Transactions on Power Systems, Vol. 7, No. 1.

February 1992 243

VOLTAGE STABILITY CONDITIONS CONSIDERING IDAD CWUZACTERISTICS
M. K. Pal, Senior Member
Public Service Electric and Gas Company
Newark, N.J. 07101

Abstract: It is shown that when the characteristic of as follows: Assuming that an equilibrium, satisfying
the composite load of a typical utility system is taken the operating constraints, i.e. a viable power flow
into account, large disturbance voltage stability is solution exists, the equilibrium state is steady-state
assured by the existence of the stable equilibrium (small signal) stable if all the eigenvalues of the
state of the post-disturbance system, as determined system, linearized around the operating point, have
from the standard power flow model. When the load negative real parts. Large disturbance stability is
contains static components, stability limits extend assured if the system state at the end of the disturb-
considerably. The exact limit for such loads can also ance lies within the region of attraction of the stable
be determined from a power flow model, properly equilibrium point of the post-disturbance system.
modified to reflect the static component of the load. Voltage instability is primarily a small signal
In specific situations, when the bulk of the load is phenomenon. In the brief period following a large
composed of fast response loads, a correct assessment disturbance most loads behave as static loads, e.g.
of voltage stability would require comprehensive constant impedance or, at worst, constant current. It
analyses, employing detailed dynamic models of all has been recognized that for such loads voltage insta-
system components. The use of the conventional power bility cannot occur. By the time the load dynamics
flow model may lead to considerable error. start having any impact on the system response, the
disturbance would have reached the small signal status,
Kewords: Voltage Stability, Voltage Collapse, Load provided of course, synchronous stability has been
Characteristics. maintained. This would be particularly true for compo-
site loads in a large utility system. Therefore, it is
INTRODUCTION only necessary to assure small signal stability of the
final post-disturbance system. A rigorous justification
The existing and continuing trend toward heavier supporting this view will be provided in this paper.
line loadings has resulted in an increased need for
reactive power management, and raised the possibilities Voltage instability is largely determined by load
of voltage instability and collapse. Concerns for characteristics and the available means of voltage
voltage instability and collapse are prompting control. For true voltage instability, at least a part
utilities to better understand the phenomenon so as to of the total load must be of self-restoring (constant
devise effective, efficient and economic solutions to MVA) type. By the same token, voltage stability prob-
the problem. The problem of reactive power and voltage lems become less severe as static or voltage sensitive
control, i.e. maintaining an acceptable system voltage loads constitute a portion of the total load. The
profile by providing adequate reactive supports at precise determination of the stability limit for such
appropriate locations so as to meet system reactive loads is demonstrated in the paper.
need efficiently, is well U erstood and reported
extensively in the literature'-' It is not, however, SYSTEM AND LOAD MODEL
well appreciated that maintaining a good voltage pro-
file does not automatically guarantee voltage stabili- We will adopt the prevailing view that voltage col-
tr, and that voltage instability need not be associated lapse is a relatively slow process and, therefore, it
with low voltages, although frequently it is. can be (although it need not be) studied in isolation,
i.e. by separating it from the angle (synchronous)
Traditional methods of voltage stability investiga- stability aspects. We will concentrate only on the
tion, for example, 1 3 ) - [ 7 ] , have relied on static ana- dynamics that are likely to have appreciable impacts on
lyses using the conventional power flow model. Defini- voltage stability. The dynamics that affect voltage
tions and tests for voltage instability have been pro- stability are the dynamics of the load and the dynamics
posed accordingly. Such analyses will, however, deter- of the voltage control devices, especially those of the
mine the power limit, which may or may not be indicat- generator excitation controls.
ive of voltage stability limit. We note that only a
dynamic system can experience instability. Voltage Important insights into the mechanisms of voltage
instability is load driven. In a voltage stability instability and collapse can be obtained through the
analysis it is, therefore, imperative that a dynamic use of a simplified model, incorporating only the
system formulation, including the pertinent load dyna- elements that are dominant in controlling these mecha-
mics, be employed. We will follow the well established nisms. The conclusions drawn from the simplified
procedures for assessing stability of dynamic systems. analysis can then be verified by detailed simulations.
In simple terms the stability conditions may be stated The power system we will analyze is shown in Figure 1.
A single generator is supplying a load over a trans-
mission line. For simplicity, all resistances and line
91 SM 490-3 PWRS A paper recommended and approved chargings have been neglected. However, any reactive
by the IEEE Power System Engineering Committee of support that might be present at the load end of the
the IEEE Power Engineering Society for presentation transmission ljne has been included.
at the IEEE/PES 1991 Summer Meeting, San Diego,
California, July 28 - August 1, 1991. Manuscript It has been recognized that the most stringent load
submitted February 1, 1991; made available for from the viewpoint of voltage stability is the load
printing May 29, 1991. that maintains a constant MVA characteristic, either
due to the nature of the load itself or due to the
action of control mechanisms that are intended to main-
tain constant voltage at the load supply point, such as
LTC's, distribution voltage regulators, etc., thereby
rendering any load constant MVA. (Note that even with-
0885-8950/92$03.oOo1992IEEE
244
an unstable equilibrium point. This may be verified by
linearizing equation (1) around the equilibrium points,
and applying the condition for stability.

vs I!? jx j B 1 VL
T VL(l-BX) VL pI
“Lo
1.0 -
tn
Fig.1 A power system with constant source voltage c,
rl C
supplying a load through a transmission s!
line. The phasor diagram applies to a unity
power factor load.
0.5 -
out voltage control action, certain apparently static
loads, such as thermostatically controlled heating
loads, due to their constant energy consumptions, tend 0.5 Po 1.0
to behave as constant MVA loads in the longer term.) Power
We will, therefore, concentrate primarily on constant
MVA loads, and show the effect on voltage stability
when parts of the load possess characteristics other Fig.2 Power-Voltagecurve Of shown in Fig.1.
than constant MVA, such as constant impedance.
It will be shown that if the state reached at the
In a voltage stability analysis, it is important to end of the disturbance lies anywhere on the portion
model the relevant dynamics of the load. Employing a V ACB of the P-V curve, the post-disturbance system
static model for constant MVA loads can lead to errone- w h be stable, i.e. the operating point will settle
ous and, often, misleading results. A constant MVA load at the stable equilibrium point A. The region V ACB,
is not a static load. This means, it cannot jump
instantaneously from one demand level to another as the
i.e. the region of the P-V curve to the right the 3
unstable equilibrium point is the region of attraction
demand changes. Following a change in the demand, the of the stable equilibrium point of the post-disturbance
load will at first change according to its instantane- system.
ous characteristic such as, constant impedance or
current. It will then adjust the current drawn from
the system until the load supplied by the system satis- From the phasor diagram (Fig.1)’
fies the demand at the final system voltage. Similarly,
when there is a sudden change in the system voltage, 2 VS2
such as, following a disturbance, the load will change
momentarily. It will then adjust the current (or imped- vL - (1-BX)’ + G2X2
ance) and draw, from the system, whatever current is
necessary in order to satisfy the demand. The process Equation therefore, be expressed as
is not instantaneous. A definite time lag is involved.
A load model, in order to be suitable for voltage d PoX2(G-Glo)(G-G20)
stabtlity analysis, must recognize this basic fact. T-G-
Ldt (1-BX)2 + G2X2
For our purpose, the dynamic behavior of constant
MVA load may be represented by a simple first order where,
delay model. Several forms of this model are given in
2
[ 8 ] - [ 9 ] . For simplicity, we will assume a unity power Vs 5 JVs4 - 4P02X2(l-BX)2
factor load, whose dynamics may be represented by
Glo,20 2P0X2
d
T - G
Ldt
- Po - VLLG (1) The two values of G correspond to the two voltage
solutions for the given power P . In Figure 2, these
where two solutions correspond to the“ equilibrium points B
V is the load voltage and A, respectively. The two voltage solutions are:
PL is the power set point
Go is the load conductance which is adjusted to
maintain constant power Vs 2 +_ 4VS4 - 4P02X2(l-BX)2
T is the load time constant
L v-2
L
- 2 (1-BX)
REGION OF ATTRACTION
The higher value of G corresponds to the low
Before a discussion of the results of the paper can voltage solution, and vice versa. This is verified by
be presented, it will be necessary to examine the noting that
region of attraction of the stable equilibrium point in
2
the context of the system power voltage (P-V) curves.
Consider the simplified radial power system with
vL1 G20 - vL22Glo -
constant sending-end voltage shown in Figure 1. Assum- Equation (2) shows that if the initial state is to
ing, for simplicity, a unity power factor load, the the left of the point A on the P-V curve (G less than
power-voltage curve of the post-disturbance system is G ) , dG/dt is positive, and therefore, the operating
shown in Figure 2. For a constant power load P , the p8Pnt will move to A, at which point dG/dt = 0.
load characteristic, as shown by the vertical ‘line, Similarly, if the initial state is anywhere on the
would intersect the system P-V curve at two points, A portion ACB (G <G<G ) , dG/dt is negative and, there-
and B, corresponding to the two possible equilibrium fore, the ope&?ing E i n t moves to A. If the initial
points. Point A, on the upper portion of the system state is to the left of B, dG/dt is positive and the
P-V curve (the high voltage solution), is a stable operating point moves further away from B. The region
equilibrium point, whereas point B, on the lower to the right of B is, therefore, the region of attrac-
portion of the P-V curve (the low voltage solution), is tion of the stable equilibrium point A.
 245

A physical explanation can be provided as follows:

Consider an initial state in the lower region of the
P-V curve but to the right of the unstable equilibrium
point B. Since in this state the power delivered is
greater than the set-point power, P , the constant
power control mechanism would decreas: the current or
admittance in order to bring the power down. This will,
however, increase the power still further, since the
voltage will increase at a faster rate with decrease in
current or admittance in this region. The operating
point will, therefore, move up the P-V curve until
point C is reached. From this point onwards the same 0.5 PI 1.0 P&7
control command will however, decrease the power. The Power
process will continue until the stable equilibrium
point A is reached. Using the same argument it can be
seen that starting from anywhere on the upper portion Fig.3 System P-V curves and steady-state and
of the P-V curve, the operating point will move to the instantaneous load characteristics.
stable equilibrium point A.
The voltage stability limit for constant MVA load
LARGE DISTURBANCE VOLTAGE STABILITY whose response time is slow compared to the speed of
response of the system voltage control, is the same as
It will now be shown that, in a typical utility sys- the power limit (sometimes called the steady-state
tem, when the overall behavior of the composite load is voltage stability limit without proper qualifications),
taken into account, voltage stability is assured by the obtained from the standard power flow model. In using
presence of a stable equilibrium point in the post- the power flow model, the status of the system voltage
disturbance system. In a typical utility system, even control is important and should be reflected in the
when the loads possess constant MVA characteristics, model. If the generator terminal voltages are not
the overall response time is quite slow. Immediately maintained by excitation control, i.e. when the
following a disturbance, all loads temporarily behave generators are operating at constant excitations, a
as static loads, e.g. constant impedance or, at worst, modification to the power flow model would be
constant current. It is either the nature of the load necessary. It can be shown that under this condition,
itself, or the action of some control mechanism that the voltage stability limit, and therefore, the stable
eventually restores the load to constant MVA. The equilibrium point, is determined by constant voltage
mechanisms by which much of the load would be rendered (generator field voltage) behind generator synchronous
constant MVA act slowly compared to the speed of res- reactance. Failure to recognize this may result in
ponse of the voltage control equipment, which are pri- gross error.
marily generator excitation controls. Voltages at the
controlled buses, e.g. the generator terminal voltages VOLTAGE STABILITY LIMIT WHEN
would, therefore, be restored well before the overall
load characteristics return to constant MVA. In
specific situations, when the bulk of the load has fast Although voltage instability cannot occur for static
load, the maximum power that can be delivered at a
response characteristics, this assumption is not
correct and can introduce considerable error. These given power factor is independent of the type of load.
cases are discussed later. Beyond the point at which maximum power occurs, the
actual power delivered will decrease with increased
Consider the system shown in Figure 1, initially demand. We will now investigate the voltage stability
supplying a load P . Assume a disturbance caused by limit when a portion of the load is static. Assuming,
opening one circui& of a double circuit transmission for simplicity, unity power factor, and a combination
line. The pre- and post-disturbance system P-V curves of constant power and resistive load, the total load
are shown in Figure 3. For initial load power P1, ,the can be expressed as
operating point is A, the intersection of the vertlcal
line at P1 and the steady state pre-disturbance P-V P = V 2(G
L
+ GL) (3)
curve. Immediately following the disturbance, the load
will behave as constant impedance (or, at worst, cons- where G is the conductance of the resistive part of
tant current). For a typical utility system, the the loazd, and G is the conductance of the constant
system voltage would be restored well before the load power part of the load whose dynamics is given by equa-
returns to constant power. The operating point will, tion (1).
therefore, temporarily move to point A ’ , the inter-
section of the instantaneous load characteristics shown From the phasor diagram of Figure 1, the real and
by the dashed curve a, and the steady-state post- reactive power balance equations can be written as
disturbance P-V curve. Since A’ is within the region of
attraction of the final steady-state equilibrium point V (G
L
+ GL)X - VS sin @ (4)
B, the operating point would move to B and the system
will be stable. The existence of a stable equilibrium VL(l - BX) = VS COS @ (5)
point in the final post-disturbance system, therefore,
guarantees voltage stability. Linearizing equations (l), (4) and ( 5 ) , and elimina-
ting the non-state variables, we obtain,
If, on the other hand, the initial load were P
2
(greater than the maximum power that can be supplied by d 2GX tan 8
the post-disturbance system), the operating point would 2
T-AG--VL [I- 1 QG
temporarily move to C ’ following the disturbance. Since Ldt (G+GL)X tan@ + (1-BX)
there is no equilibrium point for the post-disturbance
system at the assumed load, voltage collapse will ensue Using (4) and ( 5 ) . the above reduces to
if the load attempts to maintain constant power. The
collapse, however, will not be immediate, and general- d G cos28 + GL
ly, sufficient time will be available for applying cor- T-AG--V AG
rective measures to restore stability. Ldt G+GL
246

For stability, EXTENSION TO LARGE NETWORKS

G cos28 + GL > 0, which yields The above result can be readily extended to large
2 networks if we first relate the stability limit, as
cos28 > - VL GL/Po (7) determined above, to the system Jacobian, modified to

where
load.
-
Po ,G:V is the constant power part of the
reflect the static part of the load.
For the simple example considered above, the incre-
mental power flow equations cam be written as:
Voltage stability limit is reached when - (VsVL/X)cos8 2VLGL-(VS/X)sin8
cos28 = - vL~G~/P~ (8)
(VSVL/X)sine

Note that the conventional power flow Jacobian has

been augmented by terms arising from the resistive part
of the load.
We now determine the singularity condition of this
modified Jacobian. The Jacobian is singular when the
determinant is zero. After evaluating the determinant
and carrying out the necessary algebraic manipulations,
the condition for singularity is obtained as,

I I which is the same as the condition for voltage stabili-

0.5 1 .o ty limit. Therefore, the voltage stability limit, when
Power the load contains static components, is indicated by
the singularity of the modified Jacobian.
The extension to large networks follows directly
Fig.4 System P-V curve and load characteristics from the above analysis. Power flow equations are
for a combination of constant power and first written, separating the voltage dependent part
resistive load. of the load from the constant power part, as shown in
equations (12)-(13).
Referring to Figure 4, assume that the load charac-
teristic for a given load combination of constant power P + P(V) = f(e,v) (12)
and resistive load is as shown by curve a. It inter-
sects the system P-V curve at points A and A ' . It will Q + Q(V) = g(e,v) (13)
now be shown that point A (although it is on the lower
portion of the system P-V curve)is a stable equilibrium The linearized power flow equations are written as
point, whereas point A' is unstable. follows :
In order to prove the above, we first increase the
load until a limit is reached, i.e. until the load
characteristic becomes tangent to the system P-V curve.
The load increase can be achieved by increasing either
the constant power part or the resistive part of the
load, or both. where,
Jpv' = Jpv - aP(V)/aV
We have,

At the generator and other voltage controlled buses,

AV's are zero, so that the Jacobian would be adjusted,
Substituting V obtained from the reactive power as is customary in a power flow solution under perfect
balance equation (4) into (9), we obtain, voltage control. It is important to account for any
non-conformity of the voltage controlled buses. For
2 example, when any of the generators reaches excitation
Po +- (10) limit, its terminal voltage is no longer constant.
2X( 1-BX) Instead, the constant voltage under fixed excitation is
the voltage behind synchronous reactance (the field
In order to obtain the limit, we take the deriva- voltage), as noted earlier. This should be reflected
tive of either P or G with respect to 8 , and equate in the Jacobian. Voltage stability limit would be
to zero. After gome akgebraic manipulations we arrive indicated by the singularity of the modified Jacobian.
at the condition at limit as,
LARGE DISTURBANCE VOLTAGE STABILITY FOR MIXED LOADS
COS2 = - V 2G /P
L L O It has been shown that when a part of the total load
which is the same as the voltage stability limit ob- is static, voltage stability limit can extend beyond
tained above (equation 8 ) . the maximum power point on the system P-V curve. For
such loads stability can be maintained for initial
The limit shown on Figure 4 as point B, the point at pre-disturbance load greater than the maximum power
which the load characteristic, b, is tangent to the capability of the post-disturbance system. This is
system P-V curve, is, therefore, also the voltage because a stable equilibrium point may be possible in
stability limit. It follows that the equilibrium point the post-disturbance system, although the actual power
A is stable and A' is unstable. delivered will be reduced following the disturbance.
 247

In order to illustrate this further, consider the is at A, the intersection of the pre-disturbance steady
case when about 50% of the load is resistive. The state P-V curve, the transient P-V curve and the
system P-V curves and the steady-state load character- steady-state load characteristic, which is a vertical
istics for this case are shown in Figure 5, in solid line at P for the constant power load assumed.
lines. For a total load P1, A is the initial, pre-
disturbance operating point. The equilibrium points
following the disturbance are shown on the post- I pre-dist. Pre-dist.
disturbance P-V curve at B and B'. (Recall that when
/
a part of the load is static, both equilibrium points P) 1.0
P
may be on the lower portion of the system P-V curve.)
c,
rl

9
I Pre-disturbance a , 1 0.5

1.0
m
n
c,
rl
I I -1.u_
I I
9 0.5
PO
0.5 Power

Fig.6 Large disturbance voltage stability for fast

response loads using "transient" P-V curves.
At the power level shown, the system is
stable following the assumed disturbance.
Power
Following the disturbance the operating point first
Fig.5 System P-V curves and steady-state (solid moves to A ' , the intersection of the post-disturbance
lines) and instantaneous (dashed lines) transient P-V curve and the instantaneous load charac-
load characteristics when part of the load teristic. Since A' is within the region of attraction
is static. of the stable equilibrium point, B', on the transient
P-V curve, the system will be transiently stable. The
Using a similar argument as before, it can be con- operating point will eventually settle at B, the
cluded that the region to the right of B' (the unstable intersection of the post-disturbance steady state and
equilibrium point) is the region of attraction of the final transient P-V curves and the steady state load
stable equilibrium point B. characteristic.
Following a disturbance the operating point moves to
A', the intersection of the postidisturbance P-V curve
and the instantaneous load characteristic, shown by
the dashed line a'. Since it is within the region of
1.51
Post-dist.transient I
attraction of the stable equilibrium point, B , the

zi
system will be stable. The approximate path of the ,steady
operating point from initial to final state is marked 1.
by arrows as shown. P
Post-dist. 3
transient
Note that if the initial total load is P2, an c,
4

equilibrium point would not exist in the post-disturb- P

ance system, and voltage collapse would occur follow- 0.
ing the disturbance. Once again we can conclude that
large disturbance voltage stability is assured by the
existence of a stable equilibrium point in the post-
disturbance system.
e r e - G :
Pre-dist.
steady state I
VOLTAGE STABILITY FOR FAST RESPONSE LOAD

In the voltage stability analyses for constant MVA

Fig.7 Illustration of voltage instability follow-
loads presented so far, it was assumed that the mecha-
ing a large disturbance for a fast response
nisms that restore the loads to constant MVA act much
load. At the load level shown, a stable
slower compared to the mechanisms that restore system
equilibrium point exists in the post-dis-
voltages, which are primarily the generator voltage
turbance system. In order to maintain volt-
controls. When the load response speed is comparable
age stability following the assumed large
to the speed of response of the voltage control
disturbance, some control action, e.g.
devices, the assumption of constant system voltage may capacitor switching would be necessary.
introduce considerable error.
A conceptual understanding of the issues involved in Figufe 7 shows the case when the initial load is at
the voltage stability for such loads may be obtained a somewhat higher level, so that an intersection
through the use of "transient" system P-V curves. between the steady-state load characteristic and the
Figure 6 shows the steady state and transient P-V post-disturbance system transient P-V curve does not
curves when the load power is P . The transient exist. [The post-disturbance system is, however, steady
curves, shown by dashed lines, were? obtained using a state stable, at the load level assumed.] It is
fictitious voltage, held constant during the transient evident that, without some sort of control action,
period, behind a fictitious reactance. In practice, it voltage will collapse following the disturbance. With
may not be a simple matter to estimate the correct appropriate control action, voltage stability can be
value of this reactance. The present purpose is to maintained. For examgtte, if sufficient capacitors are
provide a conceptual understanding of the issues switched on promptly , stable operation can be res-
involved, and as such, the transient curves are hypo- tored as illustrated in the figure. In the illustra-
thetical. For the load Po, the initial operating point tion, the operating point first moves to A ' following
248
the disturbance. It would then move down the transient problem can be gained by considering a simple two bus
P-V curve and, in the absen'ce of any control action, system and through the use of the torque-slip curve
voltage collapse would follow. With prompt capacitor (for induction motors), or the power-angle curve (for
switching, the system state can be brought within the synchronous motors). It can be shown that, in the case
region of attraction of the stable equilibrium point on of induction motors, both small and large disturbance
the new transient P-V curve, so that it eventually stability would be more-or-lessassured by the exist-
moves to the stable equilibrium point C. Note that if ence of a stable equilibrium state in the post-disturb-
the capacitor switching is delayed so that the ance system. In the case of synchronous motors, the
operating point has moved beyond A" on the transient familiar equal-area criterion would apply for large
P-V curve, stable operation cannot be restored. This disturbances. Small disturbance stability would be
is because the initial state following the capacitor affected by the type of excitation controls the motors
insertion would fall outside the region of attraction might have.
of the new stable equilibrium point. In some extreme When the system load is composed predominantly of
situations it may be necessary to resort to SVC's in motor loads, and their stability may be in question, a
order to maintain stability. detailed dynamic analysis, considering both large and
The objective of the above discussion was to empha- small disturbances, would be necessary. This would
size the need for a detailed analysis, considering both probably apply to large industrial power systems
small and large disturbances, when the bulk of the load generating their own power and supplying, almost
is composed of fast response loads and voltage stabili- exclusively, motor loads. In addition to detailed
ty may be in question. The analysis would require generator modeling, the analysis would require detailed
detailed modeling of generators and their control modeling of motors including rotor flux dynamics,
equipment. particularly, the excitation controls. No mechanical inertias,(excitation control, if present, in
simple criterion can be employed in such situations. the case of synchronous motors) etc. The use of an
In non-critical situations, approximate estimates of equivalent motor to represent a group of dissimilar
voltage stability limits for such loads may be obtained motors served by a substation bus is not advisable, if
based on steady-state analyses, using a constant gener- a precise limit is being sought.
ator internal voltage behind a conservatively estimated Steady-state analyses using the conventional power
reactance, in place of the constant generator terminal flow model, even if the assumption of constant source
voltage as used in the conventional power flow model. voltage (instantaneous control action) is satisfied and
constant MVA load is assumed, will not reveal this kind
MOTOR LOADS AND VOLTAGE STABILITY of instability. The familiar P-V curves do not apply
when the loads are predominantly motor loads. A
Motor loads require special consideration in voltage detailed treatment covering the relationship between
stability analysis. The response speeds of these loads voltage instability and motor instability will be
are comparable to the speed of response of voltage reported at a later date.
control equipment. In addition, their instability can
become the limiting condition.
CONCLUSION
Motors tend to draw more-or-lessconstant power un-
til a limit is reached, at which point they will start It has been shown that in a typical utility system
stalling. Induction motors may stall at a relatively a detailed dynamic analysis of voltage instability and
high voltage level, i.e. at levels at which voltage collapse is unnecessary. When the characteristic of the
instability would not normally occur. The voltage at composite load is taken into account, both small and
which a synchronous motor stalls would depend on the large disturbance voltage stability are assured by the
excitation level. existence of the stable equilibrium state in the final
Instability of motors, especially of induction post-disturbance system, as determined from a standard
motors, is generally perceived as voltage instability. power flow model using a steady state analysis.
In fact, motor instability is distinct from voltage When the load contains static components, stability
instability. For example, if we consider a system limits extend beyond the maximum power point on the
consisting of a constant voltage source, supplying an system P-V curve. The exact limits for such loads can
induction motff load through a transmission line, it also be determined from a power flow model, properly
can be shown that the motor stability limit (the modified to reflect the static component of the load.
maximum torque that can be developed by the motor) will An implication of this result is that, the precontin-
always be reached before the voltage stability limit as gency load can be greater than the maximum power capa-
defined by the maximum load supplying capability of the bility of the post-contingency system, and still main-
system. tain voltage stability following the contingency.
For a given system voltage there is a maximum torque A conceptual understanding of the issues involved in
that an induction motor can develop. If the load torque the voltage stability for fast response loads has been
exceeds this, either due to an increase in the load o r provided through the use of "transient" system P-V
due to a low system voltage, the motor would stall and curves. The necessity of a comprehensive analysis for
reduce the system voltage further. This may lead to such loads, employing detailed representations of all
stalling of motors elsewhere in the system, thereby system components, when voltage stability may be in
giving rise to a cascading system collapse. question, has been demonstrated.
It is not difficult to visualize that for synchro- Motor loads present a special problem in that their
nous motors there can be no voltage instability in a instability can become the limiting condition. Instabi-
strict sense, unless other loads of constant MVA lity of motors may lead to system collapse, making it
characteristics are connected at some point between the appear like voltage instability. A detailed dynamic
source and the motor bus. For a given source voltage analysis is required to identify such instabilities.
and motor excitation, there is a maximum torque, the
pullout torque, that the motor can develop. Beyond ACKNOWLEDGEMENT
this the motor would lose synchronism and stall. The
motor terminal voltage would collapse, making it appear The author would like to thank the Transmission
like voltage instability. Systems Department of Public Service Electric and Gas
A conceptual understanding of the motor stability Co. for interest and support in the work reported.
 249
PEFERENCES

T. J . E. Miller, "Reactive Power Contrcl in Elec-

tric Power Systems," John Wiley and Sons, New
York, 1982.
"Bibliography on Reactive Power and Voltage Con-
trol - IEEE VAR Management Working Group Report,"
IEEE Trans. on Power Systems, Vo1.2 No.2,
pp.361-370, May, 1987.
N. Flatabo, R. Ognedal and T. Carlsen, "Voltage
Stability Condition in a Power Transmission System
Calculated by Sensitivity Methods," IEEE Trans. on
Power Systems, Vo1.5 No.4, pp.1286-1293,Nov.1990.
P. Borremans, et al.,"Voltage Stability: Fundamen-
tal Concepts and Comparison of Practical
Criteria," CIGRE, Paper No. 38-11,Aug.1984.
J. Carpentier, R. Girard and E. Scano, "Voltage
Collapse Proximity Indicators Computed from an
Optimal Power Flow," Proc. of the 8th Power System
Computational Conference, Helsinki, Aug. 1984.
V. A . Venikov, V. A. Stroev, V. I. Idelchick and
V. I. Tarasov, "Estimation of Electrical Power
System Steady-State Stability in Load Flow
Calculations," IEEE Trans. on Power App. and Sys.
Vol. PAS-94,pp.1034-1041, May/June 1975.
C. Barbier and J.P. Barret, "An Analysis of Pheno-
mena of Voltage Collapse on a Transmission
System," Rev. Gen. Elect. T.89, pp.672-690, Oct.
1980.
T. H. Van Cutsem, "Dynamic and Static Aspects of
Voltage Collapse," Proceedings: Bulk Power System
Voltage Phenomena - Voltage Stability and
Security, EPRI EL-6183,pp.6.55-6.79,Jan.1989.
S . Abe, Y. Fukunaga, A . Isono and B. Kondo, "Power
System Voltage Stability,"IEEE Trans. on Power
App. and Sys. Vol.PAS-101, pp.3830-3840,Oct.1982.
(101 Y. Sekine and H . Ohtsuki, "Cascaded Voltage Col-
lapse," IEEE Trans. on Power Systems, Vo1.5 No.1,
pp.250-256,Feb.1990.
Ill] R. J. Thomas and A. Tiranuchit, "Dynamic Voltage
Instability," Proc. of the 26th Conf. on Decision
and Control, Los Angeles, pp. 53-58, Dec. 1987.

received his M.Sc. from the University of

Calcutta, India, and Ph.D. from the University of Aston
in Birmingham, England, in 1961 and 1972, respectively.
He has over twenty-five years of experience in power
engineering, including teaching and research in indus-
tries, universities and electric utilities.
He is a member of the IEEE PES System Dynamic Per-
formance Subcommittee.