing transient
stability simulation took about 2 seconds t o simulate. For this
Fig. 5.1: Single Machine Infinite Bus System system, it takes about 10% more CPU time t o compute an
energy margin from the conventional time domain simulation.

Authorized licensed use limited to: ANDERSON SOSA. Downloaded on February 13,2025 at 17:48:54 UTC from IEEE Xplore. Restrictions apply.
I
1528
271 I I I I I I
c ' ' I I

0.10 Fault Clearing Jirne(s) 0.15

Fig. 5.5: Power vs Angle
Fig. 5.3: Energy Margin vs Fault Clearing Time
5.2 OH 27-Generator Svstem
The two-stage process for calculating energy margins for
stable cases is shown in Figures 5.4 and 5.5. Point B is the The second test system is used for investigating transient
system at fault clearing where the potential energy is stability problems associated w i t h the Bruce nuclear
arbitrarily defined as zero. The system reaches its dominant generating station at Ontario Hydro. This generating station
PE peak at C. The system trajectory between B and C is has eight units of about 900 MW each. It delivers power to
simulated in Stage 1. A t C, fault reinsertion is applied so that major load centers in Ontario through the FABC (Flow Away
the system increases its potential energy further until it From Bruce) interface consisting of four 5 0 0 kV lines and six
crosses the PEBS a t E. The potential energy difference 2 3 0 k V lines (Fig. 5.6). The four 5 0 0 kV circuit lines are the
between C and E in Fig. 5.4 defines the energy margin, which t w o BXM circuits between Bruce and Milton, and the t w o
is equivalent t o area C-D-E in Fig. 5.5. This area represents BXL circuits between Bruce and Longwood. The most critical
the additional transient energy that the postfault system can contingency for this system is a LG-LG fault on the BXM
absorb. As shown, this area is less than area C-D-E' based double circuit line, cleared by tripping these t w o circuits.
on the power-angle relationship obtained from a marginally
stable case. The proposed method thus yields a conservative In this test system, the Bruce machines are modelled as
energy margin, whereas the method of 161 yields higher round rotor generators, equipped w i t h Transient Stability
values. Excitation Control (TSEC). A typical field voltage behaviour is
given in Fig. 5.7. All other machines were represented by
The simulation of the pseudo-fault-on trajectory is shown in classical models. Constant impedance loads were used
Fig. 5.5. When the fault is on, the electrical power output although this is not a limitation.
drops to zero. Every time the fault is removed, the time
domain simulation program computes the electrical power
which can then be used t o compute fi in Eq. 3.1. All of these
computations can be easily handled by any conventional time
domain simulation program. Bruce E
500 kV

t E i Bruce A
500 kV

i -I

i
i
i
I- i -1
01
0.0
/
IB-4'1
0.14 0.28
I I I
0.42
I 1
0.56
I I
0.70
Fig. 5.6 The Bruce System
Time (s)
The system has one BXL circuit out of service initially, and
Fig. 5.4: Potential Energy vs Time the task is to derive the transient stability limit for the
stability interface FABC, for the loss of the t w o BXM circuits.

Authorized licensed use limited to: ANDERSON SOSA. Downloaded on February 13,2025 at 17:48:54 UTC from IEEE Xplore. Restrictions apply.
 1529

test, it w a s assumed that three reactors at Bruce, of

In this Table 5.2: OH 27-Generator System
125 M V A R s each, were not available for postfault switching.
FABC Energy Stability
Margin
(MW) (PU)

4500 5.41 Stable

4600 4.66 Stable
4700 3.95 Stable
4800 3.24 Stable
4900 2.57 Stable
5000 2.00 Stable
5100 1.55 Stable
5200 0.51 Stable
5300 0.27 Stable
5400 0.156 Stable
5500 0.001 Stable
5600 -2.800 Unstable

6.0 FROM MARGINS TO LIMITS

0 2 4 6 8 10 Energy margins may not mean much t o power system

Time (s)
dispatchers, but the sensitivity of energy margins w.r.t.
various operating parameters such as interface flows can be
used t o speed up the derivation of transient stability limits for
Fig. 5.7 Typical Field Voltage Behaviour these interfaces. Although exact analytical sensitivities can be
obtained using the recent techniques I1 21, some approximate
sensitivity information can be calculated numerically using the
proposed method. For example, in the case of the OH 27-
generator system, the first order sensitivity of the energy
margin w.r.t. FABC (the operating parameter under consid-
eration) a t the 5000 M W level can be obtained by perturbing
the FABC f l o w by 100 M W on both sides, yielding approx-
imately (1.55-2.57)/200 pu/MW or -0.005 pu/MW. Now,
using this sensitivity t o calculate the approximate stability
limit, w e get

(0. - 2.) pu = -0.005 pu/MW * ( Limit - 5000 M W )

Hence, the approximate limit is 5 4 0 0 M W as compared to the

exact limit of 5500 M W (1.8% error). Note that the use of
the energy margin sensitivity would eliminate the need t o
continue Table 5.2 beyond the F A B C = 5 1 0 0 M W point.

We should note that the use of first-order numerical

sensitivities may, in some other cases, lead t o larger errors in
estimating the stability limit. W e have found that for many
highly stressed systems with complex modes of separation,
the Energy Margin - interface f l o w characteristics become
more non-linear near the stability limit. For such cases, the
use of higher-order analytical sensitivities should be
Fig. 5.8 Energy Margin vs FABC Flow considered.

7.0 DISCUSSION AND CONCLUSIONS

The test results were tabulated in Table 5.2 and plotted in
Fig. 5.8. For a very stable case ( F A B C = 4 5 0 0 MW), the A practical method has been proposed t o derive a first swing
pseudo-fault-on trajectory took about 0.4 seconds t o reach stability index as an additional output quantity from the time
the PEBS from the PE peak point. For the marginally stable domain simulation. The two-stage method presented in this
case, it took less than 0.1 seconds of time domain simulation. paper is easy t o incorporate into any time domain simulation
To simulate the system first swing, it required about four program. All quantities required for energy margin calculation
seconds of time domain simulation. The energy margin calcu- are readily available so that only a small amount of coding
lation only increased the overall CPU time by about 10%. effort is required.

Authorized licensed use limited to: ANDERSON SOSA. Downloaded on February 13,2025 at 17:48:54 UTC from IEEE Xplore. Restrictions apply.
 ~

1530
The proposed method has been successfully demonstrated methods that use sustained faults with detailed power
using a prototype program and t w o data sets: a single system models," PES Paper 88 S M 688-4, IEEE PES
machine infinite bus system and the OH 27-generator system. Summer Power Meeting, Portland, July 1988.
Compared t o the conventional time domain simulation, the
additional simulation of the fault-on trajectory for energy 1 2 Vittal, V., E-Z. Zhou, C. Hwang and A.A. Fouad,
margin calculation requires only a modest increase in CPU "Derivation of stability limits using analytical sensitivity
time, estimated a t approximately 1 0 % . of the transient energy margin," IEEE Trans. on Power
Systems, Vol. 4, No. 4, Oct. 1989, pp. 1363-1372.

8.0 REFERENCES
9.0 BIOGRAPHY
1. D.C. Lee, and P. Kundur, "Advanced Excitation Control
for Power System Stability Enhancement," ClGRE Paper C.K. Tang: Received his B.A. Sc (19 7 9 ) and M. Eng. (1984)
38-01, 1985. degrees in Electrical Engineering from University of Waterloo
and University of Toronto respectively. He joined Ontario
2 . El-Kady, M.A., C.K. Tang, V.F. Carvahlho, A.A. Fouad Hydro in 1 9 7 9 and is currently a supervising engineer in the
and V. Vittal, "Dynamic security assessment utilizing the System Performance Section of Power System Operations
transient energy function method," IEEE Trans. on Power Division. He is also a part-time Ph.D. student at McMaster
Systems, Vol. PWRS1, No.3, Aug. 1986, pp. 284-291. University.

3. El-Kady, e t al, and A.A. Fouad, e t al, "Direct analysis of C.E. Graham: Received his B.A.Sc. (1970) and M.A.Sc.
transient stability for large power systems," EPRl Report (1 9 7 1 ) from the University of Waterloo. He joined Ontario
EL-4980, Dec. 1986. Hydro in 1 9 7 1 is currently the manager of the System
Performance Section, Power System Operations Division.
4. Fouad, et al, "Transient stability program output
analysis," EPRl Research Project 2206-3, Final Report, M.A. El-Kady: Was born in Cairo, Egypt in 1951, He obtained
1985. his B.Sc. and M.Sc. in Electrical Engineering from Cairo
University in 1 9 7 4 and 1977, respectively, and his Ph.D. in
5. T. Athay, e t al, "Transient Energy Stability Analysis", US Electrical and Computer Engineering from McMaster
Department of Energy Publication No. CONF-790904-PI. University in 1 9 8 0 . He is currently the Development Planning
Manager in Power System Operations Division, Ontario
6. Maria, G.A., C.K. Tang and J. Kim, "Hybrid Transient Hydro. He also holds the position of Professor (part-time) in
Stability Analysis," IEEE Trans. on Power Systems, Vol the Department of Electrical and Computer Engineering,
5, No. 2, May 1990, pp. 384-393. McMaster University. Dr. El-Kady's responsibilities a t Ontario
Hydro include Management of Development Work Programs
7. Fouad, A.A. and V. Vittal, "Power system response t o a and Application Software for Real-Time and Operations
large disturbance: energy associated with system Planning use. His Research and Development activities at
separation," IEEE Trans. on PAS, Vol. 102, Nov. 1983, McMaster University include Contracts from Government,
pp. 3534-3540. Industry and R&D funding Organizations, Consultations, Post-
Graduate supervision and teaching in the general area of
8. Chiang, H.D., F.F. W u and P.P. Varaiya, "Foundations of Computerization and Automation of Power System Operation
the potential energy boundary surface method for power and Planning.
system transient stability analysis," IEEE Trans. on
Circuits & Systems, Vol. CAS-35, June 1988, pp. 1 6 0 - R.T.H. Alden: Obtained his B.A.Sc. degree in Electrical
172. Engineering from the University of Toronto in 1960. He
worked as a meter design engineer for the Sangamo
9. Kakimoto, N., Y. Ohsawa and M. Hayashi, "Transient Company and returned t o the University of Toronto, obtaining
stability analysis of electric power system via Lure' type his M.A.Sc. and Ph.D. degrees. He has been a t McMaster
Lyapunov functions, parts I and 11," Trans. IEE of Japan, University since 1970, where he is a professor of Electrical
Vol. 98, No. 516, MayIJune 1 9 7 8 . and Computer Engineering in the Power Research Laboratory.
His research interests include power system stability,
10. Fouad, A.A. e t al. "Direct Transient Stability Assessment interactive programming, as well as education. He is a
with Excitation Contro1,"lEEE Paper # 88SM657-9,1988. registered professional engineer in the province of Ontario. He
has just completed a 4 year term on the IEEE Board of
11. Sauer, P.W., A.K. Behera, M.A. Pai, J.R. Winkelman and Directors: 1988-89, Region 7 Director and 1990-9 1, Vice
J.H. Chow, "Trajectory approximations for direct energy President for Regional Activities.

Authorized licensed use limited to: ANDERSON SOSA. Downloaded on February 13,2025 at 17:48:54 UTC from IEEE Xplore. Restrictions apply.