IEEE T r a n s a c t i o n s on P o w e rA p p a r a t u sa n dS y s t e m s , vol. PAS-94, no.

3, &y/June 19 75

CONCEPTS OF POWER SYSTEM DYNAMIC STABILITY

F. P.
deMello Laskowski
T. F.
Power Technologies, Inc.
Schenectady, New York

INTRODUCTIOX of two state variables, machine rotor angle

changeA6 and machine field flux linkage change
AEq'. Figure 1 shows the block diagramof this
The problem of dynamic stability syn-
of simple machine without excitation controls,
chronous machines has been explored extensive- connected through some external impedance to
ly and widelyreported inthe technical liter- an infinite inertia system.
ature. The effects of excitation controls on
ciynamic stability and the use of supplementary
stabilizing signals acting through excitation
systems have also been demonstrated ina num-
ber of theoretical investigations and practi-
cal applications.
In spite of the valuable work that has
been reported, and the general level of under-
standing that has been achievedon the subject
of dynamic stability, there still is need for
additional parametric investigations to yield
greater insight into such questions as:
1) What types of system configurations I " , I
and loading conditions are likely to
encounter dynamic instability?
2) Under what situations can the synch- IAEfd
ronous machine itself, without con-
trol effects, produce negative damp-
ing? Figure 1
Block Diagram Showing Linearized Snall
3) In general multimachine environments, Perturbation Relations o f a Single Gen-
what should bean effective distribu- erator SuDplying an Infinite Bus Through
tion of damping action among the
var- External Impedance.
ious synchronous sources?
The developments and results reported in
this paper attemptto lend insight into these Machine oscillatory behavior following
questions.Theyareillustrativeinnature
and it is hoped will motivate additionalpara- an arbitrarily small disturbance was discussed
in terms of electrical torque components de-
metric work exploring these points.
rived from these states by the coefficients

BASIC YACHINE MODEL

Although the results reported in later
sections have been computed using detailed ma- These torque components were viewed in terms
chine models with amortisseur effects, forpur- of their synchronizing action (in phase with
poses of developing concepts the basic block angle) and their damping action (in phase with
diagram presented in Ref.1 will be used. This speed).
block diagram applies to a salient pole ma-
chine without amortisseurs. Insight into factors which affect damp-
ing was developed in Ref. 1. Ons of these fac-
In Ref. 1 the linearized analysis of the tors was labeled inherent machine damping pro-
behavior of the machine was expressed in termsduced by losses in the field. In the block
diagram of Figure 1 this eFfect is produced
through the parameter K4 which accounts for
the demagnetizing actionof a change in angle.
!lamping is contributed by field losses as long
as K4 is positive. This occurs in the usual
situation where an increase in rotor angle
causes an increase in d-axis component of arm-
ature current and demagnetization of flux.
Paper T 74 509-6, recommended and approved by the IEEE Power System A way of evaluating these damping ef-
-
rn
ig Committee of the IEEE Power Eogineering Society for presentationat
the 1974 Joint IEEE/ASME Power Generation Techuial Conference,Miuni Beach, fects is to analyze the open loop function in
Fla., September 15-19, 1974. Manuscript submitted April 29,1974;mde andable Figure 1 (opening the loop at the input to the
for printing August 7, 1974. inertia block l/MS) made up of inertja, angle
 and electrical torque. This open loop function Figure 3 shows the variation of K2X3K4
with operating conditions of the machine in
the configuration of Figure 2. These results
indicate that with reduced active power load-
ing and increased reactive loading on the re-
It is evident that as long as K4 is positive ceiving end generation (approaching synchro-
the phase angle of this function at the fre- nous condenser operation) the inherent machine
quency of crossover (/G(jw)/=l) is less than damping can be negative. The tendency towards
180" indicating damping of the machine rotor negative damping increases as XE increases.
angle swings.
There can be situations, however, where Qa-. 4 Q=-.2
K4 is negative,as can be seen from the expres-
sion for K4 in Appendix A ofRef. 1:
+.4 1 / / 0-0

Eo(xd-xA)[(xe+x )sin 60-re cos 60]

K4 = r '+(xe+x;) (xe+x
q e
and as shown in Figure 5c of Ref. 3.
/// 41.6
- RE=l. 0

For the case of the machine without damp-

///, (11.8
er winding effects, K4 is negative when did/d6
is negative, i.e. when [(xe+xq)sin6,-r COS^^]
is negative (see Figure 2 for the definition
of re and Xe). In these situations the machine
is supplying part o f a l o c a l load which is also
being supplied by the remote generation. De- P
pending on the amount of power being supplied
by the remote generation, the local machine
angle can be considerably behind the angle of
the sending end machine and could result in a
decrease in d-axis armature current with in-
crease in angle causing K4 to be negative. The
torques produced by induced currents in the +.3 xE=.10
field now have components out of phase with
speed and produce negative damping. RE=l. 0
This phenomenon of negative damping was +.2
recognized in the past in single machine ver-
sus infinite bus situations where the machine +.l
considerable
through
was
external
resistance. As shown in Ref. 1 the case of
the configuration with localload canbe equiv-
alenced to the series impedance configuration
by a Thevenin equivalent which results in ex-
actly thissitiation of seriesresistance
(Figure 2). - .1

+.4 .-
Q = - .3

xE= 5
+.2 Q=-. 1
QE=l. 0
re = -
2 2
Q= 0
REE'+ .1 Q=.1
2
x I 'ERE
e 2 2 Q=. 3
RE E'+
Basic System Thevenin Eqaivalent
Configuration -.l

Figure 2
Figure 3
Plots of K K K for Different Operat-
2 3 4 in System of Figure 2
ing Conditions
 MULTI-MACHINE EFFECTS where t h el o o p i s opened a tt h ei n p u tt ot h e
b l o c kf o r m i n ge l e c t r i c a lt o r q u e .
From the above development it appears
t h ai tn h e r e nmt a c h i n e damping
of
generation
c l o s et ol o c a l o a dc a nb en e g a t i v e
l o a d i s a l s sou p p l i e d by
remote
generation
i f such
T b [ Y +T
whose a n g l e i s c o n s i d e r a b l yi na d v a n c eo ft h e 3 7 7 T(M1+M2)
a n g l eo fl o c a lg e n e r a t i o n .
The q u e s t i o nt h e na r i s e sa st o why i n - M~M~S’
s t a n c e so fp o o rd a m p i n gh a v eb e e nr a r es i n c e
t h e r ea r e numerous c a s e so fl o c a al n dr e m o t e I t i s e v i d e nf rt o m
expression (2) t h atth e
g e n e r a t i o n . The e x p l a n a t i o n seems t ol i ei n l a r g e rt h et e r m
p a r tw i t ht h ef a c t h a t h er e l a t i v ei n e r t i a (MZD1+M1DZ)
AX
otfhseo u r c edse t e r m i n etshiem p o r t a n coef 3 7 7 T(Ml+MZ) -
damping i n t h e p a r t i c u l a r s o u r c e . F o r i n s t a n c e ,
i ft h ei n e r t i ao ft h er e c e i v i n g endsource is t h eg r e a t e rt h ep h a s em a r g i no ft h ef u n c t i o n
l a r g e c o m p a r e dw i t ht h a to ft h er e m o t eg e n e r a - G(S) actr o s s o v e rh, e n ctehge r e a t e r a itsh e
t i o n , dampingoftheremotegeneration i s of damping of n a t u r a fl r e q u e n c yo s c i l l a t i o n so f
prime importance. On t h eo t h e hr a n di,tfh e t h es y s t e mo fF i g u r e 5.
r e c e i v i n eg n sdo u r c e i s small compared with
t h er e m o t eg e n e r a t i o n ,t h e nt h ed a m p i n go ft h e Examining t h i s e x p r e s s i o n f o r a range o f
i n t e r m a c h i n eo s c i l l a t i o n s i s p r i m a r i l y a f f s c t - p a r a m e t e r as n,pdo s t u l a t i nt gh at ht eo t a l
ed by t h e damping o f t h e r e c e i v i n g e n d s o u r c e . damping e f f o r t D + R z = D and t o t a l s y s t e m i n e r -
t i a M1+MZ = M, t h ee x p r e s s i o nb e c o m e s :
The e f f e c to fd i s t r i b u t i n g damping e f -
f o r t among v a r i o u ss o u r c e sc a nb ei l l u s t r a t e d (M-Ml)Dl+Ml (D-Dl)
w i t ht h es i m p l e two m a c h i n e s i t u a t i o n d e s c r i b - X =- . or
e di nF i g u r e 4 . Damping throughsupplementary
s t a b i l i z i n g i s assumed t o accomplished be MID+ (M- 2M1)D1
t h r o u g hd e v e l o p m e n to fe l e c t r i c a lt o r q u e si n x=-
phasewithspeed.
For a t o t a l amount ofdamping e f f o r t D it is
The v a r i a t i o no fe l e c t r i c a lt o r q u ef o r e v i d e ntth atth e optimum d i s t r i b u t i o n i s t o
e a c h s o u r c e i s t h u s composed of two components, p l a c ea l lo ft h e damping i nt h es m a l l e ri n e r -
onedue t o change i na n g l ea n dt h eo t h e rd u e t i a .T h i sc a nb es e e nf r o m a X / a D = (M-ZMl)/TM
t o change i ns p e e d . The l a t t e r wouldbesuch w h i c hi n d i c a t e st h a t r a d i n go f $ ’ damping e f -
f o r t from “2 t o M 1 r e s u l t s i n g r e a t e r o v e r a l l
as
might be
produced by a ni d e a ls t a b i l i z e r
which moves m a c h i nf el uwxi t hceo r r e c t systemdamplngaslongas
mean t h a t damping a p p l i etddoo ensocto n -
Mz>M . T h i sd o e sn o t
p h a s ec o m p e n s a t i o nt oy i e l d a t o r q u ei np h a s e
w i t hs p e e d . I t s h o u l db en o t e dt h a ts i n c et h i s t r i b u t et oo v e r a l ls y s t e m damping. As a m a t t e r
damping a c t i o n i s a c c o m p l i s h e de l e c t r i c a l l y , o f f a c t , s i n c e t h e amountof t o t a l s y s t e m damp-
there i s couplingbetweensources,that is, the i n g D c a nb el i m i t e d by p h y s i c acl o n s i d e r a -
damping t o r q u e on o n e s o u r c e a c t s w i t h t h e o p - t i o n s if r e s t r i c t e d t o few machines,thegen-
positesignintheothersource. e r apl r a c t i c e would be t od i s t r i b u t e it to
s e v e r am l a c h i n e sN. e v e r t h e l e s s i t i s evident
T h i s i s i nc o n t r a s tw i t ht h ed i a g r a mo f t h a t where M Z > > M 1 i t i s e s s e n t i a lt op r o v i d e
F i g u r e 5 wheredampingoneachmachine i s pro-
duced from themechanical end. The r e l a t i v e
dampingofthesystemofFigure 4 c a nb ee v a l -
u a t e dw i t hf r e q u e n c yr e s p o n s et e c h n i q u e s . One
approach i s t oa n a l y z et h eo p e nl o o pf u n c t i o n

*p12

Apw2
___c
1
x$ -
Figure 4
1-1 377 i Figure 5
Block Diagram of Two Machine System
Block Diagram of Ideal Two Maching System With DBmping Represented as Equivalent
with Damping Provided by Voltage Control. Mechanical Damping.

829
damping on the smaller machine (Dl). It is Twobasicconditionsareillustrated.
therefore possible to have D negatrve without For XE = 1.0 the operating condition (a) with
it significantly affecting tge overall damping P+jQ = 200+j700 can be seen from the plots of
of the system, if MZ>M1. Figure 3 to yield a negative K4. (Note that
the plots of Figure 3are for the machine with-
out amortisseurs.) The other operating condi-
SAMPLE CASES tion (b) chosen for purposes of illustration
is for P+jQ = 50O-jZOO and X ~ s . 5 which yields
A simulation of the system of Figure 6 a positive value ofK4.
was made to explore these damping effects with
the machine treatedinmore detail using a com- Theinherentmachinedampingcan be
prehens've powersystem 'dynamic simulation judged-by a test of dPe/db for the machine in
programt5*6). The machine model used is shown question. Figure 0 shows the response of elec-
on Figure 7 and corresponds a salient pole trical power of machine 2 to a step change in
generator with amortisseurs. (8 angle of machine 2 for these two conditions.
Amortisseur effects were minimizedby reducing
subtransient time constants to .01 sec. Fol-
lowing the initial increase in power there is
XE (IO00 I N A Base) an additional slow increase due to a negative
K4 (magnetizing instead of demagnetizing ef-
I fect) for condition (a) whereas for condition
(b) the initial rise of poweris followed by a
1000 W A 5000 W A decrease due to demagnetizing action (positive
RE=l. 0 K a Figure 9 shows the angle swings which re-
sult for the two operating conditions folldk-
(1000 WA base) ing a small mechanical torque disturbance.
L Effect of Amortisseurs
Machine Constants in Per Unit on Own N V A The undamped conditionof operating point
Base (Yeactances Include Transformer) (a) was used to assess effectsofamortisseurs.
First increasing the d-axis amortisseur time
Tdot = 5.0 constant from .01 to .15 sec. which is equiva-
lent to decreasing d-axis amortisseur resis-
Tdolt = 0.31 tance produces more negative damping in this
U
case aswould be expec,ted since d-axis amortis-
seur losses act in the same sense as field
Zq"
9
= 0.73
= 0.38
Tqo" = 0.01
H = 4.9
losses.
The increase of q-axis amortisseur time
constant on the other hand does produce posi-
Figure 6 tive damping. Figures 10 and 11 show these ef-
SamTle System Simulated With Detailed fects both in terms of the dPe/db responseand
in terms of the damping of the system oscilla-
Machine Representation tions following a 18 step change in mechanical
torque.

-
Note: For dimensional consis-
tency theX parameters in the
block diagram should be labeled
L for
inductances.
unit
per How- I.
ever, with Wo=l.O per Xunit 's
and L's are thesame. -rp

Figure 7
Block Diagram of Salient Pole Generator
 4

Ape 2
Condition (b) P+jQ = 500-j209
XE = 0 . 5
RE - 1.0

I I11

Condition (a) P+jQ = 200+j700

I1
XE
RE - ?
1.0 P e U * 1000 M I A
1.0 p . u . Base

Figure 8
Response of Electrical Powcr to
Step Change in A n g l e for S y s t e m
and Loading Conditions (a) and (b).
I- 2 . t --_._-__
F-
0 .3 sec
Figure 10
(Effect of hortisseurs)
Response of Electrical Power to Step
Change in Angle. P + j Q = 2 0 0 + 7j 0 0 ,
XE = 1.0, RE = 1.0

Figure 11
Effect of Amortisseurs
Response to Step Change in Mechanical
Torque. Condition(a) ?+jQ = 2 0 0 + j 7 0 0 ,
XE = 1.0

831
Effect of Load Characteristics
The linearized coefficients K1 to Kg de-
veloped in Ref. 1 were for the case where the
system was made up of linear impedance ele-
ments. The effect of non-linear load charac-
teristics is also illustrated in Figures 12
and 13. Representing the load as constant cur-
rent for the real part and constant impedance
fortheimaginarypartresultsingreater
damping .
Representing the load as an equivalent
induction motor with the characteristics shown
on Figure 12 also results in improved machine
damping as is illustrated on Figure 13. This
is not surprising as the induction motor load
is frequency sensitive.
It should be noted tha3 these results
only apply to inherent machine damping defined
I
as the damping provided in a machine without
regulator effects. A s was expounded in Ref. 1
the dominant effect on system damping can well
be due to the action of voltage regulators.
The more negative the coefficient Ks (det/d6)
the more pronounced is this effect. When vol-
tage regulator effects are included, the effect
of load characteristics on Kg can predominate tstand,
and cause the net effect on damping to be op- Loa
posite of that determined when considering ma-
chine inherent damping alone. 100

.I2 .012 0 5 sec

Figure 13
Effect of Load Characteristics
base Response to Step Change in Mechanical
Torque. Loading conditions (.a)
P = j Q = 200+j700, XE = 1.0
P~~~~ = 1000

I-

I APeZ Load as constantcurrent

(real part).

IR - 1.0 (1000 WA Effect of Inertia

base)
Case (a) with a constant impedance load,
yields negative damping of machine 2 which is
the lighter machine (1000 MVA machine compared
with SO00 MVA at the sending end). To illus-

- 1.0 (1000 HVA

base)
trate the effect of inertia, i.e. the impor-
tance of having dampingonthe lighter machine,
this case (a) was simulated for a step torque
disturbance, except that now the sending end
machine's inertia was reduced by a factor of
10 times. The ulots of Figure 14 show the ef-
I t---------t fect of this reduction of inertia onthe damp-
0 5 sec ing of inter-machine oscillations. The impor-
tant factor now is the positive damping on the
sending end machine which was made lighter.
The fact that the heavier machine 2 exhibits
negative damping has little influence as the
Fieure 12
system damping is primarily determined by the
damping on the lighter source.
Effect of Load Representation

-
Response of Electrical Power to Step Change in
Angle. Loading conditions (a) P + j Q = Z 0 0 + j 7 0 0
XE 1.0

832
 Sending End Yachine 1 w i t h o p p o s i t ed u et ot h e i ri n f l u e n c eo ft h em a g n i -
I i = 0.49on5000 MVA Base t u d ea n ds i g no fv o l t a g ec h a n g ew i t hc h a n g ei n
a n g l e( c o e f f i c i e n t K g i n F i g u r e 1 ) .
P a r a m e t r isct u d i eotsnh e sqeu e s t i o n s
are n e e d e dt oh e l pd e f i n et h ec o n d i t i o n sw h i c h
l e a d t o d y n a m i ci n s t a b i l i t y , a n dh e l pg u i d et h e
p r o p e ra p p l i c a t i o no fs u p p l e m e n t a r ys t a b i l i z -
t h r o u g he x c i t a t i o nc o n t r o l s .
REFERENCES
F . P . deMelloand C . Concordia, "Concepts
ofSynchronousMachine StabilityasAffec-
t e d by E x c i t a t i o nC o n t r o l , " IEEE Transac-
t i o n s PAES, Vo1.88, No. 4 , A p r i 1l 9 6 9 , p p .
316-329.
M . K . El-Sherbinyand D . M . Mehta, "Dynamic
Sending End Machine 1 w i t h System S t a b i l i t y , P a r t I . I n v e s t i g a t i o no f
H = 4.9on 5000 W A Base t h eE f f e c to fD i f f e r e n tL o a d i n ga n dE x c i -
( F i g u r e 2) t a t i oS
nystems," IEEE T r a n s a c t i o n s PA&S,
Figure 1 4 Vol.92, No. 5 S , e p t . / O c t 1. 9 7 3 p, p 1. 5 3 8 -
1546.
E f f e c to fR e l a t i v eI n e r t i a BetweenSources
w i t hP o s i t i v ea n dN e g a t i v e Damping. S . B . C r a r y . Power
System S t a b i l i t y , Vol.
C o n d i t i o n ( a ) . P+jQ=ZOO+j700, XE = 1 . 0 11. JohnWiley 4 S o n s ,I n c . ,1 9 4 7 , ~ . 161.
RE = 1 . 0 W. Mauricioand S. S e m l y e n , " E f f e c to f Load
C h a r a c t e r i s t i c s on t h e Dynamic S t a b i l i t y
CONCLUSIONS o f Power Systems, " I E E E T r a n s a c t i o n s PA&S,
Power system dynamic s t a b i l i t yp e r f o r - 1972, p p . 2,295-2304.
mance, t h a t i s t h e damping c h a r a c t e r i s t i c s of J.M. U n d r i l l , "Take t h eB l i n d e r s O f f Your
m a c h i neel e c t r o - m e c h a n i c ao ls c i l l a t i o n s , are Computer S t u d i e s , " E l e c t r i c L i g h t 4 Power
a f f e c t e d by a n u m b e r o f f a c t o r s s u c h as c o n t r o l T / D E d i t i o n ,O c t o b e r ,1 9 7 3 .
e f f e c t s , as w e l l as t y p eo fs y s t e ms t r u c t l l r e s
a n do p e r a t i n gc o n d i t i o n s . F . P . deMello, T . E . Kostyniak, R . J . Mills
and J.M. U n d r i l l ," I n t e r a c t i v eC o m p u t a t i o n
I nt h i sp a p e rt h e phenomenaofinherent i n Power S y s t e mA n a l y s i s , " I E E E Proceed-
machinedampingcontributedbyfieldandamor- i n g s ,J u l y1 9 7 4 .
t i s s e u rl o s s e s i s explored. I t i s shown t h a t
f o rc e r t a i ns y s t e ms t r u c t u r e s a n du n d e rc e r - C . C . Young, "The Synchronous Machine,"
t a i no p e r a t i n gc o n d i t i o xt h em a c h i n ei t s e l f , IEEE T u t o r i a lC o u r s eo n Modern Conceptsof
w i t h o urt e g u l a t o er f f e c t sc, a ne x h i b int e g a - Power System Dynamics. IEEE P u b l i c a t i o n
t i v e damping. 70M62-PWR.
I n a multi-machineenvironmenttheim?or-
t a n t damping a c t i o n o ni n d i v i d u a lm a c h i n e s or
groupsofmachines i s r e l a t e dt ot h ei n e r t i a Discussion
ot hf e sme a c h i n erse l a t i vtoeot h egrr o u p s R. H. Park (Fast Load Control Inc., Brewester, Mass.): This paper
a g a i n s tw h i c ht h e s em a c h i n e se x h i b i t modes o f deserves attention in that it is a step in the direction of helping Power
oscillation. Industry transmission system planners to better take into account the
effects of motor loads.
These phenomena a r e o f p a r t i c u l a r s i g n i f - In this area the paper makes reference to the fact that induction
i c a n c e t o s y s t e m s composed o f l a r g e r e m o t e g e n - motor load is frequency sensitive. While true, the extent of dependence
e r a t i o ns u p p l y i n gl o a dc e n t e r sw i t hr e l a t i v e l y is related to the nature of the device that is driven as say whether, for
m o d e s tl o c a lg e n e r a t i o n . The dynamic s t a b i l i t y example, an elevator, or a fan or pump, which being the case, the ques-
p e r f o r m a n c eo fs u c hs y s t e m sc a ns u f f e rd u et o tion arises as to the nature of the driven load that was assumed when
lackofinherentmachinedampinginthesmaller the data of fig. 11 were being computed.
i n e r t i a group' of receiving end machines. bnuscript received September 30,1974.
The a l l o c a t i o n o f damping e f f o r tt h r o u g h
s u p p l e m e n t a r ys t a b i l i z i n gs h o u l dn a t u r a l l yb e
c o n c e n t r a t e d on t h el i g h t e rg r o u po fm a c h i 2 e s F. P. de Mello and T. F. Laskowski. The authors wish to thank Mr. Park
i ns u c hc a s e s . for his interest and comment on the paper.
The driven load characteristic used in the motor representation is
Load c h a r a c t e r i s t i c sa f f e c t machine i n - typical for fans and centrifugal pumps when operating near rated speed.
herentdamping. I t was shown t h a t a l o a dr e p - That is, load torque was represented as a constant times the square of
r e s e n t e da sc o n s t a n tc u r r e n t f o r t h er e a lp a r t motor speed. This representation gives maximum frequency dependency
and c o n s t a n ti m p e d a n c ef o rt h ei m a g i n a r yp a r t of motor power. We add that motorinertia is also important in dynamic
is n o t as s e v e r ei n i t s e f f e c t s o ni n h e r e n t stability studies. As motor inertia is increased, the motor, as a system
machinedamping as a c o n s t a n it m p e d a n c el o a d . load,approachesa constant impedancecharacteristic. The damping
from a given motor and driven load characteristic will increase as the
S i m i l a re f f e c t sa r en o t e df o rt h ec a s eo fi n - natural frequency of system oscillations increases (assuming that con-
d u c t i o nm o t o lr o a dI. nt h er e a l power system stant impedanceloads give best damping in system with voltage
e n v i r o n m e n t ,h o w e v e r ,w h e r ev o l t a g er e g u l a t o r s regulators).
have a d o m i n a n te f f e c t ondamping, t h ee f f e c t
o fs u c hl o a dc h a r a c t e r i s t i c sc a nb ej u s t h e Manuscript received December 27, 1974.
833