Analysis of

Series Capacitor Application Problems

By J. W. BUTLER C. CONCORDIA
ASSOCIATE AIEE ASSOCIATE AIEE

Introduction oscillate or "hunt," under certain circuit conditions. The

tendency to hunt becomes more pronounced as the effec-
N THE past several years series capacitors for the com- tive ratio of line resistance to line reactance increases, and
pensation of line drop in power circuits have found putting in a series capacitor to decrease the reactance ob-
increasing use,'l because improved and automatic viously increases this ratio. Consequently, if the machines
voltage regulation can, in many cases, be obtained more are susceptible to hunting, the condition is further ag-
economically by this method than by any other means. gravated by the addition of capacitors. The case of hunt-
Most of the applications have been entirely successful ing of a single synchronous machine operating from an
in improving the system performance. However in a few infinite bus was investigated several years ago.""12 In
instances unforseen difficulties of an unusual character section II of the present paper, further analytical methods
have been encountered. These difficulties were the oc- are developed involving 2 or more machines which enable
casion of a systematic and detailed analytical studv of the size and location of a capacitor to be so chosen as to
series capacitor performance. This study has been paral- product the least tendency to hunt consistent with proper
leled, as far as possible with laboratory tests and correlated voltage regulation.
with field experience. It is believed that a satisfactory In case 3, under certain circuit conditions an induction
explanation and understanding of the abnormal system motor operating with series capacitors will be self-excited
behavior have now been obtained and that adequate and will generate other voltages of lower than normal fre-
methods for predetermining the system performance have quency and thus produce objectionable voltage fluctu-
been developed, although it is recognized that the investi- ations. Analytical methods are presented in section III,
gation cannot be regarded as absolutely final and complete. which make it possible to determine the effectiveness of
The objects of this paper are to present a survey of some methods for eliminating this trouble.
of the troubles which may be encountered in the applica- It has been found in this investigation that the addition
tion of series capacitors to power circuits, to present cri- of a resistor in shunt with the series capacitor will eliminate
teria for determining the range and probability of abnormal the difficulties of cases 1 and 3, while the synchronous ma-
operation, and to describe methods for minimizing the chine hunting of case 2 is minimized or not obtained by
difficulties. proper choice and location of the capacitor. The shunt
Three types of abnormality have been investigated: resistor is of such a high value as not to interfere with the
capacitor's effective operation as a line drop compensator
1. Distorted and excessively large transformer exciting currents, and the value of the power loss in the resistor necessary to
'due to saturation. ensure complete freedom from these circuit troubles will
2. Hunting of synchronous machines. in general be less than 10 per cent of the resistance loss
3. Self-excitation of induction motors. in the line to be compensated. The proper value of re-
sistance to be used in any given case may be determined
In case 1, the application of voltage to an unloaded or from the analysis presented here. It is concluded, therefore,
very lightly loaded transformer through a series capacitor that the material offered in this paper, will enable prac-
may result in abnormally large and distorted exciting cur- tical and economically sound series capacitor applications
rents. These large currents are not simply transients but to be made with confidence that the operating results will
persist as a steady-state condition. This type of circuit, be in accordance with predictions.
that is, a saturable inductance in series with a capacitor,
has been found to have at least 2, and probably more,
different steady states, depending on how the voltage is Section I-Distorted and Excessively Large
applied. Transformer Exciting Currents
Solutions of the equations for this type of circuit have
been found by the differential analyzer'0 at the Moore A. General
School of Electrical Engineering and are discussed in sec-
tion I. In the usual case of application of voltage to an unloaded
In case 2, interconnected synchronous machines will transformer it is known that if the voltage is applied at or
-
A paper recommended for publication by the AIEE committee on power
~~~~~~~near
the zero point of the voltage wave a high inrush cur-
trans.
mission and distribution. Manuscript submitted June 1, 1937; released for rent may result. This is because the transformer flux
publication June 28, 1937 tends to rise, in the first half cycle, to double its normal
3. W. BUTLBR and C. CONCORDIA are in the engineering division, cenltral station value, with consequent excessive magnetizin cret du
1. For all numbered references, see list at end of paper. to the very high degree of saturation. Stated in another

AUGUST 1937 Butler, Concordks-Series Capacitors 975

 way, at high values of flux the effective inductance of the capacitor. However, as soon as saturation is reached, the
transformer becomes very low. The high inrush current current drawn becomes very large and a voltage sooIn ap-
lasts for several cycles and then the transformer draws pears across the capacitor. The voltage absorbed by the
only its usual low value of exciting current. On the other capacitor prevents the flux from building up to as high a
hand, if the transformer is supplied through a series capaci- value as it would have without the capacitor, so the first
tor an abnormal current may persist in the steady state. half cycle inrush current is somewhat less. Moreover,
That is, the energizing of an unloaded or very lightly as the flux tends to remain offset it does not reach a high
loaded transformer through a series capacitor may result value in the negative direction in the second half cycle. To
in a continuous abnormal flow of exciting current. This the contrary, it may hardly dip below zero. Thus the
current may be equal to or even greater than full load cur- exciting current remains very small (or, the transformer in-
rent and is of an apparently very low frequency, resulting ductance remains very large) for the interval following.
in a badly distorted current wave. Under this condition The charge is therefore left on the capacitor, which simply
of excitation the secondary voltage is also badly distorted, acts as an additional voltage source. This additional
and it likewise is largely composed of a lower frequency voltage tends continually to depress the transformer flux,
component. Oscillographic records of the magnetizing just as in the first half cycle it prevented the flux from
current, the secondary voltage and the voltage across the reaching double its normal value. Thus in the seconid
capacitor when a circuit is in this state of excitation are cycle the flux may not go high enough to saturate the core
shown in figure 1. so no additional charge is put on the capacitor. On th
These abnormal steady-state currents flow when the other hand, as practically no charge has leaked off, the
voltage is applied at or near the zero point of the wave capacitor voltage is still decreasing the flux. The general
just as the high transient occurs for this condition in the effect is as shown in figures 4b and 4c and is that of first
ordinary case without the series capacitor.
This phenomenoni has been reported before"3-" in the
technical press and is sometimes called "ferro-resonanice."
The peculiar circuit behavior depends essentially on the
nonlinear transformer characteristics and not on any
secondary features such as hysteresis, eddy currents, etc.
The nonlinear characteristic of the transformer makes
the mathematical analysis very difficult and to the authors'
knowledge there has been no satisfactory analytical solu-
tion made as yet.
The differential analyzer at the Moore School of Elec-
trical Engineering was used, with the able assistance of
Irven Travis and C. N. Weygandt, to obtain solutions to
the nonlinear differential equations of this circuit. Ap-
pendix A gives the equations of the circuit and the form
into which they were put to adapt them to the analyzer.
The single-phase circuit, as shown by figure 2, was set
up and investigated thoroughly before the 3-phase case
was studied, as it was believed that the essential features
of the phenomenon could be obtained more directly in this
manner. A schematic diagram giving the analyzer set-up
for equations 2a and 3a is shown in figure 3.
The 3-phase circuit was set up as a final check and the
results obtained indicated that the single-phase results were
qualitatively applicable.
Typical curves drawn by the analyzer as solutions to
equations 2a and 3a are shown in figures 4, 5, and 6.
A short simplified explanation of the physics of the
phenomenon will be made here, for the sake of complete-
ness.

B. Qualitative Explanation of Phenomenon

Figure 1. Oscil-
These abnormal currents may be explained as follows: lograms of abnor-
Consider that the voltage is applied to the transformer mal currents and
at the zero point of the wave. During the first part of the voltagesobtained
cycle the flux in the core builds up just as though the when transformer
capacitor were not present, since the current drawn is so is excited through
small that there is no appreciable voltage drop across the series capacitor
976 Butler, Concordia-Series Capacitors ELECTRICAL ENGINERRING
Figure 2. Circuit dia-
gram oF transFormer en-
ergized through a series
capacitor, series impe-
dance, and shunting re- c Y
sistance R E E SIN x t o)
L ~~~~~~~~RcXL2
tending to reduce the offset and then actually to build up
an offset in the negative direction. This process continues,
at a rate proportional to the magnitude of the voltage
trapped on the capacitor, until saturation is reached in the
negative direction. When this occurs a large negative
-current is drawn, discharging the capacitor. This large
negative current may discharge the capacitor partially,
completely, or may put on a charge in the reverse direction.
It is easily seen that, except for a fortuitous combination
of circumstances, the discharge will be either partial or
an overshoot. If it is partial the flux will reach a larger
negative value at the next cycle, and then the capacitor
will receive an opposite charge. This negative charge will,
of course, have the effect of reducing the negative offset
and then building up a positive offset, so that the cycle of
abnormal operation is complete.
Since the rate of decrease of offset is proportional to the
voltage on the capacitor, the apparent frequency of oc-
currence of these large currents (due to saturation being
reached) is necessarily dependent on the voltage applied
and on the angle at which it is applied. For instance, the
larger the initial charge on the capacitor, the sooner it
will build the linkages down to the saturable place on the
negative side, and give a higher frequency appearance to
the exciting current.
In order to demonstrate this property of the circuit,
solutions, as given by the analyzer with different values
of voltage applied at the zero point of the wave, are shown
in figure 4. The periodicity increasing with the applied
voltage is clearly seen. If the voltage is increased suf-
ficiently, the circuit can be made to saturate every half
cycle and results such as those obtained by Suits'7 are then
obtained.
C. Summary of Results Obtained
From the Differential Analyzer
It is felt that the major contribution this investigation
made was that of explaining and clarifying to a certain de-
gree the physics of the phenomenon. In order to appreci-
ate the "goings on" in the circuit, fundamental circuit con-
cepts must be kept in mind, as vector diagrams, resonance,
sustained harmonic solutions, and other linear circuit
concepts fail when nonlinear parameters are employed,
The effect of a damping resistor shunting the capacitor
-was investigated and the results are summarized in figure 7.
In these figures, the region abov7e each curve represents
an unstable or abnormally excited circuit condition; that
is, if a given circuit has avalue of capacitance and shunting
resistance that fall in the area above the corresponding
AUGUST 1937 Butler, Concordict-Series Capacitors
IPTOUTPUT INPUT
TABLE TABLE TABLE
L L _ I
i, t
E SIN (Q,t+d)
LxLl
LGEAR BOX
RATIO w/t
. +,
t
_ [-sN(w+-ri2 d1t !
?f(iiCL-i2)dtl dt
INTEGRATOR INTEGRATOR
Figure 3. Schematic diagram oF differential analyzer connec-
tions For solution of circuit equations
curve and the voltage is applied at or near the zero point
on the wave, the circuit will draw a sustained large dis-
torted magnetizing current. All points below each curve
represent a so-called stable area where the circuit will have
its normal low value of magnetizing current. The various
curves show the effect of changing the line reactance, the
line resistance, and the transformer iron characteristic.
The 2 iron characteristics are shown in figure 8. Curve 1
represents the magnetization curve of a modern trans-
former. Curve 2 was taken arbitrarily to have the current
scale of curve 1 multiplied by 3 merely to show the effect
of a different magnetizing characteristic or of a larger
transformer.
In figure 7 the value of capacitance, to which the left
side of each curve is asymtotic, represents the value of
capacitive reactance that must be used to prevent the
circuit from going into a state of overexcitation without a
damping resistor. Unfortunately these values are so low,
that they can seldom be useful in correcting a bad voltage
condition.
These curves were all determined for normal voltage
applied at the zero point of the wave. In certain cases,
it may be found that a resistance sufficient to insure
normal operation under these conditions will not be low
enough if the system is operated at a different voltage.
For example, figure 5 illustrates a case in which reducing
the voltage caused an abnormal state of excitation.
Curve 1 shows normal operation at normal voltage, while
curve 2 shows the overexcited condition occurring when
0.75 voltage is applied. Both voltages were applied at the
zero point of the wave. This indicates also that in some
cases the application of voltage at the zero point of the
wave may not be the worst condition since in a sense a
shift of angle is partially equivalent to a reduction of
voltage, at least as far as the initial charge given to the
condenser is concerned.
A circuit having a resistance shunting the saturable
inductance, simulating a load being taken off the circuit
977
 ical equations for a general linear circuit, consisting of any
number of meshes, the characteristic determinant is in-
dependent of the magnitude or location of the applied
voltages. It is therefore evident that the form of the tran-
/L=1.7 i 1{11E 1 l 1°l 1:1 I sient response, or the building up or dying away of a tran-

\~1,1 \ , /|sient
llAlE=IO
I.0 11 lll I l 111\ 1} l current, is also independent of where the voltage is
-g1 , -. {il 111 1# 1 , \ ,/ll
{ applied.
1
/ \tV /{, ' o } ,0JX e \1 XThisof reasoning,
\l8t1 1, \ o \X J \\t|o A showing that in linear circuit, the loca- a
'\ 'V l I I 'v1 R A l, lltion
1\ 1 1 / ' l the resistor does not affect the periodicity of the
circuit, combined with the differential results for analyzer
the nonlinear circuit, indicate that this may well be a
general conclusion. It has been known in practice that
intervening or secondary loads prevent this overexcitation
tVC=0.45 U (a) condition, but the effect had not been previously investi-
gated quantitatively.

P=1.sA5 D. Conclusions
11W=0.75 I \ / } 10 An unloaded transformer, operating at normal densities,
Xif exited through a series capacitor of reasonable reactance,
is susceptible to conditions of large distorted magnetizing
-11
-K/ 1w ICcurrents.
\'-11 These abnormal conditions can be restored to
normal by the addition of a suitable shunting resistor to
(b) the capacitor, as indicated by figure 7, or by the locating
of a certain amount of load on the load side of the capaci-
tor.
The resistor shunting the capacitor, introduces addi-
/E=\06A tional losses. However, in general, this loss is less than
-,:i\0.15y \ , \ per
10 { of the normal line loss and this can usually be
f\ >cent
justified by the improvement in voltage regulation. More-
v. -' / t j ' / '' over, since the presence of the abnormal condition depends
"tVC=O-033
\ / \ /on the initial transient, it may be eliminated by shorting
out the capacitor only while the starting switch is being
Figure 4. Effect oF applied voltage on the apparent perio- closed, or by having the resistor connected only during the
dicity of exciting current energizing period. The shorting switch or temporary re-
sistor may readily be arranged so as to function automati-
r1 = 0.049 1/C = 0.214 (a) E = 1.0 cally whenever the abnormal condition tends to develop.
Rc =o Magnetization curve 1 (b) E = 0.75
L =0.197 a=0 (c) E=0.6
Section 11-Hunting of Synchronous Machines
between the transformer and the capacitor, was also in-
vestigated on the differential analyzer. The critical points A. General
obtained with this circuit agreed very well with the points
for the same resistance when it was shunting the capacitor, An electric power system consists in general of a group
indicating that it made little difference where the resistor of interconnected synchronous machines together with
was located. impedance and induction motor loads. Any such system
From figure 2 it may be seen that the 2 locations of the is subject at practically every instant of its existence to
resistor are equivalent to 2 locations of the applied voltage, disturbances caused by changes in load, switching, pul-
the circuit remaining unchanged. Further, in the canon- sating loads, pulsating driving torques, changes in excita-
r 2CURVE 1

tion,e caacto voltae buivlds upA2CRE

g X /RIX fN1\/
antztion,cpcorvlgebld curv 2 \1 J\/1X8\
978 Butler, Concordvia-Series Capacitors ELECTRICAL ENGINEERING
tion, etc. These disturbances tend to set up rotor oscilla-
tions of the synchronous machines, with consequent volt-
age and frequency pulsations. Therefore, for satisfactory
operation, the system constants must be such as to cause
any such oscillations to disappear rapidly if due to a
momentary disturbance, or to remain of small amplitude
if due to a continuous pulsating force. A momentary dis-
turbance will, in general, produce oscillations of the syn-
chronous machines at all of the principal natural frequen-
cies of the system, while a sustained -pulsating force of any
frequency will of course produce machine oscillations of
the same frequency.
B. Effect of Line Resistance
anSeisCpaiac
The inherent damping of the system is usually sufficient
to prevent severe hunting. However, in certain cases the
damping may be either too small to be effective or may
actually be negative. That is, the rotor oscillations may be
amplified rather than damped out. There are present in
all systems 2 conflicting effects. Resistances in the rotor
circuits, some types of load, and mechanical friction will
produce true damping of rotor oscillations, while resistance
in the lines or armature circuits tend to produce amplify-
ing rotor oscillations (Qr negative damping).
In the case of a single synchronous generator connected
to an infinite bus, it has previously been shown11"2 that
the ratio of line resistance to line reactance determines
whether or not there is negative damping, and that if all
rotor circuits except the field winding are neglected, there
is a critical operating angle determined by this ratio above
which the machine is unstable. The critical angle is
always decreased by the presence of rotor windings, so
the method provides a conservative criterion.
If a series capacitor is used in the line the ratio of line
resistance to line reactance is increased, since the reactance
has been reduced. Consequently the tendency toward
negative damping is increased.
C. System of Several Machines
In case of a system consisting of several machines, it is
shown in appendix B that there is a corresponding relation
among the line resistances and reactances which deter-
mines the negative damping and that therefore such damp-
ing will be affected by the insertion of series line capacitors.
However, it cannot be expected now to be a simple re-
sistance-to-reactance ratio; instead the criterion for stable
operation is a ratio of certain combinations of line con-
stants. For example, if the case treated in reference 11 is
extended to include shunt loads taken off the line, the
criterion becomes the ratio of the resistance to the react-
ance components of the transfer impedance between the
synchronous machine and the infinite bus (see appendix
B, equation 8b).
In appendix B a system consisting of 2 synchronous
generators, an infinite bus, and a shunt load (see figure 9)
has been analyzed. This is believed to be sufficiently
general to form a basis for the study of almost any par-
AUGUST 1937 Butler, Concordiii Series Capacitors
2
CURVFS 2
2
CURVESI1
Figure 6. Linkage and capacitor voltage curves with critical
values oF R~
Calu c
Curve 1- Rc = 6.7. Abnormal, Cdpacitor voltage builds up
Curve 2- R, = 4.2. Normal, capacitor voltage decreases
a = 0 ri = 0.197 Magnetization curve 1
L = 0.394 1/C = 0.21
ticular case which may arise. The critical operating angle
of either generator in this general case is shown to be de-
pendent not only on the ratio of combinations of the line
constants, but also on the operating angle of the second
machine and on its relative amplitude and phase of oscil-
lation. The line constants enter the equation as ratios of
the resistance to the reactance components of the several
transfer and driving point impedances and so are correlated
to the criterion developed by Nickle and Pierce."
The equations of appendix B may be used to study any
particular proposed series capacitor application to deter-
mine whether or not it is in the danger zone and, if it is,
to determine the optimum location and maximum per-
missible reactance of the series capacitor consistent with
proper line drop compensation. These equations also
enable the effectiveness of putting additional series react-
ance in certain machine circuits to be evaluated.
D. Discussion of Results
As an example of the method of analysis, a problem in
which the circuit may be represented as in figure 9 has
been worked out. In this case the synchronous generators
at 1 and 4 ordinarily supply the load, line 3 merely acting
as a tie to the large system at 3. However, occasionally
power must be supplied to 4 from 3. Under this condition
the voltage regulation at 4 is rather poor and it is desired
to correct it by the insertion of a series capacitor between
points 3 and 4. As considerable line drop occurs in line 3
the logical place to put the capacitor might seem to be in
that line so as to give loads at 2 some benefit. Unfor-
tunately, when this is done the hunting of machine 1 is
found to be excessive, while if the capacitor is placed in
line 4 hunting is no worse than without the capacitor. The
desired result of correcting the bad voltage condition at
4 is accomplished by putting the capacitor in line 4. How-
ever in this location the additional advantage of correcting
the voltage at the junction is not obtained.
Table I shows the results of calculations of the minimum
permissible load angle of machine 1 for stability, under the
3 conditions of no capacitor, capacitor in line 4, and capaci-
979
tor in line 3, and also for different reactances in branch 4 Table I
and with and without shunt load. It is evident from this
table that the capacitor in 4 causes a much smaller differ- Critical Load
ence from the original condition than it does when in of MachineI
branch 3. Condition (Degrees)
Although a picture can thus be obtained of the relative No capacitor .r2 ..........=
........,
X4 = 0.2, ....r,= 0.03 . 1.38
tendencies toward hunting of several proposed arrange- Capacitor in number 4....r2 = CO, X4
= 0.2, r4 = 0 .05. 1.32
ments, it has not been absolutely determined by calcula- Capacitor in number 3... r2 = X4= 0.2. r4 = 0.05.4.14
No capacitor .7........ .2 = 1, X4
= 0.2, r4 = 0.05... 1.48
tion that one arrangement will operate satisfactorily Capacitor in number 4.... r2 = 1, X4
= 0.2, r4 = 0.05. 1.40
and another will not. This is still a matter of judgment Capacitor in number3.... = 1, X4 = 0.2, r4 = 0.05 3. 94
fortified by a correlation of field experience with the calcu- Capacitor in number 4.... r2 = O, x = 1. 07, r4 = 0.049 . 1.70
lations. Capacitor in number 3 .... r2 = 1.07, r4 = 0.049...........2.46
co, X4 =

The critical angles recorded in table I are the load angles No capacitor.i.n. ..... r2=1, X4 = 1. 07, r4 = 0.049. 1.88
1i-&i' where 6, is found as described in appendix C and a1' is Capacitor in number 3.... r2 = 1, x4 = 1.07, r4 = 0.049...........2.67
the no-load angle of machine 1 referred to the infinite bus r1See=figure
0
9 for circuit.= 0.023
r3 xg = 0.65 X3 = 0.0516
3. The oscillation of machine 4 was neglected (i.e., k = r2 =co or 1 r4 = 0.05 X2 = 0 Xg = 0.20 or 1.07
0), since it was assumed that the machine t eing investigated xc = 0.105 = capacitive reactance
would have the largest amplitude of oscillation. Also it Note: xq and X4 include both line and machine reactanees.
was found that the necessary load angles were only slightly
affected by the operating angle 64 of machine 4. It is
therefore possible to use the simplified formula (8b.1) for modes of vibration corresponding to the 2 principal natural
nearly every case by simply substituting the value of frequencies of the system of figure 9, or to any impressed
Z3 (Z3 = r3 + jx3) paralleled with Z4 for the Z3 of equation forces, can be very simply calculated if the inertia constants;
(8b.2). of the machine are known. This will not usually be neces-
A more exact study of a system may be made by taking sary, however, since sufficient information can be obtained
the amortisseur windings of the machines into account from a study of the system neglecting amortisseur windings
and computing the damping coefficients by the methods and a knowledge of the general effects of such windings.
of reference 25 for every probable mode of vibration. The (See, for example, reference 19.)

8 + 0l
1 T gl97 e0.497 oi gnormal
tion excita
as a function
Section Ill-Self-Excitation of Induction Motors
MAG. c
- 'CURVE~2 _ OF RC and I1/C.
A
A
General
eea
o '] /7
rl "| Area below each It is well known that an induction machine will operate
-- Icurve is normal, as a generator and that such operation depends on the
L Area aove each
supplying of excitation from either a synchronous machine
4 - - - curve s abnorm or a capacitor even though in case of the capacitor the
4 - ! hG \ - operation has not previously been very well understood.
\|\\l | / SMAG. Now consider an induction motor supplied through a
z
D __ w R line containing a series capacitor. Then in addition to the
e \ / / Figure 8 (below) currents flowing due to normal operation of the motor, un-
2t \ tt / r 1 8 Magnetization der certain conditions the motor may act as an induction
curves. Curve 2 is generator of current of lower than normal frequency. This
curve I with cur- low-frequency current is limited only by the impedance of
rent scale multi- the supply circuit at the low frequency, and may reach
o 0.2 .- 0.4 0.6__I pliedby3
pidb relatively large values. These large low-frequency cur-
CAPACITIVE REACTANCE YC rents manifest themselves not only as current surges and
voltage swings but also as strong oscillations of the motor
1.6e rotor caused by the large pulsating torques produced.
1.4 XtiXl X atm00
- --0 lfl25H 1111111 I This phenomenon of self-excitation will not always take

1s.0 ___L2T EjIIlUlllllll 11|111 line or shunt resistance, or by judiciously locating the
Z0.8titittiji902 - ~~~capacitor. It may, moreover, be calculated with reason-
Z0.6tX 0. z;tXz iH ttx 111111W1^ 1111111111able
t 9 tt0 ] ! 11{211 [
certainty and accuracy by equations given in ap-
J] {SlS11]pendix C, if the circuit constants and operating conditions
Figures 10 to 13 show the regions in which self-excita-
0o0 .1 .
PER UNIT CURRENT -PEAK
. tion is likely to occur for the case of a purely induction
motor load. Figures 10 and 11 show the effect of series
980 Butler, Concordia~-Series Capacitors ELECTRICAL ENGINEERING
 Xq Figure 9. Circuit at normal load, a sustained overload may produce self-
El r1 d INFINITE diagram of 2 syn- excitation.
rx BUS chronous ma-
E4 rx E chines connected C. Calculations and Test Results
X2 LOAD to infinite bus and
impedance load In order to check the theory offered for this so-called
"self-excitation phenomenon," tests were conducted on a
miniature set-up, in order to get good control of the vari-
line resistance in preventing self-excitation, while figures ables. A 3-phase 3/4-horsepower wound-rotor induction
12 and 13 show how a resistance in shunt with the capaci- motor was connected to a power source through a series
tor can eliminate the phenomenon. For any practical capacitor and 0.39 per unit external reactance. The
case the reactance of the series capacitor will of course be values of series capacitance were varied over the complete
much less than x', probably about half of x', since x' is range of self-excitation, and the corresponding value of
the short-circuit reactance of the motor including line shunting resistance to insure normal operation was de-
reactance and since the capacitor reactance will usually termined. These tests results, and corresponding calcu-
be no greater than line reactance. Therefore figure 14 lated points on the curve are shown plotted in figure 16.
is given, showing the shunt resistance necessary to pre- The agreement between test and calculated values was
vent self-excitation in the useful range of capacitance, for very gratifying. It is felt that the equations offered can
2 sets of system constants. be used with confidence in predetermining these regions
The values of shunt resistance given in the curves will of abnormal operation.
usually be conservative because the effect of series line The motor was tested running light, hence the speed was
resistances and machine losses, which themselves tend to taken as unity to simplify the calculations. The armature
damp out self-excitation, has been neglected. Equations resistance of the motcr and external circuit was 0.10 per
are given in appendix C which take both shunt and series unit. This high circuit resistance is the reason for the
resistance into account and which should be used to ana- curve in figure 16 intersecting the a = 0 point at such a
lyze any specific problem. These are based on the an- high value of x,. The rotor resistance was likewise very
alysis presented in reference 20. Figure 15 shows an ex- high, hence the motor had a correspondingly small rotor
ample of how the value of a necessary for stability de- time constant. The small rotor time constant gives the
creases with increasing line resistance. motor a greater tendency to self-excite as shown in figures
10 and 12.
B. Discussion of Results In view of the type of motor tested, that is, one having
high resistances, these test results should not be generalized
Figure 10 shows that as long as the capacitive reactance as significant of normal motor performance.
is less than half of the system short-circuit reactance, The line of demarkation between normal operation and
self-excitation will not occur for a normally loaded group of
induction motors. This is because the line resistance is
usually great enough to damp out any tendency to self- 4 - - 1
excite. However, as soon as a certain critical value of - - 1
short-circuit reactance is exceeded the amount of series 1.2 - - | - --
resistance required begins to increase rapidly. This crit- - - - - . - -
ical value of reactance is, of course, a function of the motor -. - - -S ll
characteristics and the impedance of the connected circuit. /x UNSTABLE
If the line is relatively short (i.e., if its reactance is less w
than the motor reactance) it may be compensated more
- /__i.
/
than if it is long, unless a shunt resistor is added to pre- ( 1|
vent self-excitation. 3 06 17.14 { -
Figure 11 shows that as the motor speed decreases the - .
region of self-excitation moves down along the xc ordinate. 04 L
That is, the tendency to self-excite occurs at a lower value
of capacitive reactance at lower motor speeds. Such low 0.2
To= 171.4
speeds occur during starting and during periods of over- UNSTABLE
load. We should thus expect some oscillations as the motor L ' tiL 11 711 I llIi 7T]i
comes up to speed. This iS especially true if a starting 0) 10.4 0.8 1.2 1.6 2.0 2.4 2.87a 32
resistance is- used in the rotor, as figures 10 and 12 indicate 2xt SERIES LINE CAPACITIVE REACTANCE = XC I2x
that an increase of rotor resistance increases the tendency Fgr 0 efectto fidcinmtr iksre
to self-excite. In general it will not be necessary to elimi-
nate these transient oscillations. Actually, in service capacitors
when the system behaves correctly under normal operating Showing series resistdnce required for Stdble operation aS d function of
conditions no trouble during starting has been experienced, the line capacitance and showing effect of varying rotor resistance
If, however, the system does not have very much margin x' =0.4 x = 3.24 To 171.4 and 17.14 X=0.975
AUGUST 1937 Butler, Concordia-Series Cap?acitors 981
self-excitation was very sharp up to x, = 0.6. Calcula- of a found for each value of x,. The current wave was
tions and tests both showed this. Beyond xc = 0.6 the watched through an oscilloscope and self-excitation could
calculated curve became more difficult to determine be- be readily recognized by a change in the wave. In the
cause it seemed to be very broad in the upper regions, lower regions below xC = 0.6, the phenomenon was more
that is, it was difficult by ordinary slide rule methods to violent and critical. A change of a few per cent in the
get a smooth curve, indicating the critical region was not value of shunting resistance R in this region, would be
very sharply defined. Tests bore this point out in the up- sufficient to change the circuit from normal to a definite
per regions. The test points plotted are the largest values state of self-excitation. The region of self-excitation in
the practical and useful range of xc can be readily recog-
nized by calculation since this range is below xc = 0.6.
1.4.- - - - - - Figure 17 is an oscillogram of the current flowing in one
1.4 _
ZI

_ / _E _phase during the motor starting period, and of the subse-

quent flow of current when the circuit is in the state of
1.2 - - - _ .self-excitation. It will be noted that in this case the motor
came up to speed in the normal manner, with normal
e .0 - - .inrush and sustained current. However in several cycles
-J_ - _ X- the current built up to several times normal with obvious
z frequencies, which are identified as the natural
_ .superposed
6 o.e - - -
currents of the system having positive decrement factors.
These currents are ultimately limited in magnitude by
w 0.6 CZ rSTIE TT T1TT TT X l 0 stants
saturation andtothea value builds
currentthat up until the circuit con-
_J
-tt change will not permit further self-
0.4 excitation.

0.2. I
D. Self-Excitation of Synchronous Machines
A B C ID
0 I__ It has been stated above that in circuits containing
0 .0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 series capacitors and induction motors, there may be a
SER IES LINE CAPACITI VE REACTANCE- XC
tendency to self-excite and that in circuits containing series
Figure 11. Self-excitation of induction motors with series capacitors and synchronous machines there may be a
capacitors tendency toward hunting. It is evident, however, that
the tendency to self-excite (that is, to generate amplifying
A--w = 0.2 5 Bow = 0.5 C-or 0m707
= D-or 1e 0 current oscillations even though the rotor inertia prevents
x' = 0.24 x = 3.24 To = 1 71.4 hunting) exists also in synchronous machines. It has been
possible to obtain a satisfactory explanation of operating
1.4 ... experience with synchronous machines on the basis of
| | |
... I l hunting alone but it is conceivable that under certain
1. lconditions purely electrical self-excitation may be en-
countered.
1.o C
t w /W < X <
N I I 11|
While
tF i this
phenomenon
matter has not been thoroughly studied, the
T S1is familiar in relation to the determination of
- - -.I
, I -\ I -} the line charging capacity of synchronous generators.
V0.8 41>1=4+ + t+ < |\| S. B.Crary,2M. Takahashi,22 and Y.H.Ku23 have made
X 17.4 Xa study of this subject and have shown the tendency of
0iUNTAL line resistance to reduce the self-excitation. By means of
the theory given in reference 20, it may further be shown
0
171.4 - - .that a resistance in shunt with the capacitor may also be
0.4>tW 8 used to eliminate the self-excitation (see appendix D).
I.W \ Figure 18 shows, for the same generator as in figure l of
0.2 - -. - reference 21, the effect of shunt resistance.
-+-UNSTABLE - - - - - - Since the effect of rotor circuits in addition to the main
o - ; 1 -I - - - - - field has not yet been determined, no definite conclusions
SERIES LINE CAPACITIVE REACTANCE -Xc can be drawn from this curve, but it is believed worthwhile
to point out the existence of such self-excitation and to
Figure 12. Self-excitation of induction motors with series indicate that it may be corrected exactly as in the case of
capacitors induction motors. The phenomena of self-excitation and
Showing resistance (in shunt with the capacitor) required for stable hnigmyo oreeitsmlaeul n r
operation
as a Fucto ofteln-aaiac n hwnfeto mutually dependent to a limited extent. Their mutulal
varying rotor resistance effects have not yet been studied exactly.
x'= 0.4 x =3.24 X=0.975 r=o The ideal procedure, from a mathematical point of view,
982 Butler, Con.cordia-Series Capacitors ELECTRICAL ENGINEERING
 1.2--.--
1 Figurel14. Self. 60 -- -
1.2- F Texcitation
= W W T TTofTFin -__
tt
t tttttttt
-- Z < duction motors
Xt
iX
ttXt.-8
X e $ WW
X with series capaci- 20 t
tors Lu
12-I-0.-6 _ Showing resistance \
io
8 (in shunt with the 8
0.4 .capacitor) required 6
.~~~~~~~~~\~~~~~~~~~~~~~~ ~~~~for stable operation CURVS_
0o2 ISd function oF the Z4
0.l _
___
-A/B -
i
- - - -
1
j 4 4 4line capacitance For
various systems
x=0.4T0=l
x02-4 T 17

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7. 0.8 0.9 I.0 x = 3 r 0 Z \ x=0.16,To=171
SERIES LINE CAPACITIVE REACTANCE -Xc 9x0.4,To=I7.I_0
co 0.975 ML
Figure 13. SelF-excitation of induction motors with series x' =0.4,To=t71
capacitors 0.6 H CUR'VES _ x'=0.4,To= 17.1
Showing effect of varying motor speed 0.4
A-w = 0.25 B- =, 0. 5 C-cX = 0.707 D-X = 1.0
x' 0.4 = x = 3.24 To = 171.4 r=0 xl=0.6,To=I7I
0.2
0 0.2 0.4 0.6 0.8 1.0
RATIO CcIx'
would be to set up the equations of motion of the circuit
including both electrical and mechanical motions, and
then to find out whether the system as a whole was stable R = resistance in shunt with the capacitor
or unstable. This method brings out very clearly the basic k = 6m
identity of the phenomena of hunting and self-excitation abn
but is, from an engineering point of view, at present im- m, n = subscripts given to currents, voltages, impedances, and
angles in relating them to certain machines and circuits
practicable because of the excessive labor involved in the as used in section II
calculations. It is believed that the results obtained by
the approximate studies are practically the same as would
be found by the unified method. Appendix A-Derivation of Equations
Nomenclature Set Up on the Differential Analyzer
The single-phase circuit analyzed is represented in figure 2. The
Section I magnetization curves shown in figure 8 are the 2 iron characteristics
that were investigated and discussed in the body of the paper.
E = applied voltage, maximum value The differential equations applying to this circuit are:
I = time in radians
a = phase angle of applied voltage E sin (t + a) = r1ii +L- + (il - i2)dt + (la)
ri = line resistance dt C dt
ii = line current 1 C
i2 = current through Rc i2Rc = C J( i - i2)dt (2a)
Re = resistance shunting the capacitor
L = line inductance Rearranging and integrating (la) into a form applicable to the
c = capacitance differential analyzer, there is:
4' = fluxlinkage rr 1 C 1
= voltage across capacitor Li, +4' = ) [E sin (t + a) - rii- C J (i - i2)dtjdt (3a)
Sections II and III The schematic hook-up of the analyzer, giving the solution to
E = internal voltage equations 2a and 3a is shown in figure 3. The input voltage was
ed, e5 = direct and quadrature axis terminal voltages actually not cranked in by an operator. Instead an auxiliary set-up
4'd, 4'.
direct and quadrature axis flux linkages
=
was made that gave the solution to the differential equation
id, ij= direct and quadrature axis currents d2x
Xd, Xq = direct and quadrature axis synchronous reactances dt2 - x (4a)
(section II)
r =
stator line resistance per phase which is a sinusoid, and it was impressed automatically into the
a =angle of voltage E from some reference point hook-up. An operator was needed to crank the current into the
0 =
angular displacement of rotor system by keeping a pointer representing 4', the linkage, on the
XL p6 =rotor speed
= linkage-current characteristic. Values of applied voltage, linkage,
Xc =
line series capacitive reactance capacitor voltage, and total current were plotted by the output
To rotor time constant with open-circuited stator tables.
x = direct or quadrature axis synchronous reactance of induc- The procedure used to determine the points on the resistance-
tion motor including line reactance capacitive reactance curves of figure 7 was as follows. The differen-
x'= direct or quadrature axis transient reactance of induction tial analyzer was started at a point corresponding to application of
motor including line reactance the voltage at the zero point of the wave and the several dependent
a = xc/R variables of the circuit plotted automatically on an output table.

AUGUST 1937 .Butler, Concordia-Series Capacitors 983

 0.12 Figure 1 5. Relation 2 different values of shunt resistance, one just sufficient to produce
C C- - - - - - between line resistance normal operation, the other slightly higher.
0.10 -
_ - and a required to pre-
\ S_ _ ~~~~~ventself-excitation
W
ventsell-excitation
3 = 0. 4 Appendix B-Criteria for Negative Damping of
0.08 X 0.2 Synchronous Generators Neglecting Amortisseur
x,=
_ -^ v
° 0.06 - Figure 16 (below). ~~~~~~~~Winding
- - - - - - +- SeIf-excitation oF in- Consider the circuit of figure 9 with 2 generators connected to an
0.C4 - - duction motors with infinite bus with a shunt load at the branch point and with resistances
- -X- - - - series capacitors and reactances as indicated.
\ Comparison of calculated The effects of salient poles will be considered only in machine
0.02 -t tt
- -- Compdrison o dcauted number 1, the stability of which is being investigated, while machine
tand test resul number 4 is considered to be round rotor. Moreover, the reactance
0 - X-Test points x4 will usually be taken as the transient reactance or some modified
0 0.01 0.02 0.03 0.04 Curve-Calculated value depending on the frequency of the oscillation being studied.
SERIES RESISTANCE - r x' = 1.03 x = 2.32 The steady-state equations for this circuit are:
= 1.0 r = 0.10
O.3C -.-- - ~~~~~~~~~~~~~~~~~~~~~~~~~~ ql + q2 + q3 itq4 = 0
0.30 r z l t |
Xdid1 +X2id2 +riiq + r2iq2 =E
0.25 - X2id2 -X3d3 +r2iq2 -r3iq3 = E3 cos 81
di ± id2 + d3 id4 = 0
0.2C Zq2
~~~~~~~~~~~~~~~~~~~~~~~~~~
+.qq± r2id2 r1idl-~ ~ ~ ~ ~ ~ ~- ~
r2id2 - r3id3 -2Xq2 +X3iq3 = E3 sin 6,
___.15
__ \ ___ - - X2id2 +X4id4 +r2i.2 +r4iq4 = cos ( &-4)
-
r2id2 +r4id4 X2iq2 X4iq4 = E4 sin (8'-
-4)
0.10 ~~~~~~~~~~~~~~
0.05 _ _ If the effect of the amortisseur windings is neglected it has been
shown12 that the load angle below which negative damping of
generator number 1 may occur is accurately given by the condition
0 0.2 0.4 0.6 0.8 1.0 .2 1.4 1.6 1.8 2.0 2.2 24
PER UNIT xc (when the frequency of the mechanical oscillation is small compared
to normal-voltage frequency),
d.
As indicated by the equation above there are several variables which d = 0 (2b)
can be used in the interpretation of the analyzer results. The
quantities which would ordinarily be measured in test of an actual Solving equations lb for idi, we find
circuit are the transformer voltage and current. However, it was 0 0 -1 1 0 1 -1 0
found by experience that the most significant quantities, as far as E ° ° r° X2
the efficient determination of a criterion for proper circuit operation lE 0
X2 - X3 ri r2 0
E3cos81 x2 00 0 r2 -r2 00
was concerned, were the capacitor voltage and transformer flux. 0 1 1 -1 0 0 0 0
Since, as shown in the oscillograms, the current only appears as an Aid1 = 0 -r2 0 0 Xq x2 0 0 (3b)
appreciable magnitude at widely separated intervals of time (e.g., E0sin
8,
E cos(n -6 )
r2 -r, 0 0 -X2 X3 0
figures la and 4c) and, when it does appear, it is likely to bear no x2 0 X4 0 rX2 0 4
obvious relation to its previous value, thus it is very difficult to use E4 sin (81 - 84) r2 0 r4 0 -X2 0 -X4
current as a continuous indication of the behavior of the circuit.
On the other hand, the transformer flux, which bears a direct func- where A is the determinant of the coefficients of the currents in
tional relation to the current as given by the saturation curve, is equations lb.
always of a reasonably large magnitude and of a nearly sinusoidal Differentiating with respect to 8,, there is
wave shape. It appears, as shown in figures 4 to 6, as a sinusoidal
wave having a slowly varying displacement, and the rate of change 0 0 0 0 -1 1 1 -1
and magnitude of this displacement were found to be useful factors 0 x2 0 0 r, r2 0 0
in determining as quickly as possible from a given analyzer run, -E3 sin b1 X2-XS 0 0 r2 -r3 0
whether or not the circuit would approach its proper steady state.
The capacitor voltage was found to be still more useful in this
Ad i °0 1 1 -1 0 0 0 0
dbl 0 -r2 0 0 xq x2 0 0
respect; that is, if one had to judge from the behavior of only one E3 cos Al r2 -r3 0 0 - X2 X3 0
quantity whether the circuit was to approach a normal or abnormal -E4( 1-k)sin(Bl- 64) X2 0 X4 0 r2 0 r4
condition, one would choose the capacitor voltage. This voltage E4( 1-k)cos(8, -84) r2 0 r4 0 -X2 0 -X4
determined by its magnitude the rate of change of flux wave dis- (4b )
placement, by its rate of discharge the effectiveness of the shunt
resistor, by its duration the continuance of the shifting of displace- week= 884 ,le,t1 h al ftemto fmcienme
meat of the flux wave. In general, it was discovered that if ever a
the capacitor voltage increased in magnitude from cycle to cycle 4 to that of machine number 1. For sinusoidal oscillations k may
(it will be understood here that by "cycle" is meant the apparent be taken as the ratio of the amplitude of the machine oscillations.
cycle of the low frequency current pulsations and not that of the Equation 4b, together with equation 2b, may be written
applied voltage), for example, if it ever went higher the second time
than it did the first, the system behavior was abnormal. On the 0 d1 E(nsn8 I4 o ,
other hand, if it continually decreased in magnitude from cycle to 5ld8(+
cycle the circuit operation was sure to be normal. This is illus- E4[M7, sin (8 - 84) + M,, cos (8 - 84)1(1 - k)
trated by figure 6 which shows curves of capacitor voltage taken at (5b.1)

984 Butler, Concordia-Series Capacitors ELECTRICAL ENGINEERING

or I.0 --

'f6,~ + (I - k) - (MS, cos 5~ - M71 sin 54) 0.8 THIS AREA APES FOR FIELD
a -, E
+ ( - t ) (MIf cos 4 M7sin CIRCUIT D OR CLOSED

The relation- ~di,= OJthus

E3 leads to atranscendental equation if04'__ __ _ _ _ _ _ _ __ __ - _ --
tan x., - (52 [_ II

(I0 xg -X2 0 0. 4 0. 8 0.12 0.16 0.20 024 0.28 0.32 0.36

Al31 = 7r2 + 7M -r3 x3 X2 + X, x3 (6b 1)S=(CR4
laX2 X4 0 -rt r4 Figure 18. Regions of self-excitation of synchronous
Ihmyb r4 0 fx2 machine with armature capacitance as affected b4y shunt
In eqai re r. 0I resistance R
X2 + x3 -x8 -ri r2 + -3
-r3
M6,1= -X -X 2xg X2 0 (6b.2) xd=.0 X= 0.6 Xd= 0.30 T0o= ,000 r0=O0
X2 X4 0 r2 C4
r2 r4 0 -X2 -X4
-713r 0 X3 -x2 0 (6b.2) or 0 .0 T ,0
-r3~~~~~~~~~~_ Z33(r1 + rs) + xg(7ax2 -x3r2) + r1(x,x, + 737,) (8b.2
X x .
d
X2 0 ri -r2 Or
X2-r2
x3°
M1= -r 3 0 -X2 -° ( anb.3) Z23(X + x3) + X,(x3x3 + r2r3) + r(,(r,- X2r) (*)

X2 ra 0 x2 he
r2 + r4 -r4 X, X2 + X4 -X4
x2 0 r3 72 ° I22+ X22
X2 -X3 0 r2 -ra The right-hand side of equation 8b.2 is recognized as the ratio of
Ms17= -r2 0 X, X2 0 (6b.4) the real to the imaginary part of the transfer impedance from ma-
r2 ra -X2 X3 chine number 1 to the infinite bus ifsx is used for the machine re-
[r + X4 X4 r4 72 + 74 74 actance. Further, if the shunt load is removed (Z2 = x) it checks
the criterion of reference 11.
In the special case for which the motions of the 2 machines are If the shunt load is removed but machine number 4 is kept, the
the same (k = 1), equations 5b reduce to form of equation 5b is unchanged but equations 6b are reduced to:

tan 5, = _ M-1
MJI
(7b) -r3 x, + xa xabl
- Mai = X4 -rl N (9b. 1)
but this is not likely to be a useful case. If k = 0, there is no sim- r4 xg X4
plification. Xa ri + rs r3
If machine number 4 is eliminated by letting x4 and/or r4 be - MG, = -X4 r -7r4 (9b.2)
infinite, M71, Ms, become negligible compared to M31, Mal, and r4 Xg X4
equations 5b reduce to:
-3 ri r
X2 7r 7 l -Al'71 = r3 xg X I (9b.3)
X2 + X3 r3 r2 +r3 -r4 Xg +X4 -X4 1
tan
rg =
L
__
X (8b.1) - Mg, = -3x3 (9b.4)
Xii ri -r2
+ 74
73
x4 4
i
+q 3
- r2 XI -X2
tr2 + r3 X3 X2 + x3 If in equations 5b E4 = 0, or if x4 and/or r4 = there is,

6, = -tan--'- ( lOb)

which is equation 7b if E4 = 0, and is equation 8b if x4 or r4 = .

If E4 = 0 but x4 and r4 are finite, equation lOb may of course be put
in the form 8b by paralleling Z2 and Z4, i.e., in equation 8b replace
r2 and x2 by the corresponding real and imaginary parts of the
parallel values of Z2 and Z4. This reasoning shows that since tan 5
in (8b.2) equals the ratio of the real and imaginary components of
the transfer impedance from machine number 1 to the infinite bus
(number 3) therefore -MM,,/M3, where M61 and M31 are given by
(6b.1) and (6b.2), must also be the ratio of the real and imaginary
parts of the transfer impedance of machine number 1 to the infinite
bus with Z4 finite. In fact, directly from equation 3b it may be
seen that since:

Figure 1 7. Starting and self-excited current of induction Ai4j = M21E, + M,E3 cos 5, - M6,E3 sin S, +
motor with series capacitance in primary circuit MnE4 cos (a, - 54) - M,E4 sin (51- 64) (llb)
AUGUST 1937 Butler, Concordia-Series Capacitors 985
Then:
-M2,/A is the imaginary part of the driving point admittance
at machine number 1
M311 A is the imaginary part of the transfer admittance from i to 3
M61! A is the corresponding real part
M71/ A is the imaginary part of the transfer admittance from 1 to 4
M81! A is the corresponding real part
It follows that since
real
. = _ *
real
. l ~~~~~~~~(12b)P xop
)maginarYadmittance (imaginary)impedanceT 1xb
we have *
_ M61l = ( imaginary
real ) impedance 1-3
Ms,
and
M81 ( real
__
M71
=
imaginary/impedance 1-4
1(14b)
All of these transfer impedances and admittances are computed for
a direct-axis current in machine number 1. Also, since it is only the
relative magnitude of the components of impedance which enter into
the final criteria, the impedances calculated entirely by means of
xq can be used. That is, even though Xd occurs in A and affects
the true values of the impedances, A does not appear in the results.
The method of extension to more than 2 machines is evident from
equations 2b, 3b, and 4b.
Composite Loads
If the shunt load of figure 9 is a composite load'8 of any kind all
the criteria given here may be used unchanged if r2, x2 are defined
by the relation:
de2
= r2 + jx2
di2
at the load voltage e2.
Appendix C-Self-Excitation of Induction Motors
The object of this appendix is to present criteria to determine the
limts f sablopraton f iducionmotrs ithseres .ol
it s line caapai-.
tors. Under certain conditions of load it is found in practice that
there will exist undamped current pulsations of apparently low
frequency and large magnitude. These pulsations are here con-
sidered to be caused by undamped or negatively damped (ampli-
fying) free electrical oscillations. Thus, to determine whether a
given (electrical) system is stable or unstable the natural currents
are examined. The time variation of these natural currents is
specified by the roots of the characteristic determinant of the system.
If the real part of any one or more of these roots is positive the
system is unstable since then the corresponding component of the
transient current is amplifying rather than decaying and will tend
to increase indefinitely until limited by changes in the circuit caused
by saturation, slowing down or oscillating of the motor rotor, etc.
Thus, it is not necessary actually to solve for the roots but only
to find the signs of their real parts. To find the actual frequency
of the unstable oscillation it is necessary to solve the equation,
which is of the sixth degree (or of the third degree if expressed with
complex coefficients) and has in general 3 conjugate pair of complex
roots or 3 natural frequencies.
Anlyi
For an induction or synchronous machine with capacitance and
resistance in shunt connected to the terminals, there are the equa-
986 Butler, Concord ia-Series Capacitors
tions (see reference 20):
(p + a) [ed - Pd + rid + (L4tq I + Xcid = -[e5--;t + riq -@d
(p + a)[e - Pq + riq - (A'dI + xciq = -w[ed P-d + rid +wkq]
(Ic)
where, for an induction motor with completely symmetric rotor and
stator,
{'d = X(P)id
{f= -X(p)iq(C
(2c)
and
xTp + x
x'Top (3c)
Substituting (2c) in (1c) and rearranging; there is:
(13b) {(p +r}ia) [px(p) ++e)ed
r] + xc-wx(p)} c4
id-< (p +a)x(p) + px(p) +
q= -(p (c)
+eq
co{(p + a)x(p) +px(p) + r}Iid ± (P + ot)[px(p) + r] + X, -
( co 2x(p)}iq = -ed - (p + a))eq
and for constant rotor speed the characteristic determinant (the
denominator of the expression for current as a function of voltage),
A = { (p + a) [px(p) + r ] + xc - c2x(p) 1 2 + @2 { (p + a)x(p) +
px(p) + r }2 (5c)
tita
p, there is:
(Top + 1)2A - A' = p6(x'2T02)
+ P'[(2x'To)(x + ax'T0 + r TO)]
+ p4[(x + ax'To + rTO)2 + (2x'To) X
(ax + W2x'To + r + arTT, + xcTO)]
+ pI[2x'To(x, + w x) + 2(x + ax' To + rTTO) X
(ax + 2x'TO + r + carTo + xcTO) ]
W
+ p2[(ax - Wo2x'To + r + arTo + XCT,)2 +
2(x + ax'To + rTo)(xc -co2X) +
(ax'TTo + 2x + rT, )2c2 + 4W2X'TO,(ax + r)]
(15b) + p[2(ax - C2x'T,O + r + arTo + x,TO) X
(XC- 2x) + 2W2( ax'To + 2x A- rTTO) X
W
(ax + r)] + (xc - w2X)2 + (aX + r)2w'
(6c)
Equation 5c is of the form (A' = a2 + b2) and so may be factored
as [A' = (a + jb)(a - jb)] thus at once reducing the order of the
equations to be solved by half. However, this is accomplished at
the price of introducing complex coefficients. The real parts of the
roots of the 2 factors are identical in pairs; the imaginary parts
eequal and opposite. Thus only one factor need be considered, but
it is found that all methods for determining the character of the
roots or solving the equation require about the same amount of
work as doublmg the order and in effect obtaining equation 6c
a + jb = x'Tp'3 + [ (x + ax'T, + rT,,) + jco2x'T,,]p2 +
[(x,T, - cw2x'T, + ax + r + arT,) + jw(2x + ax'T, +
r To) ]p + [(xc - W2x) + jw( ax + r) I (7c)
with a similar conjugate expression for (a - jb).
It will be observed that equations 6c and 7c are almost symmetrical
in r and a except that ca has sometimes the coefficient x' and some-
times x, which destroys the symmetry. It is therefore expected
that all the conclusions and curves obtained will be very similar if
the effect of shunt and series resistances are investigated separately,
using a as the variable in case of the shunt resistance. Here a
may be thought of as the reciprocal of the time constant of the RC
circuit considered by itself. It is the negative of the coefficient of
time in the exponential solution of the RC circuit equations.
APPROXIMATE CRITERION FOR INSTABILITY
It will be found that a very simple and sufficient condition for
instability is that the coefficient of p in equation 6c be negative.
That is, if this coefficient is negative the system is surely unstable,
ELECTRICAL ENGINEERING
but if it is positive the system is not necessarily stable. Putting Since at low speeds the capacitive reactance at which the maxi-
the coefficient equal to zero there is: mum series resistance or shunt conductance is required #.A.ifted
1 (x + r) toward the actual capacitive reactance likely to be used in the
2 |(X
XC = - + X')u2 -ar a circuit, equation lOc or 8c then becomes a useful criterion. Having
L To the boundary curve of instability at any one speed, those at any
(X-XI)2C,O4 C2 1cr (ax+ r) 2 other speed may then be quickly estimated by the relations llc.
- 2k
)+ 4 L Figures 11 and 13 show the effect of varying the rotor speed as
4

4 To
Wtt2 (axC + r ) l/2
- u2(XXla2 + r2) - 2 (3x + x') ( +
2 To
The 2 values of xc obtained from (8c) for a given system determine
the limits of a region which is theoretically surely unstable.
If a set of circuit constants is selected, leaving r and xC free, then
r is varied and x, computed, a region is outlined as in figure 10 equating
(dashecurv).Thifigueshos functionofvery
(dashed curve). This figure shows r as a function of xc if there iScuv.Te
r as a
determined by equation lOc. Figure 11 shows the effect on the
(8c) value of series resistance required if there is no shunt resistance,
while figure 13 shows a similar curve of a versus xc if r = 0.
Equation lOc also provides a means of determining quickly an
aprxmtepesinfrhevleoarquedoeiiae
instability coprequired
to eliminate
the radical in equation 10c to zero, which locates a point
close to the maximum of the a versus x, or r versus xc boundary

no shunt resistor (a = 0). If the effect of shunt resistance alone is curve. Then
determined, letting r = 0, the dashed curve of figure 12 is obtained. a (x + 3x')r' 1
Equation 8c outlines the unstable region very close to the true ku/max =-4xx' - r' +2 4xx' - _ r'r x
boundary over most of the range of instability. However it indi-
cates that instability lies wholly in the range,
-\// + 3x'2'2+ (4xx' - r")[(X X')2- 4r'2] (12c)

co2x' < x < c 22X (9c) where r' = r/,.

If r = 0, there is
so that if xc < u2x' the system is stable when in fact there is required
a certain finite series or shunt resistance to assure stability. Equa- -a (x-x') (13c.1)
tion 8c is therefore not a good criterion when x, is small as in the ku/max 2V\/xx'
case of series capacitors for line reactance compensation except which occurs at
when the motor is running at low speeds.
xc x + x'
EXACT CRITERION u2 2 (13c.2)
To find the true boundary in the region xc < u2x' the criterion of Similarly if a = 0, there is
Routh24 is used. This is carried out numerically by assigning values
to all of the system constants and testing equations 6c or 7c for (r A (x - x') (13c.3)
stability with various values of line or shunt resistance. The boun- ku/max 2
daries found in this manner are shown in figures 10 and 12 (solid
lines). These figures are calculated for a motor of 0.16 per unit which occurs at the same xc as in equation 13c.2. These last rela-
transient reactance and an external reactance of 0.24. In figures tions determine equations,
10 and 12 there are also plotted boundaries corresponding to opera- -
tion at the same speed but with some external rotor resistance, the r2 = (x for a = 0 (14c.1)
total rotor resistance being 10 times the rotor winding resistance.* 2(x + x')
The effects of series and shunt resistors have been determined sepa- and
rately in order to show as simply as possible the orders of magnitude
of the various quantities involved. However, since there is always ,2 (xf-=x0( )
some line resistance present, the value of a required need not be as xx a = 2 (x + x') for r 0 (14c.2)
large as that given by the curves of figure 12. Figure 15 shows an
example of how the a necessary for stability can be decreased with which are the equations of the loci of the maxima of the r versus x,
increasing line resistance, and has been computed from equation 6c and a versus xc curves as the speed c is varied. Then, if the capaci-
by Routh'S24 method. tive reactance x, in a given circuit is known equations 14c determine
the value of r or a needed to eliminate self-excitation at all speeds.
EFFECT OF ROTOR SPEED However, as mentioned in the body of the paper, it will not in general
be necessary to do this.
Equations 6c or 7c may be used to determine exactly the regions
of unstable operation at any rotor speed. However, the effect of FREQUENCY AND TIME CONSTANT
speed may be more clearly and easily seen from the approximate OF THE AMPLIFYING OSCILLATION
expression (8c). Since the rotor open-circuit time constant is
usually rather large, equation 8c may be further simplified by neg- The frequencies and time constant of all these components of the
T. , whence
lecting terms containing transient currents may of course be found directly by solving equa-
tions 6c or 7c. However, it has been found that within the normal
OCC range of circuit constants only one of these currents is amplifying
a /o: A/r \~

CW)2 2 co + V vJV JJand

so likely to be observed as a large current pulsation. Moreover,
this current is that corresponding usually to a small root of 7c
4__x)2_ r__
Lx3' ft2
I@ r>
uc/2 X3'a\/r\
2 \ \ 1/a\2/r\2
\ X over most of the ranlge of xc. It should therefore be possible to
find the root approximately by neglecting higher powers of p in
~~u}j
- 2 ku/ku(lOc equation 7c whence the solution is

Equation lOc shows that as the speed uo is varied the scales of the (XC- ux) + juo(ax + r)(5)
r versus.-r or a versus xc diagram are simply changed so that Pr [(xc-u13x' + ar)]?0 + aX + r] + ju [(aox' + r)T,, + 2x](
XC is proportional to u> (11 c) A further obvious simplification is obtained by neglecting the
r and a areproportional to u
~~terms in the denominator which do not contain ls0.
If P, = Pr + jq
* f otr esstnc i zro(T =o~ ter i n istbiit. Aloi.T then Pr is the decrement factor and (uo + q) the frequency of the
or X = 0 there is obviously no instability, possibly amplifying oscillation. The frequency is u + q rather than

AUGUST 1937 Butler, Concordia-Series Capacitors 987

q sirj the solution is for id, i. The phase currents are obtained by manner similar to that of reference 21 there is found an approximate
Parik t .9 ptjops formula for the lower region:
ia = id cos 0, etc.
- iq sin + 2 (X--Xe)(Xd' Xc) (5d)
and it may be shown that the component of phase current of fre- + xjxd
quency c- q vanishes identically. or
If the real part Pr of p, is equated to zero (corresponding to the = Xq + Xd /(x Xd)2
Xd (6d)
point at which the oscillation is just neutral) equation 8c is arrived c 2 4 xqd6
at by a different route.
The approximate equation 15c holds only for x0 > W2x' and so in Similarly, an exact formula for the upper region is obtained by
the useful range of capacitance at normal motor speeds, equation 7c replacing Xd' by Xd in equation 5d, whence
must be solved directly. It is still possible to avoid the solution of (Xq-XC)(Xd-Xc)
a cubic equation if it is desired to find the frequency only at points +a = -x (7d)
on the boundaries between the stable and unstable regions, since - XQXd
at such points the root is obviously pure imaginary and if p = jq or
equation 7c may be split up as Xq + Xd /(xq - Xd)2 x0Xdc2 (8d)
a + jb = -[(x + ax'T', +JrTq)g2 + (2x + ax'T, + rT,)q - 2 4
(XC '-co2x) ]- j[x'T,q1 + w2x'T,q2-(xcT.- 2xT,, +
ax + r + arT,)q - w(ax + r)] (16c) References
and the real and imaginary components of (16c) equated separately
to zero. Setsto of system constants on the boundary may then be 1. SERIES CAPACITOR INSTALLATION AT BALLSTON, N. V., E. K. Shelton.
zero.
ofsystem2costants4ot.bundarymytnenD
Sets General Electric Review, volume 31, August 1928, pages 432-4. SERIrES
first found by Routh'S24 method and q then obtained by solving the CAPACITORS FOR TRANSMISSION LINES (BALLSTON, N. Y.), E. K. Shelton. Elec.
quadratic equation a = 0 in equation 16c. Of the 2 real roots thus Engr. of Australia and N. Z., volume 5, October 15, 1928, pages 245-7.
found only one will in general satisfy the cubic equation b = 0 of 2. APPLICATION AND PERFORMANCE OF SERIES CAPACITORS, M. I. Alimansky.
(16c) and this will of course be the proper frequency. For example, General Electric Review, volume 33, November 1930, pages 616-25.
taking circuit constants corresponding to the points xc t=n 0,2, c3.GGeneral
SERIES CAPACITOR IMPROVES 4,150 VOLT LINE OPERATION, W. S. Hill.
Electric Review,volume 36,October 1933,page 461.
c

r = 0.035 in figure 10, the 'frequency is found to be very nearly ~~~~~~~~~~~4.NOt RieS CAPACITORS OrDSRUI FeER C Dudley
WHY NOT SERIES CAPACITORS FZOR DISTRIBUTION FEEDB:RS? C. L. Dd
0.7 times normal (rotor speed). At values of xc > x' the frequency and E. H. Snyder. Electrical World, volume 103, June 30, 1934, pages 942-5.
is always very close to, and slightly less than that corresponding to 5. SERIES CAPACITORS, R. E. Marbury and W. H. Cuttino. Electric Journal
the rotor speed. volume 33, March 1936, pages 145-9.
M. Takahashi25 has an approximate solution of equation 7c for 6. LARGEST CAPACITOR BY INSTANTANEOUS RESPONSE ELIMINATES STUBBORN
FLICKER,Electrical
the case of no shunt resistance and has determined regions of self- Lewis. ORIGINATING IN INDUSTRIAL LOAD OFF7,RESIDENTIAL
World, volume 106, November
FEEDER, J. J.
1936, pages 3488-90.
excltation for this case. His results are in general agreement with
7. SERIES CAPACITOR PROVES ECONOMICAL, A. E. Perreten. Electrical West,
those given here. volume 77, December 1936, page 48.
8. IMPROVED VOLTAGE REGULATION WITH SERIES CAPACITORS, R. C. Buell.
General Electric Review, volume 40, April 1937, pages 174-9.
9. EFFECT OF SERIES CAPACITORS UPON STEADY STATE STABILITY OF POWER
Appendix D-Sel-Excitation SYSTEMS, E. S. Allen and J. L. Cantwell. General Electric Review, volume 33,
of Synchronous Machines May 1930, pages 279-82.
10. DIFFERENTIAL ANALYZER ELIMINATES BRAIN FAG, Irven Travis. Machine
Design, July 1935, pages 15-18.
From equations 15 of reference 20, the equations of a synchronous 11. STABILITY OF SYNCERONOUS MACMINES As AFFECTED By ARMATURE
machine operating at synchronous speed, with only one rotor wind- RESISTANCE, C. A. Nickle and C. A. Pierce. AIEE TRANSACTIONS, volume 49,
ing, with series capacitance and resistance in the armature circuit, June 1930, pages 338-350. Discussion, pages 350-1.
and with additional resistors shunting the capacitors, are: 12. EFFECT OF ARMATURE RESISTANCE UPON HUNTING OF SYNCHRONOUS
MACHINES, C. F. Wagner. AIEE TRANSACTIONS, volume 49, 1930, pages
(p + aX)(ed -POd + rid +=,6) + X,id e -p4q + ri, -{d 1011-26. Also refer to discussion of reference 9.
(p + a)(e. P4Pd P'Pq+ri
(P + )(eq + ri, --bq)
-
'P) + Xciq
Xtq = ed P'qd rid
+PV'd
ed + rId 'Pq - -
(d)
13. LEs RhSEAUX DE TRANSMISSION D'ENERGIE (a book), Jean Fallou.
Pages 390-408.
where 14. P. Boucherot, Comptes Rendus des Sgances, Academie des Sciences, volume
4'd -Xd(P)id T= P +
Xd'T.P + Xd. 170. Paris.
= 1 (2d) 15. Jean Fallou, Revue Ginirale de L'Electricite, volume 19, June 1926, page
987.
.1.
q== *x.(pM =-Xq$q
-Xqr$q -x =
(3d)
16. TMEORY OF ABNORMAL LINE-TO-NBUTRAL TRANSFORMER VOLTAGES,
C. W. LaPierre. AIEE TRANSACTIONS, volume 50, March 1931, pages 328-42.
If (ld) are solved for the currents id, i4 in terms of the applied 17. STUDIeS IN NoN-LINEAR CIRCUITS, C. G. Suits. ALEE TRANSACTIONS,
voltages ed, eq, it is found that the operational common denominator volume 50, June 1931, pages 724-32.
of the 2 formulas for current is: 18. STABILITY COMPOSITE
STEADY STATE OF
ENGINEERING, volume 52, November 1933, pages 987-92.
LOADS, S. B. Crary. ELECTRICAL

A' = P5xqXd To 19. FIELD TESTS TO DETERMINE THE DAMPING CHARACTERISTICS OF SYN
4Xqd+rox d)+2TxX CHRONous GENERAT6RS, F. A. Hamilton, Jr. AIEE TRANSACTIONS, volume
+ p4[2xqXd' + XcXq+ Xd') + 2a XXoxrxd
1[(2XqXd'
] 51, September 1932, pages 775-9.
Two THEORY OF SYNCHRONOUS MACHINES WITH ANY BALANCED
+ P + Xe Xq + Xd } ) 7'e + r(xd + Xq + ri'0) +
20.
a(2xdXq + Lxaxqd' To) + 2aerTe(xq + Xd )] TERMINAL REACTION
IMPE:DANCE, C. Concordia. Submitted for publication.
+r P1I2XdXqz + XC(Xd + xqz) + r( To4 2xc + xq + Xd' I + r) + 21. Two REACTION THEORY OF SYNCERONOUS MACEINES, S. B. Crary. ELEC-
a(C ToXe Xq + Xd' } + 2 TeXqXd' + aFXdXq) + ar(2xd + TRICAL ENGINEERING (ALEE TRANSACTIONS), January 1937.
2Xq + 2 Tar + lisa { xq + Xd' )1] 22. TRANSIENT PEENOMENA OF AN ALTERNATOR UPON CONDENSI'JE LOAD,
+ P [ To(Xc -xq)(xc -Xd') + rt(xd + Xq + 2xc + r TO) + M. Takahashi. Researches of the Electrotechnical Laboratory, Tokyo, Japan,
c4x , Xd+X XX iqdT) + s(xTe+2 23. TRANSIENT ANALYSIS OF ALTERNATING CURRENT MACRINERY, V. H. KU.

ctIIXd+tXQA + err?0O Journal of Electrical Engineering, China, volume 3, number 4, AugUSt 1932,
+ [(Xd - Xc)(x - xc) + r1 + c11XdX5 + 2arxc + (ar)1] (4d) pages 179-206.
If the same generator as in reference 21 iS taken, with r = 0, and 24. ADVANCED RIGID DYNAMICS (a book), E. J. Routh. Page 168.
the stability of the system as affected by a is studied by means of 25. Two REACTION THEEORY OF SYNCHRONOUS MACMINE:S, R. H. Park. Part I,
Routh's14 criterion, the results of figure 18 are obtained. Also in a ACTIONS, VOlUme 52, 1933, pages 352-5.

988 Butler, Concordia-Series Capacitors ELECTRICAL ENGINEERING

 D.N. Wdker C.E.J. Bowler ELL Jrkron D.A.Hodgea
GeneralElectric Co. Southern CaliforniaEdison Co.
Scheneetdy, N.Y. Rosemead, California
ABsrRACT condition cawad the occiurtionr to gr6w insbad of decay r e d t i i
The d t
s of the sulmynchronoucl m om ce teat conducted on inrteatiy~inttabilifywhichinturncaumdthedamagetothe
the number 2 turbinegenerator at the Southern Cdifomia Ediron Mohm maemtor.excitar hafts.
Company Mohave Generating W o n are prwented and dsicwed. To better undentandtheproblem a teat was conducted on
M o h unit No. 2 to meam the torsional dynamic characteristics
C o m p d o n a are made between memured and calculated tonional
~ t u n irequencies
l and mode shhapeq and the damping d etd of the mechmicdd- system. A li& of the mort importanttert
with each torsiond mode stimulated b presented at variour lord* In objectivesm given below:
addition, the methodr of &imul.tingthe mechanical tordonab of the 1)Mescure the totriodri natural frequencierand correspondingmode
turbine-generator set are dtcumed along with a deecription of the &4)=
instrumentationsyehm that W I used to obhin thed a h 2)Measure the modaldamping valuer of the t u r b i i t o r at
aeverrl loads.
3)conelate the m e s u m l shaft strain values to mewred vibratory
INTRODUCTION rotational dbplacements g
in&
4)Correkte measured and calculated mechanical respom of the
Subsynchronour Fkomce t u r b i i e m t o r to imposed electrical transients in the
tnnanhion linea
Thephenomenon of subsynchronous m nance on alternating
current power syetem wm Tirat treated in the technical literature tu
early as 1937.1 At lee& five more papen, dwelt on the subject up POWER sY!3" CONFIGURATION
through 1948.2131496~6Two additional papemdealingspecificaUy
with the phenomenon werepublished in 1970.'+* All of these
treated the subject as strictly an electrical phenomenon. It was not
until two shaft failures occurredat theMohave Generating Station in
-
General
The generating station is comprised of two 909 MVA
Southern Nevada that the theory of interaction between a Beries cmwompound units. The h & pressure (HP)generators ate two
compensated tmsmimion syehm inelectricalresonance, and a pole, 483 MVA, 22-KV machines.These are drivenby HP and
turbin-nerator
mechanical
stem torsional
in mechanical intermediate pressure t u r b i i ~8 shown inFa.1.
resonance, wan developed.gJ0~lf~2918P STRAH
The failures occurred in the shaft section between the generator
and exciter at the main generator collector due to torsional fatigue. TELEMETRY all
The mechanical strain cyclingwhichinvolved plastic deformations
caused the shaft to heat up to temperatures which resulted in the
breakdown of theinsulation between thecollector rings and the
shaft. The heavy current flow that resulted from the positiveand
negative Onerator fiild short c h i t s eroded large pockets of metal
from the shaft and the collector ring. Analysis of linecurrent HIGH PRESSURE (HPjjGOO R P M UNIT 483 MVA 4.fMvA
oscillograms taken during the disturbance on the line indicated the A8 A8 Ab
presence of appreciablecurrents of suboynchronous(leeo than 60 H z )
frequency. In general, these subsynchronous currents do not cause
any harm to the electrical system. They do, however, flow in the II II 1
generator armature and react with the main flux of the generator to GENERATOR
develop a pulsating torque on the rotor system at the difference or Y
'f 1 Fyc I
slip frequency between the synchronously rotating main generator FLEXIBLE~ 4iiiVA
flux and thesubsynchronouscurrenta flowing inthe electrical coupiiffi
network. The slip fkequencyfollowing thedisturbance which LOW PRESSURE (LP) 1800RPM UNIT 426 MVA
precipitatedthe Mohave failures verynearly coincidedwiththe
second flexible tolsional natural frequency of the turbinegenerator Fig. 1. Mohow Generating Station 909 MVA Crow Compound Units
rotor system, thus effectively amplifying the magnitude of the shaft
response torquea For this mode of vibration, maximum twist occurs The lowpressure (LP) generators are four pole, 426 MVA
in the h a f t span between the generator and exciter.
The machines each driven by LP turbii. The cross-compound units are
turbinegenerator rotor motion that was produced caused voltages to connected to the Mohave 5oOgV bus through 825 MVA, 16351,
be generated in the armaturecircuits, and the electrical currentsthat 525/22-kV transformer b w . The generating station and ita
these voltages produced reinforced the original subsynchronous associated traasmission was integrated into the 500kV system that
currents. The resulting armature current produced by the Mohave existed between Four Cornersand h g o , see Fig. 2.
turbinegenerator rotor oscillations was largeenough b produce The Eldorado-Lug0 5 W V line W B ~already designed with 7096
surricient torque to sustain the motion of the rotor that caused this series compensation.TheSouthern California E d h n Company's
component of armature current in the T i t place. At Mohave this share of Four Corners generationis 760 megawatts. Stability criteria
for the tawsmixion of this 760 MW plus their 840 MW share of
Mohavepower dictated the 70% series compensation in both the
Eldoradu-Lug0 and Mohave-Lugo 5oOgV lines. The 60 mile 5WkV
line to Eldorado Substation is uncompensated. Each of the series
capacitor banksie rated 36.72 ohms, 1600 amperes continuous.Each
is composed of four 9.18 ohm modules in series per phase. Each of
theae modules is equippedwith a triggered protective gap and a
vacuum bypass breaker for high speed reinsertion. In addition to its
automaticfunction,the vacuumbypass breaker is capable of
F'aper T 75 176-3, recommended and approved by the IEEE Rotating
manuallremoteoperation for b m i n g or insertii each module.
Machinery Committee of the IEEE Power En@eering Society for presentation at Another integral part of the stabiility criteria design for this system
the IEEE PES Winter Meeting, New York, N.Y., January 26-31,1975.Manuscript was the use of hi& initial response, Althyrex excitation systems on
submitted September 3, 1974; made available for printing November 22, 1974. the Mohave generators.

1878
 LUG0 500 kV LINES. EACH -I SYMBOL
REPRESENTS 6.75% OF THE TOTALLINE
REACTANCE.

c@ UNIT 2

PHOENIX

Fig 2. Schematic Dicrgmm of Transmission System into Which the Mdurve Generating Station was Connected

During Test Fig. 1shows the mechanical test arrangements for the HP and LP
It was knownfromexperience and system analysis thatthe machines of unit 2 including the rotational displacement and strain
Mohave units were subjected to the greatest subsynchronous hazard mewrement lomtiom. Fig. 3 shows the electricaltest configuration
when the MohaveEldorado 5OOkV line was open. Under these for the HP machine of unit 2, which was identical to that for theLP
conditions, the critical number of series capacitor modules per phase machine of unit 2.
in the MohaveLugo line was known to be men. Because safety of Determination of shaft torsional hquencies, damping and mode
the Mohave generating units w s of prime importance, the shapes required accurate measurement of torsional oscillation. Two
LugddohaveEldorado 5OOkV loop was retained inhctfor the methods were employed to obtain this information. Shaft torsional
entire test and no more than four series capacitor modules per phase strain w s measured diRctly on the exciter shaftof the HP machine,
were wed in the Mohave-Lug0 line. At three points on each shaft, instantaneous rotational velocity was
measured. This w s processed as described below to obtain angular
displacement due to torsional oscillations. Coxrelations between
DESCRIPTION OF TEST INSTRUMENTATION these two measurements were conducted to confirm the validity of
using rotational
displacement
measurement to calculate
shaft
torques.
General Considerations Toadequately d e& the forcing functioninbothsinusoidal
A number of criteria governed the selection of parameters to be excitation tests and transient events, the several electrical quantities
mewled. shown in Table I were measured. Tbese signals were p d to
obtain the subsgnchronous components.
1.Sufficient information to satisfy all test objectivesmust be
obtained.
2. Safety and integrity of theturbinegenerator andaesociated
equipment was of prime importance.
3. The instrumentation techniques employed were to have no effect
on normal station instnunentation.
4. Installation and operation of test equipment shouldhave no effect
on scheduled operationof the turbinegenerator.
Table 1 lists the parameterswhich were measured.
TABLE I
Parameter No. of Channels
HPshaft
exciter torsional strain 2
HP rotational
machine
displacement 3 Fig 3. Illushation of Method U s d to
LP machine rotational displacement 3 Sepamtely Excite Torsional Frequencies
HP&LP generator field current (If) 2 Shaft Strain Measurement
HP&LP generatorvoltage
field (Vi) 2
HP&LP generator
armature
current (Ia) 4 The strain gages mounted on the shaft between the HP generator
HP&LP generator power 2 and exciter weae to provide a precise monitor of shaft strains during
Generator tenninal voltage (q) 2 the tests and to confirm calculated relationship between rotational
displacementmeasurements and shaft torques. Measurements
involving transducers mounted on rotatingshafts require special

1879
techniques for connecting leads to stationary instrumentation. Slip
rings are often used, hut they generally must be mounted at theend
of the shaft and the transducer leads run to them. In this case, the
measurement hardwarewas confined to a small section of shaft
between two rotors, so it wasnecessary to use a radio telemetry
system as a wireless data link. Two strain channels were used in this
application to increase the reliability of the measurementand permit
two strain ranges The low sensitivity channelcovered the maximum
strain encountered in the test and the high sensitivity channel gave
improved signal-to-noise ratio
for
the lower
level strains. A
conventional 4 active arm strain. gage bridgewired to measure
torsional strain andcancel bending strains wasinstalled on the
exciter shaft. The remaining shaft mounted hardware was packaged
in a titanium ring, fabricated in two halves and clamped around the
shaft next to the strain gages. strain gage leads were run to their
respective transmitterconnections wrapped
and with
fiberglass-epoxyband.Eachhalf of the titanium ring contained a
transmitter, 2 battery modules, a “g”-switch, and an interconnection
circuit board.The batteries had sufficientcapacity to power the
systemfor 6 days of cumulative operation.The “g”-switch
connected the battery to the transmitter when the turbine-generator
reached a pre-selected speed, generally about 109% of rated speed.
“hiis feature conserved battery life during those periods when the
machine was not operating. The stationary equipment forthe system
consists of the receiving antenna and receiverdemodulator.The
antenna is a continuous loop device, mounted around the shaft in
close proximity to the transmitters. The receiverdemodulator
produces a signal proportional to the bridge output, i.e., torsional
Strain.
Rotational Displacement Measurement
Because the measurement of shaft torsional
with
radio
telemetry
equipment is complex
and costly,
the
rotational
displacement measurementsdescribedbelowwere used as the
primary indicator of shaft torsional oscillation. A magnetic pick-up is
located next to a shaft mounted toothed wheel or gear as shown in
Fig. 4. For aconstantshaftrotational speed, S revolutionsper
second, and N teeth on the toothed wheel or gear, the fundamental
frequency component &om the pickup will be N x S Hz. If the shaft
oscillates torsionally, this valuewillbe frequency modulated. It is
necessary to demodulate the signal prior to recording to obtain the
torsional oscillation information in its direct form. A phase locked
loop discriminatorw a designed to accomplish this function.
Fig. 4. Toothed Wheel
Generator Electrical Parameters
Measurement of generator output current andvoltageandfield
current andvoltage
required current
traasformers,
potential
transformers, current probes and voltage attenuators respectively, to
convert these large signals to suitable instrumentation s ig
n& They
were then processed to obtain signals representing subsynchronous
oscillation components.Generator powerwas computedfromthe
current andvoltagesignalsandsimilarlyprocessed to obtainthe
subsynchronous components.
Data Recording and Monitoring
In each of the foregoing measurement descriptions, avoltage was
obtained representing the quantity beingmeasured.Each of t h e ,
was recorded on magnetic tape for permanent record and later analy-
sis.Many of them were monitored during the testwith oscillosmpes
or oscillograph chart recorders. Fig. 5 shows some of the recording
=and monitoring equipmentat thetest site.
a
Fig. 6. Test Site Equipment
‘Ihe monitoring of data during testing wasrequired for several
reasons. First, the test procedure required setting forcing functions
to achieve a specifiedvalue of machine response whichrequired
real-time observation of the shaft strain and rotational displacement
signals. To ensure safety of the machine, maximum values for each
parameter weredefinedand a minimum of one of each type
measurement was continuously monitored during testing. Several of
the signals were incorporated into an automatic trip circuit where the
limits were preset. If the machine response were to exceed the limit
values for any parameter, the trip circuit automatically removed the
test oscillator forcing function. An additional function of the chart
recordingswas to provide sufficientdatafor preliminary o n d e
analysis of mechanical damping and mode shapes.
METHOD OF TESTING
Methods and Procedures
Implementing test procedures to precisely me8w1R the torsional
mechanical frequencies and damping as a function of electrical load
on the turbinegenerator was quite simple in concept, but required
utmost care to ensure that no damage was incurred by the
turbinegenerator. Tbe following is description
a of the tests
conducted.
The requirement of excitingthenaturaltorsionalmodes of
vibration of the turbine-generatorwas met with fourforms of
stimulation. The f i i t test with the turbinegenerator set stationary
was engagement and disengagement of the turning gear. The turning
gear is provided to slowly rotate the turbinegenerator when hot to
eliminaterotor bowingand theresultant balanceproblems.The
turning gear is physically located between the generator and turbine.
The force transmitted torsionally dueto gear engagementproduces a
general multi-modal responsewhich can bemeasured using strain
gages on the shaft. Frequency analysis of the strain signals served to
identifythe natural frequencies that were stimulated.Thestrain
responses filtered through band pass filters at the measured torsional
frequencies provided a measure of the damping of eachmode
stimulateddue to strain energy dissipationdue to mechanical
hysteresis and bearingfriction.
The second test to provide a torsional stimuluswas that produced
by manual synchronization of the generator to the power system.
The mall mismatch of angle between the incoming and running bus
voltages provides an impulsive force to the generator to produce a
response of thetorsionalsystem which is somewhatdifferent in
modal content from that produced by turning gear jogs due to the
different axial location of the forcing function.
The third method of providing a torsional stimulus and by far the
most accurate is that produced by sinusoidal variation of the exciter
power output.This test allows accuratedetermination of the
torsional frequencies,damping,and mode shape as afunctionof
turbinegenerator output power.Fig. 3 is aschematic of the test
system providing a general picture of the essential ingredients of this
test. The excitationsystem supplied on the Mohave units is the
Althyrex system. This system embodies rotating
a 4.7 MVA
synchronous machine mechanically coupled to the main generator,
whose output is fed to a full wave three-phase thyristor controlled
rectifier system. The thyristor’s ruing is controlled by the voltage
regulator whose output is derived from the setting of the voltage
1880
reference and theerrorbetweenthissetpointandthe sensed
terminal voltage.
Sinusoidal forcing of exciterandgeneratorpower output was
obtained by enteringa signal froma voltage controlledoscillator
(VCO) at the summing junctionof the regulator. The VCO produces
a variable frequency, constant amplitudeoutput, with sweeprate and
sweep widthadjustablefor each test condition.The VCO was
adjusted to produce a sweep of output frequency of about t0.5 Hz
around a torsional frequency deduced from spectral analysis of the
earlierperformedturning gearand synchronization events. The
sweep rate was adjusted low enough t o e m r e that no SignXicant
resonant amplitude or frequency distortions were obtained by the
rate being too high. The s u c w of this test was predicated on the
ability to use the exciter generator as a wide-band, large, dynamic
range power amplXer providing torsional excitationin the frequency
range of 10 Hz to 60 Hz.
This excitation system meets these requirements as it is a high
initial response system whoseresponse time is onlylimited by
transport delay in fiing the thyristor. Having a full wave thyristor
This sample clearly
shows that
the mechanical system is
responding principally in two of its normal modes 88 indicated by
thebeatfrequency of about 3.3 Hz which corresponds to the
difference between the f i i two flauile torsional modes of the
system. It also indicates the lightness of the mechanical damping
present as there was no appreciableattenuation of the strain
amplitude over a period of five seconds of which Fig. 6 is part.
The frequency analysis of the strain response shown in Fig. 6,
which gives the relative amplitude of the strain signal as a function
of frequency, gave peaks at 26.75 Hz and 30 Hz which
corresponded to the first two flexible torsionalnatural frequencies
of the H p machine. Fig. 7 shows the results of a frequency analysis
of the strain signal for the HP machine following synchronization
of the generator to the line.
I
!
I
I
I I

1
bridgeprovided thecapability of full exciterpower output both
pogitive and negative due to the transient inverting capability. The
objective of sinusoidalforcing is to set up puremodaltorsional
vibrations of aconstantamplitude,then remove the excitation
source at a zero crossing and record the decay of shaft strains and 60
rotational displacementg The rate of decay and the instantaneous 0 IO 20 30
magnitude and phssebetweenmeasurementsatdifferentshaft FREP-Hz
locations provides torsional damping constant and mode shape data
respectively. Fig. 7. Synchronization Frequency Amly&
The fourth method of introducing a torsional disturbance is that
produced by insertion and bypass of transmission line compensating
capacitors.
The
accurate
determination of thedynamics of It is seen that in addition to the identification of the first two
interaction between theelectricalpowersystem and the torsional torsional natural frequencies at 26.7 Hz and 30.1 Hz, the third
mechanical system is vital forcalibration of mathematical models torsionalnaturalfrequency was also identirled at 56.1 Hz. By
used in the prediction of the onset of a subsynchronous resonance similar
analysis of rotational
displacement signals on the
event. The measurements taken during a capacitor switching event l8oORpM L2 machine following synchronization of unit 2 to the
provide a measure of the mechanical response
and also the line, the f i t two mechanical torsionals were identified as 10.3
amplitude,
frequency and damping associated with the and 17.9 Hz.
subsynchronous cunents in the power system. The forcing of the The calculated torsional natural frequencies for the HP machine
mechanical system due to capacitor switching approximates a step using a lumped inertia model are compared below withthe
function of torque on the generator as far as the torsional responses memred values. It was found that great careneeded to be
are concerned. A sudden change of systemreactance causes the exercisedin thepreparation of the mathematical model if
generator to accelerate or decelerate to a new angular position with accuracies better than 3% are to be attained.
respect to the synchronously rotating reference frame. This would
include oscillationsat about 1-2 H z dependent on the total inertiaof
the system and the electrical stiffness. The Value
insertionCalculated Value
of capacitance Measured
and the chargingenergy requirementproduces an oscillationin
transmission system energy storage causing subsynchronous currents
to flow intothe generator and produceatorque at afrequency ~

controlled by the total 29.8

system LC product. 30.07,30.16
These four methods were incorporated into a test procedure to 55.6
enable isolation of the torsional mechanical parameters as a function
of turbine generatorloading.
Torsionalnaturalfrequencies were also obtainedbyidentifying
the resonance peaksinthesinusoidalexcitation syeep testing
DISCUSSION OF TEST RESULTS conductedat several electrical loads.
Within measurement
tolerances, thefrequenciesobtained were the same as those
a) Torsional Natural Frequencies obtained from the synchronizationand the turning gear jog events,
The toGsional natural frequencies were first identified during the and were found to be independent of load.
test by analyzing the torsional rotational displacement and shaft
strain signals following transient torques applied to the
b) Mechanical Damping
turbinegenerator due to engagement of the turning gear and the The mechanicaldampingassociated with the f i i t three flexible
synchronization of the generatorto the line.
Fig. 6 shows a sample torsion modes for the HP and the f i t two modes for the LP
of datafroma strain gage mounted on theshaft between the machine were obtained from the sinusoidal excitation decay tests
generator and exciter of the HP 360" machinefollowing and from analysis of the mechanical signals following
engagement of the turning gear. synchronization of thegenerator to the line and capacitor
switching at several electrical loads. In the sinusoidal excitation
TUR)IING GEAR JOG EVENT decay tests conducted at several fiied loads, the turbine-generator
EXCITER STRAIN SIGNAL was stimulated at each torsionalnaturalfrequencyin turn as
I SECOND describedpreviously. Aftersufficienttime hadelapsed forthe
mechanical system to settle down to steady state conditions, the
sinusoidal drivingsignalwasremovedand the decay of the
mechanical oscillations recorded. As the driving signal was set as
close as possible to each torsional natural frequency in turn the
mechanical systemrespondedin verynearly apuremode of
vibration in each case. The damping values measured, therefore,
Fig. 6 Tmnsient Decay corresponded veryclosely to thedamping associated with each
1881
 normal mode and are referred to hereafter as the modal damping EXCITER STRAM
values. Most modem vibration response programs use these modal ta% FULL LOAD
damping values directly and the task of deducing the individual
dashpot constants between shaft spans and from rotors to ground
from the measured modal values even if it were poaxible, is not
required.
Fig. 8 shows a typical set of decays of the mechanical signals in
the f -torsion mode for the HP machine at 18% of net full load.
EXCITER STRAIN

-
0 77% FULL LOAD
rtd
=
zs- s

YZ

Fig. 9. Fir& Torsion Mode Decays

the synchronization and capacitor switching events through a band

pass filter to isolatethe decay of eachmode inturn. This
also served to give damping dataat different loadpoints for modes
that could be stimulated sinusoidally as shown in Fig. 10 which
summpiza the damping BB a function of load for the f iithree
modes of the HP machine. Fig. 11shows a typicalHP t h i mode
decay envelope that was obtained from the turning gear location
bybandpass fiitering the rotationaldisplacement signal atthe
thirdtorsionalnaturalfrequency and plottingthe peak value
following synchronization of the generator to the line. The relative
amplitudes are plottedindecibels andhence, the slope of the
envelope is proportional to the damping present 88 shown in
Appendix A.
0.016r I I I I I I I

0.0 I4
"0.0 I 2
%.OlO
80.008
-I 0.006
0.004
0.002

m ~ l , ' l I l l , , I 1
% FULL LOAD
Fig. 8. First Torsion Mode Decay -
Fig. 10. Modal Damping Load

The bottom trace on the charts is the sinusoidal driving signal and
the time at which thii signal is turned off which conesponds to
the start of the decays may be idenM1ed. The top threetraces are
the decays of the oscillatory
component of rotational
displacement atthefmnt standard,turning gear andexciter
locations respectively. The signals were obtained by processing the
output from amagnetic pickup lookingat themotion of a toothed
wheel or gear withaphasedetector to yield theoscillatory
rotationaldisplacement at eachlocation. The fourth and fifth
traces are the oscillatory component of strain measured with a
strain gage on the shaft span between the generator and exciter
and the fluctuation of generator fieldvoltage,respectively. The
reason that the signals at the exciter location were superior to Fig. 11. HP Mode 3 Decay Envelope Following Synchronization
thoae measured at the turning gear and front standard locations,is
that in the first tomion mode the exciter motion is significantly
higher than at the other locations and hence the signal to noise c) Sinusoidal Torsional Response Characteristicsof the Rotor System
ratio was correspondingly better.
Fig. 9 shows the effect of load on the vibration decay rates in the Mode Shapes
first torsion mode for the HP machine. It shows that there is a Fig. 12 shows the calculated torsional normal mode shapes for the
marked increase in damping with anincrease in load, however, the HF' rotor system. The circled pointsrepresentthe measured
damping even near full load on the HP machine is extremely light rotational displacement signals under steady state conditions due
correspondingto a system magnification factor of over 200. to sinusoidal variations in the exciter power output, obtained by
Mode 2 of the HP machine was found to have gimilar modulating the voltage regulator reference supply at the fmt two
characteristics It was not possible to excite mode 3 of the HP torsional naturalfrequenciesinturn. These measurements have
machineormodes 2 and 3 of the L?' machine by sinusoidal been normalized to the same base ES the calculated mode shapes 80
perturbation of the voltage regulator reference supply for reasons that a direct companion can be made. The measured results shown
to be discussed in the next section. Damping information was, in Fig. 12, for the P i t two modes correspondto a load of 77% of
however, obtained for these modes, at various loads, by passing full load, and are seen to be in good agreement withthe calculated
themulti-modal mechanical response signals obtainedfollowing shapes. At lower loads because of decreased damping itwas found
1882
 Agreement is seen to be good andour abilityto derive strains from
vibratory rotational signals has M e r been c o n f i i e d by more
extensme testing recentlyconducted on atandemcompound
EN EXCITER turbinegenerator for which several shaft spans were strain gaged.
I 2 3 4 5 6 7 8 910
Double ResonancePeak Phenomenon
Fig. 13 shows the form of the torsional amplitude response that
was obtained at all measurementlocations on the HP machine
when sweeping sinusoidally through the frequency range in the

Fig. 12. HP N o m 1 Mode Shapes
that agreement was not as good due to the difficulty in exciting
the rotor system precisely at its natural frequencies because of the
increasing sharpness of the resonance peaks. It was not possible
during the test to excite the third torsional mode by the above
method due to the small motion of the exciterinthis mode.
Calculations of thefrequency response characteristics of the
generator - exciter - voltage regulator electrical system,showed
that by modulating the v o l m regulator reference supply at the
third torsional natural frequency, the majority of the oscillatory
torque will be applied at theexciter location.Fig. 1 2 shows that in
the third mode the motion of the exciter is very small relative to
that in the firet two modes. This means that it will be much more
difficult to excite the third mode in comparison to the first two
by applying torques at this location. In fact, in the limiting case
wheE it is assumed that the exciter is at a node inthe third mode,
it would be impossible to stimulate this mode at all by a torque
applied at theexciter. An additional important effect in producing
torsional displacement is the relative phase angle of the torques
applied at the generator and exciter locations with respect to the
phase angle of the modal displacement at these two locatio= If
the modal displacement of the exciter and generator rotors are in
phase opposition then the applied torques should ideally respond
inthe samerelativephaseand proportions, else the torques
produced may cancel one another to produce no net displacement,
or a very small value. This is believed to be the principal reason for
our inability to stimulate the second and third modes of the LP
generator by sinusoidal perturbation.
Shaft Strain Correlation
An important aspectof the test program was to determine the
accuracy with which shaft strains could be predictedfrom
measured rotational displacement signals and a knowledge of the
calculated mode shapes. Successful correlation would mean that
torsionalvibrationmonitoringequipmentcould bebased on
rotational displacement rather than rotorstrain measurements, the
latter requiring a more complicated telemetry system to transmit
therotor strain signals to stationary equipment. Appendix B,
shows the relationship betweenthe shaftstrains and the rotational
displacements when the turbine-generator is vibrating in oneof its
normal modes of vibration. Table II shows comparisons between
the measuredand calculated strain values for the HP machine
when vibrating in the modespecified.
*
proximity of the calculated second torsional natural frequency.It
willbe noted that two resonance peaks were obtained- 0.09 Hz
apart. Torsional natural frequency calculations for the HP rotor
shaft system showed that it was inconceivable that two torsional
natural frequencies could be so close together. It was therefore,
postulated thatthe electrical power system was couplingthe
motions of the HP machines of Unit 1and Unit 2.
30.07 Hz 30.16 Hz
FRONT STANDARD
DISPLACEMENT
Fig. 13. Double Resonance Peak
If vibration analysis is conducted on the two HP units, considered
as a singIe system and interconnected at thegenerator locationsby
a spring to represent the electrical cross coupling, torsional modes
of vibration w li be calculated that involve the motions of both
units. These modes can be conveniently grouped in pairs, due to
the modes in each pair having frequencies extremely close to one
another, the separation beiig determined by the magnitude of the
electrical cross coupling between the two units.One mode of each
pair will correspond to both units moving in phase with respect to
one another with a frequency and mode shape for each unit very
nearly equal to that obtained from the analysis of one unit on its
own with no electrical coupling effects. The second mode in each
pair will differ in characteristics only in that the units in thiscase
will be vibrating in antiphase to one another. It was observed that
the double resonance peak phenomenon was only observed in the
region of the conventional second torsion mode and only then at
low load. Appendix C postulates the reasons for these effects by
the use of a simple mathematicalmodel. Further work needs to be
conducted to obtainabetterrepresentation based on the test
results, particularly in the context of quantifying the magnitudeof
the electrical cross coupling present.
Theunderstanding of thedouble resonance phenomenon is
important as it may directlyreflect on the design of auxiliary
equipment required to help eliminate the subsynchronous resonance
problem. It should as0 be noted that ingeneral if there axe N
nominally identical units in a station feeding into the same line, the
electrical cross coupling will result in families of N closely separated
modes. Non duplicate units will have simiiar characteristics except
the modes are no longer closely separated. Subsynchronous currents
flowing in the line due to disturbances will be shared between units
inthe ratio of theirreactances.This implies duplicate machine
installations are advantageous due to equal current sharing and the
lesser number of distinctlydifferent machine torsionalnatural
frequencies that require protection.

TABLE I1 CONCLUSIONS

Rotational The measurements obtained during the test led to the following
displacement conclusions:
% o f full ModeNo. measurement Measured Strain 1)Sinusoidalmodulation of theexciter power output is a good
load excited location calculated strain method of stimulating the turbine-generator set to respond in pure
modes of torsional vibration, enabling accurate measurements of
Turning gear torsionalnaturalfrequencies, modeshapes,andmodal damping
1.026 Exciter values to be obtained.
1.027 Exciter 2)Good correlation was obtained between measured and calculated
2 26.7 Exciter 0.993 torsional natural frequenciesand mode shapes.
2 66.7 Exciter 0.956 3)The modal damping values increase significantly with increases in
17.3 2 Exciter 0.927 electrical load, but even at the highest load for which data were

1883
 obtained, the damping is shown to be extremely light.
4) The damping information obtained from filtering the multi-modal
responses following transients caused by switching in and
bypeeeing modules of capacitance in the trammission line was in
good agreement with that obtained in the sinusoidal excitation
tests.
5)The calculated shaft strain valuesbased on measured rotational
displacement signals and a knowledge of the normal mode shapes
agreed wellwith the measured strain values.
6)If N nominally identical turbinegenerators are electrically
coupled, the number of torsional natural frequencies will be N
times the number for a single set. The modes will be in groupsof
N that are cloeely separated in frequency,each mode involving the
motion of all the turbinegeneratom
This test served to obtain a better understanding of the torsional Fig. 15. Tmnsient Decay Envelope
dynamiccharacteristics of theturbinegenerator,includingthe
interactive effects of the electrical trammission system. These tests I€theamplitudeatt=t,isA,
have
laid the
foundation
for seeking
a solution to the
sub-synchronous mnance problem in a systematic fashion.Work is
required and is presently underway in thedesign and development of andtheamplitudeat t= is A,
auxiliaryequipment to helpeliminatethis serious application
problem. The solution to the problem will involve close cooperative
effort between the Electric Utilities and the manufacturers of steam
then V by definition

-
A,
=20LoG10 (")
turbine generators.
'** AN = lOv/zo

APPENDICES hN=f(tN-tro)
where f = frequency of the fdtered mode in the decay
(Hz)
Append= A
The mechanical damping values quoted previously are in terms of substituting into (1)yields
the logarithmic decrement(LOG DEC). It is defined as:
LOG DEC=O.l151V/(f (tN -to ))

Appendix B
where the parameters involved are defied in Fig. 14. Relationship between the strain gage signal and the toothed wheel
signals.

Fig. 14. LOG DEC Definition

The LOG DEC is related to other commonly used dehnitions and Fig. 16. Principal Stresses For Shaft in Torsion
for convenience theyare listed below :
The principal stresses in a shaft subjected to pure torsion are at
angle of 45" to the shaft axis as shown in Fig. 16 and are equal to the
nominal shear stress, i.e. P = sxY.
Hence, the strain measured at 45" to the shaft axis which is how
the strain gages were aligned in the test is given by:

For LOG DEC values less than 0.6, the relationships given below E=&+ us,
2
are accurate to better than 1/2%. E E

5 = LOG DEC/2n where Y = Poissons Ratio and

E = Youngs Modulus
Q = magnircation factor= n L O G DEC = 1/21
However, for any isotropic material
Dampingvalueswere obtainedfrom the sinusoidalsweepand
decay tests using the d e f ~ t i o nfor LOG DEC given above.
Damping values were obtained following a transient by filtering
the response to pass only each modal frequency in turn. The filter
$ = 2(l+v) where G is the material Rigidity modulus.
band width wzs sufficient to have negligible effect on the transient
decay. Fig. 15 shows an example of a typical decayenvelope plotted
in decibels

Consider the case of relating the strain gage signal on the HP

exciter shaft to the rotitional displacement signal 6 m e d by
the exciter toothedwheel.

1884
 -
UNIT I UNIT 2

1 2 3 4 5 6 7 0 9 10

Fig. 17. S h i n -Dbpplocement Correlation

2T
The alternating shear stress in shaft span 9-10 i
s- where T is
d
the altemating torque and r is the equivalent shaftradiua
The torque T in span 9-10 d e f i i in Fig. 17 is given by k (6 -
tig) where k is the torsional stiffness of the shaft and 69 & 610 axe
the rotational displacements. ' le2

~ i g18.
. Double Reeononce Peak Model
4 in Fig. 18.
The analysis that follows will predict the form of the resonance
c u m for units1 and 2 to compare with them e d results.
"%e equations of motion are:
where k has units LB IN/R
6 hasunitsdegrees -k 0
rYas units IN k+el+e2 0
G has units Lam2 0 k
Note that 6 and E are quantitiea measured in the test and -
69 -e 1 -k
610
can be determined fromthe mode shape.
For simplicity in what follows, it is assumed that I1 = I2 = I, as the
mode shapes then degenerate to all unity terms as shown below.
For low values of electrical c r o s coupling stiffness relative to the
shaftspan stiffness, as is the case at Mohave, theflexiblemode
shapes of the systemare as given below:

E- = k
'6
r3 G 180
where 61 a d are the meleuredvibratorydisplacements atthe 'Ihis result would be obtained by solving the eigenvalue problem
for the system shown in Fig. 18, andinvolvesapair of modes in
front standud and turning mu locations respectively %,2,'10 which both rotor strings vibrate in phase and in antiphase with one
'1 61 '6 another respectively. At a forcing frequency thatis close to either of
& -
69
cpll be obtained from the calculated modeshapes. these modes, the system will respond in a linear combinationof each,
and the motion can be represented by two generalized coordinates
'6
91 and 92.

Double ResonancePeak Analysis Hence

Fig. 18 is the simplest model that can be used to investigate the
double resonance peak phenomenon and yet preserve the essential
featules mquiredto describe themechanism.
Fig. 18 shows two nominally identical rotorsystems designated as
units 1 and 2, each containing two rotors of polar moment of inertia
I1 and 12 and connected by a shaft of stiffness k. Springs q and 9 where [MI is the mode shape matrix. Substituting equation (3) in
reprwentequivalent mechauicalsprings, the formercoupling each theequations of motion (2) and premultiplyingtheresulting
unit together to simulate the electrical cross coupling that would equation by [MIT yields two uncoupledequations of motion in
result if two turbinegenerators were feeding into the same terms of the generalized coordinates.
transmission line in close proximity. The value'of these springs is
controned by the load and powerfactor and the frequency of
perturbation of the generator masses. Because of this, the evaluation
of the equivalentspringconstants is very complex. Simplifying
assumptions assuming the machine is atno loadwithconstant
excitation and speed and treating only the fundamental frequency These equations may be rewritten as
term will reduce the spring constants to the familiar synchronizing 2
torque relationship. Sl + WN1 91 = -TI41
During the test, sinusoidal torques were applied at the generator 2
q2 + w N 2 q2 = TI41
and excitor rotorsof unit 2 which is simulated by a torque applied at

1885
 where the natural frequenciesare given by: be seen in part from Fk. 19 where it is seen that as the damping is
increased, the peaks merge together to form a single broad peak,
witness the full load damping curve. The extent to which a double
resonance peak is formed is also a function of the frequency
separation between the in phase and antiphase modes. It was found
by doing a detailed analysis using ten inertias for each Mohave unit,
For low values of elthese two frequencies will be extremely close, that the frequency separation was much smaller in the first and thiid
as o b s e d at Mohave and in the limit when q = 0, the frequencies pair of flexible modes comparedto thesecond and this is thought to
will be identical, W i g the natural frequencies of each rotor string be the reason for the apparentabsence of the double resonance peaks
separately, Le., an uncoupled system. in these modeseven at lowloada
If a sinusoidal response solution is assumed and the modal
damping terms added, the followingsolutionfor the generalized
coordinates w libe obtained. REFERENCES
-T
“= 41 EwN~’ ~ w 2 ) + 321 w N I 4 1) J.W. Butler and C. Concordii “Analysis of Series Capacitor
application Problems,” AIEE Trans,1937.
2) C. Concordia and GX. Carter, “NegativeDamping of Electric
Machinery,” AIEE T r q V 0 1 . 60, pp. 116-120,1941.
+T 3) CP. Wagner, “Self Excitation of Induction Motors with Series
92 = Capacitors,” AIEE Trans, Vol. 60, pp. 12414247,1941.
42 [ w - ~w 2 ) ~+ at2
~ uN2w where21and 22 are the 4) RD. Bodine, C. Concordia, G. &on, “Self Excited Oscitions
modal damping values. of CapacitorCompensatedLong Distance Transmission
Systems,” AIEE Trans, Vol. 62. pp. 4 1 4 discussion pp.
The rotational displacement responee may now be obtained by 371-372,1943.
back substitution into thetransformation equation (3). 5) RB.Marbury and JB. Owens, “New Series Capacitor Protective
By way of example, the mnance shape was calculated for this Device,” AIEE Trans.1946.
system with a frequency separation that corresponds to the value 6) A.A. Johnson, “Application
Considerations of Series
measured for theMohave HP second pairof flexible modeswhere: . Capacitors, W&gho& Engineering Journal, 1948.
7) H. Rustebacke and C. Concordia. “Self Excited Oscillationsin a
~~ ~ ~~ ~~

uN1 = 30.07 Hz & uN2 30.16 Hz ’ Transmission Systemusing se;ies CapacitoA,” IEEE ~rans,
PAS89, No. 7, pp. 1504-1512,1970.
8) L.A. Kilgore, L.C. Elliott, E.R. Taylor,“ThePredictionand
Control of Self Excited Oscillations Dueto Series Capacitors in
Power Systems,” IEEX Trans,PASSO, Vol. 3,1971.
w2 9) J.W. Ballance and S. Goldberg, “Subsynchronous Resonance in
SeriesCompensated Trananission Lines,” IEEE Trans, Vol.
2
v) PAS92, NO. 5, pp. 1649-1658.1973.
!$i 10) C.E.J. Bowler, D.N. Ewart, C. Concordia, “Self Excited
cu
Torsional Frequency Oscillations with Series Capacitors,” >
E
5 w V 0 1 . PAS92, NO. 5, pp. 1688-1695,1973.
11) R.A. Hedin, R.C. Dancy, KB.Stump, “AnAnalysis of the
SubsynchronousInteraction of Synchronous Machinesand
Tr&ission Networb,” Roc. h e r . Power Conf., Vol. 35, pp.
30.25 30.20 30.15
30 30.10 30.05 1112-1119.1973.
FORCING FREQUENCY 12) L.A. Kilgok, E.R. Taylor, Jr., D.G. h e y , RG. Farmer, E.
Katz, A.L. Schwalb, “Solutions to the
Problems of
Fig 19. Forced Response Subsynchronous Resonance in Power Systems with Series
The results are shown in Fig. 19 for three different mechanical Capacitors,” Roc. Amer. Power Conf., Vol. 35, pp. 1120-1128,
damping values, beiig zero and the values m e d for the second 1973.
mode at no load and full load respectively. W i g the test, the 13) C. Concordia, J.B. “ice, C.E.J. Bowler, “Subsynchronous
double monance peak phenomenon WBS only observed when Torques on Generating Units Feeding Seridapacitor
sweeping through the second flexible torsional natural frequency at ~ opie nsae
td~ines,”ROC. h e r . Power Cod., VOI. 35, pp.
low loadswhich corresponds to low damping, Thereason for this can 1129-1136,1973.

1886
 Discussion The frequency results are tabulated in Figure 2. As in the Mohave
tests, frequencies were measured by three methods, turning gear jogs,
K. Murotani (Nissin Electric Co., Ltd., Kyoto, Japan): The authorsare synchronization, and by sinusoidallyvarying the exciter power output
to be congratulated for having presented such useful and fundamental at each natural frequency in turn by a method similar to that described
data on the torsional natural frequencies and mode shapes which are in the paper. Theslightdownward trend in themagnitudes of the
very important tostudythe sybsynchronousresonanceproblem. torsional frequencies as the electrical load is increased is thought to be
I would like to raise two questions aboutthis paper. due to the increase in the temperature of the rotorshaft system as the
In the paperpresented by theauthor*atthe1973 Power load is increased, resultingin a slight reduction of the torsional stiffness
Engineering Winter Meeting they reported that the turbine generator of the turbine-generator.
torsional damping, as defined in Dl 2, is very important and theyused No double resonance peaks were observed in the test, which fits
three values, such as 0.0,0.6 and 1.2 for their example of the hydraulic thehypothesis givenin the paper fortheoccurrence of this phe-
system. nomenon at Mohave, as at the time of the test, Navajo 2 and 3, which
I wish to knowtherelationshipbetweentheturbine-generator are duplicates of No. 1, were not in service.
tortional damping in that paper and the logarithmic decrementgiven in
Figure I O of the author's paper.
Then what kindofcounter measures theauthorsare now
proposing to prevent theshaftfailuresdue to thesubsynchronous
resonance at theMohave Generating Station orhas any countermeasure
actually been installed?
(*) Reference Number 10 W
Y

Manuscript received February 10,1975.

R. Quay (General Electric Co., Schenectady, N.Y.) and A.L. Schwalb

(Salt River Project, Phoenix,Az.): This paper has presented the method
of testing and the results obtained in the subsynchronous resonance
test conducted on thecross compound Mohave No. 2 turbine-generator.
A similar test was completed in May 1974onthe SaltRiver
Project's tandemcompound Navajo 892,000 KVA unit No. 1. Selected
results of this test are presented below to supplement the resultsgiven
in the paper and endorse the approach of obtaining accurate torsional
natural frequency and modal damping data by sinusoidallyvarying the
directly connected exciterpower output.
A future paper on the Navajo tests is planned. We are taking this 5% -0.75
opportunity now to present basic results urgently needed for system 30%-0.51)
design studies. 50% -0.72
92% -0.75

-
NAVAJO #l I u o u w K * L yLI#IuyLIsIJ

FIG I Figure 3 shows relative rotational displacementat various positions

FS HP RHT Lm LPB G€N EXC
along the shaft for thef m t mode with a calculated natural frequencyof
15.8 Hertz. The solid line is calculated. The circled points represent the
positions at which the rotational displacements were measured during
the sinusoidal excitation tests at several electrical loads. These
measurements werenormalized tothe samebase as thecalculated
results, so that a direct comparison couldbe made. In this case the base
motion is at the exciter end,and the comparisonof test and calculation
is at the front standard. As in the Mohave tests, good correlation was
obtainedbetweenthecalculated andmeasured results - with the
points for all loads falling on top of one another.
Figure 1 shows the principal rotor elementsof the Navajo turbine-
generator, and thelocations of the mechanical measurements. This
turbine-generator has five torsional natural frequencies less than 60 Hz,
and the frequencies and modal damping constants were measured for
the first four modes.

MEASURED TORSIONAL NATURAL FREQUENCIES

FIG 2
1
OI NRNlMGOE*R*XI 16.0 a.4
ox sYNcHnDNIzITK*I 166 Z0.2
0.77% SIIWOOIDUWTATWN 15.78 a26
6- SIIWOOIDALEXCITATION 16.- a38
33.6% SIWDUEXCITATION lW73 7.0243
43.0% OIWOOIDALEXCITATION - -
443% SIlWgWDALEXClTATlDN - -
10.7% SIIWOOIDUUCITATKIII 16.77 ZOW
64.8% SINUSDIDALEXClTAllDN 16.n 2023
74.8% ONUCOlOALEXClTATlON 16.71 2023
S
ZB SINISDIDALD(WTATKIII 16.m a021
@7.4% SINUPQIDALU(CITAT~ 15.77 2
0-
CALCULATEDVALUES 151 Z0.J Similarly, Figure 4 is the second mode at 20.3 Hertz, normalized
to the base motion at the exciter, figure 5 is the third mode at 25.8
Hertz, normalized to the front standard, and figure 6 the fourth mode
Manuscript received April 18, 1975. at 33.9 Hertz,normalized at the low pressure turbine.

1887
 MEASURED MODAL DAMPING vs LOAD
FIG 8
r

/
MODE 3

0.018

0.016

1.0 AT FRONT STANDARD) 0.020:

0.014
5%-0.296
30%-0.260
5 0%-0.220
75%-0.250
92%-0.227
9?%-0.24?

Y “ E D VALUES ( BASED
1.0 ON 0.72 AT FRONT STANDARD)
a 0.6
0.6
6- 0.4
0.2
J
a 00.002
~ 0 0 4 ~ 0 D E
2 0
2
t -0.2
I-
&-0.4
!
W
Z -0.6
s -0.8 L I I I I 1 I

0 20 40 60 80 100 12(
-1.0 %OF FULL LOAD

Figure 8 gives a summary of the net modal damping in the first

four modes of the Navajo unit as a function of the electrical load. These
results further emphasize the conclusions givenin the paper that the
modal damping values are extremely low and are highly dependent on
the electrical load.

D.N.Walker, C. E.J. Bowler, R. L. Jackson, and D. A. Hedges: The

authors thank thediscussors for their comments.
The data provided by Mr. R. Quay and Mr. A.L. Schwalb on the
results of the subsynchronous resonance test at Navajo supplement the
results givenin the paper and endorse several of the conclusions.
With regard to the first question by Mr. K. Murotani, the relation-
ship between the mechanical dashpotconstantsand thenon di-
mensional modal logarithmic damping values. (such as those measured
in the test), is obtained as follows.
If the equations of motion of a torsional systemare written in
terms of discrete mechanical dashpot constants, which are dimensional

U J
-0.4 and may act between adjacent inertias and or from each inertia to
Z -0.6 ground, the corresponding non dimensional modal logarithmic damping
5 -0.8
W
K
values may be obtained by uncoupling and then non dimensionalizing
theequations of motionby using thetransformation shown in
-1.0 Appendix C.
If the form of one of the uncoupled modal equations is:
Figure 7 shows the calculated mode shape for the fifthmode with I q + &+Kq = T(t)
a calculated natural frequency of 51.5 Hertz. The fifth modecould not
be stimulated by sinusoidallyvarying the exciter power output. The it may be nondimentionalized by dividingby the modal inertia I to
explanation is similar to that given in the paper for mode 3 of the yield:
Mohave high pressure machine. This figure shows that in mode 5 the
exciter and generator are virtually stationary and it would, therefore, + 2cWN(1+ W N 2q = T (t)/I
require extremely high torque values to be applied at either of these
locations to stimulate this mode. Thefrequency of this mode was, where W N = an undamped natural frequency and { is the corresponding
however, measured to be 53.2 Hertz by passing a long time sample of non dimensional damping ratio.
normal operating data from thefront standard rotational displacement, Appendix A, shows that the modal“logdec” = 2n X modal {.
(for which there is large relative motion in this mode),into a frequency
spectrum analyzer. Manuscript received April 18, 1975.

1888
 The definition of o shown in Fig. 5 of paper Reference 10 is given
by FUNS for a pure modal response, henceo = LOG DEC X f, where
f is the undamped natural frequency in Hz.
Figure 10 of this paper shows that in mode 2 at full load the log
dec = 0.013 and for mode2, f = 30.1 Hz.
Hence using the relationship above, o = 0.3913 rad/sec. This result
is to be compared with a value of 0.6 in Reference 10. It should be
appreciated that the dampingusedin Reference 10 was estimated in
advance of any test results, which in this case gave reasonable values.
It must also be pointed out that the discrete dampers can be assigned
1889
arbitrarily provided the net modal damping is achieved. Most modem
torsional vibration response programs use the modaldampingvalues
directly and the task of deducing the individual dashpotconstants,
even if it were possible, is not required.
With regard to the second question, the counter measure adopted
to prevent subsynchronous resonance failures at Mohave has been the
reduction in the amount of series capacitor compensation used. This
measuremoves the critical electrical resonance away from the
mechanical resonance such that there is no significant torsional
interaction.
ArticleCode : pqa_3311

Mitigating SSR in Hybrid System with Steam and Wind Turbine by

TCSC
Hossein Hosseini* and Behrooz Tousi**
*Engineering Department of Urmia, Urmia, Iran, uh.hosseini@gmail.com
**Engineering Department of Urmia, Urmia, Iran, b.tousi@urmia.ac.ir

Abstract: The increasing requirement to the clean and 1. Introduction

renewable energy has led to the rapid development of wind
power systems all over the world. With growing usage wind With raising worries about environmental pollution and
power in power systems, impact of wind generators on possible energy lack, great labors have been taken by the
subsynchronous resonance (SSR) is importance. SSR is a well- governments surrounding the world to achieve renewable
known phenomenon in a series compensated power systems energy programs, established on Hydropower,
which can be mitigated with Flexible ac transmission systems Biomass/Bioenergy, Geothermal Energy, Wind Energy
(FACTS) devices. In this paper for damping the SSR, a and Photovoltaic (PV) Cells, etc[1].
Thyristor Controlled Series Capacitor (TCSC) has been used. Wind power is the most fast growing technology for
This paper used wind and steam turbine as a hybrid energy renewable power generation systems [2]. The first wind
production system. In order to have an optimal control on pitch
turbines have been appeared at the beginning of the last
angle in high speed of wind, fuzzy logic damping controller
(FLDC) has been used. The main objective of this paper is to century and its technology was increased step by step
investigate the ability of the Thyristor Controlled Series from the early 1970s. At the end of 1990s, wind energy
Capacitor (TCSC) for mitigation of SSR. In order to conduct the has reappeared as one of the most important maintainable
studies, the IEEE second benchmark model on SSR is adapted energy resources, partly because of the increasing price of
with the combination of synchronous wind generator based the oil, safety worries of nuclear power and its
wind turbine. environmental issues [3]. Based on predictions 12% of
the total energy demand all over the world will create
from the renewable wind power by the end of 2020 [4].
Therefore with the fast development of installed capacity
Keywords: Wind Turbine, Steam Turbine, of wind farms, the large wind turbine generators (WTGs)
subsynchronous resonance, SSR, Thyristor Controlled are widely used into electric power grids. Series
Series Capacitor, TCSC, fuzzy logic damping controller, capacitive compensation is a very economical method to
FLDC. enhance the system stability and power transfer capability
especially where large amounts of power must be
Nomenclature transmitted over long transmission lines. However, this
General also leads to occur the phenomenon of subsynchronous
fe Electrical frequency of the power system resonance (SSR) [5],[6]. SSR could contain either
synchronous frequency of the power torsional interactions (TI) or induction generator (IG)
f0
system effect [7].
reactance of series capacitor In recent years, there are many researches which have
Xc
been done in order to damp the SSR. M. S. El-Moursi et
Xl leakage reactance of compensated line
al are proposed a novel STATCOM controller for
Cp power coefficient mitigating SSR [8]. Impacts of Large-Scale Wind Power
A Area swept by the rotor R the Integration on Subsynchronous Resonance have been
radius of the blade [m] analyzed by Tang Yi, and Yu Rui-qian [9]. R. K. Varma
Vw Wind speed et al Mitigated Subsynchronous Oscillations in a Series
ωw Angular velocity of rotor [rad/s] Compensated Wind Farm with Thyristor Controlled
Vω Wind speed upstream of the rotor [m/s] Series Capacitor (TCSC) [10].
R Rotor radius [m] This paper focuses on a Thyristor Controlled Series
P Pitch angle [°] Capacitor (TCSC) based on a supplementary controller
for damping SSR. In this paper wind and steam turbine
have been used by means of a hybrid energy production
system. The IEEE second benchmark model on SSR is
occupied with integrating aggregated synchronous namely SSR [5]. Two types of SSR connections which
generator-based wind turbine to carry out the studies. may occur in a power system are as follows [14]:
This paper shows that the TCSC with fuzzy logic
damping controller (FLDC) in supplementary controller 1. Self-excitation or steady state SSR
is able to mitigate SSR. 2. Transient torques or transient SSR

Self-excitation is separated to two major parts:

Induction Generator Effect (IGE), and torsional
2. System Configuration interaction (TI). The IGE is not possible in series
The arrangement of the study system has been shown in compensated power system. But, the TI and transient
figure 1. This figure shows the IEEE Second Benchmark SSR are usually happen in series compensated power
Model (SBM) combined with TCSC in line 1 which is systems [15].
mainly used for SSR studies [11]. This system comprise
of steam and wind turbines with synchronous generator
supplying power to an infinite bus per two parallel 4. Modelling of Wind Turbine
transmission lines, and one of them is compensated by a A fixed wind speed turbine system based on a
series capacitor by TCSC. A 600 MVA steam turbine- synchronous generator (SG) has been investigated in this
generator and 60MVA wind turbines (which comprise of
work; wind speed is 13.5 m/s and the base speed is 11
40 turbines and each turbine generate 1.5MVA) are
m/s. Following equations described, power extracted
coupled to an infinite bus, and the rated line voltage is
from the wind.
500KV, while the rated frequency is 60Hz. The shaft
system comprise of four masses: a high pressure turbine (2)
1
(HP), low pressure turbine (LP), the generator (G), PW =Cp ρ AVw3
rotating Exciter (EX). Elastic shaft linked all masses 2
together mechanically. -21 (3)
116
FLDC pitch
angle control Cp=0.5 [ - 0.4θ. - 5]e λi +0.0068 
λi
SG Turbines ωwR (4)
λ
Excitation EX G LPB LPA IP HP vω
system

1 1 0.035 (5)
PSS
  
bus1 bus2 X L1
TCSC X system λi λ  0.08θ. θp3  1
G
Generator X L2 Infinite
Wind farm Bus θp is the angle between the plane of rotation and the
Three phase
to Ground blade cross-section chord [16],[17].
Fault

Fig. 1: IEEE SSR second benchmark model supplied by the TCSC

3. Subsynchronous Resonance Rotor plane

SSR is a potential phenomenon which may appear in a

series compensated power system when the mechanical
system (turbine-generator) exchanges energy with the
electrical network [12],[13]. The power system
compensated via a series capacitor has a sub-synchronous (a)
natural frequency (fe) which is assumed by:

Xc (1)
fe  f0
Xl
The generated subsynchronous currents will outcome in
rotor torque at the complementary frequency, i.e.

fr=f0-fe (2)
If fr=f0-fe is near to one of the torsional frequencies of
the rotor shaft, the torsional oscillations will be excited (b)
and this condition lead to detrimental phenomenon Fig. 2: Blade pitch angle θp (a), Power coefficient versus blade pitch
angle (b)
 In recent years, Fuzzy Logic Damping Controllers
(FLDCs) have been appeared as an effective implement
to stabilize the power network. [19], [20]. FLDCs are
more robust and effective than Conventional Damping
Controllers (CDCs) in the power. In this paper two Fuzzy
Logic Damping Controllers (FLDCs) have been applied
to pitch angle control in wind turbine and TCSC
controller.
In this article Power deviation ( ΔP ), its derivatives
ΔP
( ) are inputs of fuzzy controllers and pitch angle
t
( θp ) is output of fuzzy controller.
Fig. 3: Wind Turbine modeled with Simulink 1n 1p
1 2n 2p
1

input1 of wind turbine fuzzy controllers

input2 of wind turbine fuzzy controllers

0.8 0.8

Figure 3 displays the model of the wind turbine 0.6 0.6

equipped by FLDC of pitch angle. Wind power plants are 0.4 0.4

not sustainable like fuel energy production. The wind 0.2 0.2

power generators installations must use a system to 0

-3 -2 -1 0 1 2 3
0

-3 -2 -1 0 1 2 3

support the power generator for the time that wind turbine Power deviation

1
on oz op
Power derivatives

output of wind turbine fuzzy controllers

produces energy is little. In this paper wind and steam 0.8

turbine have been used as a hybrid energy production 0.6

system. 0.4

0.2

5. Modeling of Thyristor Controlled Series 0 5 10 15

Pitch angle
20 25 30

Capacitor (TCSC) Fig. 6: Membership functions for the FLDC of pitch angle controller
A TCSC comprise of a capacitor in parallel with an
inductor that is connected to a couple of opposite-poled The control rules of the fuzzy controllers are showed
thyristors. By adjusting the firing angle of the thyristors, below. The fuzzy sets have been determined as: N:
the inductor reactance is varied and it can lead to change negative, Z: zero, P: Positive, respectively.
the effective impedance of TCSC. The TCSC is ΔP
1. If ( ΔP is P) and ( is P) then ( θp is N)
commonly operated in capacitive region, while inductive t
mode operation can be used during severe possibilities ΔP
[18]. Schematic diagram of TCSC has been shown in figure 4. 2. If ( ΔP is P) and ( is N) then ( θp is Z)
t
ΔP
3. If ( ΔP is N) and ( is P) then ( θp is Z)
t
LTCR ΔP
TCR 4. If ( ΔP is N) and ( is N) then ( θp is P)
t
In TCSC controller, rotor speed deviation ( Δ ),
iTCR Δ
its derivatives ( ) have been used as the fuzzy
t
controllers inputs and firing angle derivatives is output of
i i-iTCR CTCR fuzzy controllers.
Fig 4: Schematic diagram of TCSC 1
mf1 mf2
mf1
1
mf2
input2 of TCSC fuzzy controllers
input1 of TCSC fuzzy controllers

0.8
0.8

0.6 0.6

0.4
Vabc Impedance 0.4
Controller
0.2 0.2
Impedance capacitive mode Alpha
Iabc calculation calculator 0 0
Irms Controller -0.03 -0.02 -0.01 0 0.01 0.02 0.03
-0.03 -0.02 -0.01 0 0.01 0.02 0.03
rotor speed derivatives
inductive mode rotor speed deviation
mf4 mf2 mf3
1
Anfis controller
dw
Output of TCSC fuzzy controller

0.8
+
0.6
ωr Alpha 0.4
t TCR pulses
0.2
Firing unite
0

-15 -10 -5 0 5 10 15

Fig. 5: schematic of Auxiliary FLDC controller of TCSC Reference Voltage

Fig. 7: Membership functions for the FLDC of TCSC controller

6. Design of Fuzzy Logic Damping Controller
(FLDC) And ANFIS Controller
The control rules of the fuzzy controllers are showed 50
below. The fuzzy sets have been determined as: N: 40

negative, Z: zero, P: Positive, respectively. 30

Torque (Gen-LP) and (LP-HP)

20

Δ
1. If ( Δ is P) and ( is P) then ( Δ is P)
10

t 0

Δ -10

2. If ( Δ is P) and ( is N) then ( Δ is Z) -20

t -30
Δ
3. If ( Δ is N) and ( is P) then ( Δ is Z) -40

t -50
0 1 2 3 4 5 6

Δ Time(sec)

4. If ( Δ is N) and (
t
  90  Δ
7.
For verifying the efficacy of the offered control method
to mitigate the SSR phenomenon, the IEEE Second
Benchmark combined with the TCSC is modeled in
MATLAB/Simulink. Two cases for studying are
is N) then ( Δ is N) c
Fig. 8: Simulation results for un-damped mode: (a): rotor speed
(7) deviation, (b): speed deviation between generator, Low pressure
and high pressure turbine, (c): torque between generator, Low
pressure and high pressure turbines.
Results and Simulink in Matlab
SSR oscillations in wind turbine have been shown in
figure 9.
80
Mechanical Electromagnetic Torque
60
Mechanical Electromagnetic Torque

considered. Initially, the power system without any 40

damping controllers and secondly with FLDC controller 20

is simulated. When fault is cleared, large oscillations will 0

be happened among the different parts of the turbine- -20

generator shaft, as shown in Fig. 8. -40

-60
0.2
Rotor speed deviation -80
0 1 2 3 4 5 6
0.15
Time(sec)

0.1 a
Rotor Speed Deviation

0.05 0.05
Rotor Speed Deviation
0
0
-0.05
-0.05
-0.1
Rotor Speed Deviation

-0.1 -0.15

-0.15 -0.2

-0.25
-0.2
0 1 2 3 4 5 6
-0.3
Time(sec)

a -0.35

-0.4
0.25
-0.45
0.2 0 1 2 3 4 5 6
Time(sec)
Speed Deviation (Gen, LP and HP)

0.15
b
0.1
50
0.05
Electrical Power
40 Pitch Angle
0
Mechanical Power
30
-0.05
20
Power and Pitch Angle

-0.1
10
-0.15
0
-0.2
-10
-0.25
0 1 2 3 4 5 6 -20
Time(sec)
-30
b
-40

-50
0 1 2 3 4 5 6
Time(sec)

c
Fig. 9: Simulation results for un-damped mode: (a): mechanical
electromagnetic torque, (b): rotor speed deviation in wind turbine, (c):
mechanical and electrical power and pitch angle.
 12
Figure 9 shows the oscillations in wind turbine due to Mechanical Electromagnetic Torque
10
SSR phenomenon without TCSC.

Mechanical Electromagnetic Torque

8
In this section, a novel FLDC controller has been added
6
to the TCSC for observing specification deviation of the
4
system. Rotor speed, rotor speed deviation and torques of
2
generator are shown in Fig. 10.
0

-2

-4
0.025
-6

0.02 0 1 2 3 4 5 6 7 8 9
Time(sec)
0.015
a
Rotor Speed Deviation

0.01

0.005 Rotor Speed Deviation

0.04

0 0.03

-0.005 0.02

Rotor Speed Deviation

0.01
-0.01
0
-0.015
-0.01

-0.02 -0.02
0 1 2 3 4 5 6 7 8 9
Time(sec) -0.03

a -0.04

-0.05
0.04
-0.06
0 1 2 3 4 5 6 7 8 9
0.03 Time(sec)
Speed Deviation (Gen, LP and HP)

b
0.02
14
0.01 Pitch Angle
12 Mechanical Power
Electrical Powwer
0 10

8
Power and Pitch Angle

-0.01
6

-0.02 4

2
-0.03
0 1 2 3 4 5 6 7 8 9 0
Time(sec)
-2
b
-4
4
-6
0 1 2 3 4 5 6 7 8 9
3 Time(sec)

2 c
Torque (Gen-LP) and (LP-HP)

1
Fig. 11: Simulation results for results with damping SSR by
TCSC with FLDC for pitch anlgle controling: (a): mechanical
0
electromagnetic torque, (b): rotor speed deviation in wind
-1
turbine, (c): mechanical and electrical power and pitch angle.
-2

-3 Figure 11 shows that the TCSC with auxiliary FLDC

-4 controller can be able to mitigate SSR. With optimal
-5 pitch angle control by using FLDC in high wind speed,
0 1 2 3 4 5 6 7 8 9
Time(sec) mechanic and electric power were closed to together. By
c this work wind generator produced power in near the 1pu.
Fig. 10: Simulation results with damping SSR by TCSC with
FLDC controller: (a): rotor speed deviation, (b): speed deviation
between generator, Low pressure and high pressure turbine, (c):
torque between generator, Low pressure and high pressure
turbines.
8. Conclusions

Figure 10 illustrates that the TCSC with auxiliary The increasing requirement to the clean and renewable
FLDC controller can be able to mitigate SSR. energy has led to the rapid development of wind power
systems all over the world. With growing usage wind
power in power systems, impact of wind generators on
subsynchronous resonance (SSR) is importance.
This paper proposed a technique for damping the SSR
by TCSC. The IEEE second benchmark system equipped
by steam and wind turbine as a hybrid energy production
system was studied. A TCSC is originally employed to 122. [17] S. Mishra, P. K. Dash, P. K. Hota, and M. Tripathy,
“Genetically Optimized Neuro-Fuzzy IPFC for Damping Modal
increase power transfer capability of the transmission. As Oscillations of Power System,” IEEE Transactions On Power
shown The TCSC with auxiliary FLDC controller can be Systems, Vol. 17, No. 4, November 2002.
able to damp the oscillations of SSR phenomenon. [20] H. He, “Fuzzy Modeling and Fuzzy Control [Book Review],”
With applied FLDC controller to TCSC and applied Computational Intelligence Magazine, IEEE,
10.1109/MCI.2008.926613.
FLDC for controlling pitch angle of wind turbine, TCSC .
mitigated the SSR and mechanic and electric power of
wind generator were closed to together. By this work
wind generator produced power in near the 1pu.

References
[1] Yazhou Lei, Alan Mullane, Gordon Lightbody and Robert
Yacamini. "Modeling of the Wind Turbine with a Doubly
FedInduction Generator for Grid Integration Studies".IEEE
TRANSACTIONS ON ENERGY CONVERSION, VOL. 21, NO.
1, MARCH 2006.
[2] Rajiv. K. Varma, and Soubhik Auddy,” Mitigation of
Subsynchronous Oscillations in a Series Compensated Wind Farm
with Static Var Compensator”, IEEE, 2006.
[3] Alejandro Rolan, Alvaro Luna, Ge rardo Vazquez, Daniel Aguilar
and Gustavo Azevedo. " Modeling of a Variable SpeedWind
Turbine with a Permanent Magnet Synchronous Generator". IEEE
International Symposium on Industrial Electronics, 2009.
[4] T. Ackerman, Wind Power in Power Systems. John Wiley&Sons,
2005.
[5] P. Kundur, Power System Stability and Control. New York:
McGraw-Hill, 1994.
[6] Luiz A. S. Pilotto, André Bianco, Willis F. Long, and Abdel-Aty
Edris, “IEEE TRANSACTIONS ON POWER DELIVERY”,
VOL. 18, NO. 1, JANUARY 2003.
[7]. A. B. L. Anderson, P. M. and J. E. Van Ness, Subsynchronous
resonance in power systems, New York: IEEE Press, 1990.
[8]. M. S. El-Moursi, B. Bak-Jensen, M. H. Abdel-Rahman “Novel
STATCOM Controller for Mitigating SSR and Damping Power
System Oscillations in a Series Compensated Wind Park,” IEEE
Trans. Power Delivery., vol. 25, no. 2, pp. 429–441, Feb 2010.
[9] Tang Yi, and Yu Rui-qian, ” Impacts of Large-Scale Wind Power
Integration on Subsynchronous Resonance”, IEEE, 2011.
[10]. R. K. Varma, Y. Semsedini and S. Auddy, “Mitigation of
Subsynchronous Oscillations in a Series Compensated Wind Farm
with Thyristor Controlled Series Capacitor (TCSC),” in Proc.
IEEE Power Eng. Soc. General Meeting Conf, pp. 1–7, 2007.
[11] IEEE SSR Working Group, Second Benchmark Model for
Computer Simulation of Subsynchronous Resonance, IEEE Trans.
on Power Apparatus and Systems. 104, N0.5 PP:1057-1064, 1985.
[12]. IEEE Committee Rep., “First benchmark model for
computersimulation of subsynchronous resonance,” IEEE Trans.
Power Appl. Syst., vol. PAS- 96, no. 5, pp. 1565–1572, Sep./Oct.
1977.
[13]. IEEE Subsynchronous Resonance Working Group, “Terms,
definitions and symbols for subsynchronous oscillations,” IEEE
Trans. Power Apparatus and Systems, vol. PAS-104, no. 6, pp.
1326–1334, 1985.
[14]. A. B. L. Anderson, P. M. and J. E. Van Ness, Subsynchronous
resonance in power systems, New York: IEEE Press, 1990.
[15] K. R. Padiyar, Power System Dynamics: Stability and
Control.Bangalore, India: Interline, 1996.
[16] Alejandro Rolan, Alvaro Luna, Ge rardo Vazquez, Daniel Aguilar
and Gustavo Azevedo. " Modeling of a Variable Speed Wind
Turbine with a Permanent Magnet Synchronous Generator". IEEE
International Symposium on Industrial Electronics, 2009.
[17] Nicolae MUNTEAN, Octavian CORNEA and Diana PETRILA,
"A New Conversion and Control System for a Small off – Grid
Wind Turbine".12th International Conference on Optimization of
Electrical and Electronic Equipment, OPTIM 2010.
[18] I.Mohammad Hoseiny Naveh, Elyas Rakhshani, Reza Heidari,
”Application of Linear Observer on Control of Subsynchronous
oscillations Using TCSC”, IEEE, POWERENG, Lisbon, Portugal,
pp. 540–545, March 2009.
[19] Qun Gu, A. Pandey, and S. K. Starrett, “Fuzzy Logic Control for
SVC Compensator to Control System Damping Using Global
Signal”, Electric Power Systems Research, 67(1)(2003), pp. 115–
 Transfer Capability Study of Tie-line Installed TCSC Considering
Transient Stability Constraints

WU Yanjuan[1㧘2] LI Linchuan[1]
1. Key Laboratory of Smart Grid of Ministry of Education, 2. School of Automation, Tianjin University of Technology,
Tianjin University;,Tianjin 300072, China Tianjin 300384, China

Abstract—The transfer capability of tie-lines installed Thyristor
Controlled Series Capacitor (TCSC) is studied in multi-machine
systems considering transient stability constraints. Since the
conversion time of TCSC mode has the important effect on
transient stability, the discrete control strategy(DCS) is presented
based on corrected transient energy function. The multi-machine
system is divided up sending-end and receiving-end two
equivalent machine systems. The moment just at which the
corrected kinetic energy reach the maximum value is used as the
start moment of TCSC maximum capacitive impedance in order
to increase compensation degree, and the moment just at which
the corrected potential energy reach especial value is used as the
start moment of TCSC bypass mode in order to reduce short-
circuit current. IEEE 4-generator system and New England 10-
generator system are used as case studies, transfer capabilities of
tie-lines considering transient stability constraints are calculated
with the conventional control of TCSC and with the discrete
control of TCSC, respectively. The calculation results show that
the discrete control of TCSC enhances significantly the transfer
capability of tie-lines in multi-machine system considering
transient stability constraints compared with the conventional
control of TCSC.
Keywords- TCSC; Transient stability; Transfer capability;
The study concerns transfer capability of tie-line installed
TCSC which applies DSC to satisfy transient stability
constraints. It is a key to select the time of DSC, otherwise,
instability of the power system is deteriorated probably in the
case of selection of DSC improper time. According to the
corrected transient energy function theory[5], the special
moments at which the corrected equivalent generator angular
velocity and the corrected potential energy derivative reach
the special values act as the moment of discrete control. The
method is verified by the study systems of 4-generator-10-bus
and 10-generator-39-bus.
II. MATHEMATICS MODEL
A. Flow equation
Polar form of flow equation acts as the following based on
continuous flow method. 

­ P gi − P Di − V i ¦V j (G ij cosθ ij + B ij sin θ ij ) = 0 ˄1˅
° j∈ N
®
°Q gi − Q Di − V i ¦V j (G ij sin θ ij − B ij cosθ ij ) = 0
¯ j∈ N

Corrected transient energy function; Discrete control strategy

Where,
i ∈ N 䯸P =䯴 1+ λ䯵P 䯸P =䯴 1+ γ䯵P
g gi
0
gi Di
0
Di
I. INTRODUCTION
Frequent Power exchange forces the improvement of P°gi ,P°Di indicate respectively the initial values of the
transfer capability in increasingly large interconnected power
present active power and the present active load. λ,γ indicate
system. Moreover, active power transfer is one important
their incremental coefficient. Gij, Bij indicate real and
indicator on weighing transfer capability of interconnected
imaginary of generator potential admittance according to
power system. There are mainly three research branch to
additional equipment in tie-line .
discuss separately how to improve transfer capability of weak
tie-line based on ensuring power system stability[1~3]. Firstly,
the research is focused on area transfer capacity considering B. Constraints
steady-state constraints. Secondly, dynamic stability of power Static constraints consists of control variable constraints
system is probably damaged only considering steady-state and state variables constraints. Control variable constraints
constraints. Especially when area power transfer of weak tie- include active power of generators, adjustable reactive power,
line is close to thermal stability limit, the power system loses node voltage, active power of tie-line, etc.
probably transient stability as suffering serious disturbance. In
view of this, many documents add transient constraints to their
research model so as to ensure dynamic stability of the system, ­x≤x≤x
in spite of its transfer capacity lower than the former. Thirdly, °
the two formers study from system themselves features ®u≤u≤u ˄2˅
without additional equipment. However, it is contradictory ° h ( x, u ) ≤ 0
¯
between the improvement of transfer capability of contact
section and assurance transient stability. The successful
application of TCSC in power system verifies the feasibility System dynamic model using the classical model, the
that advanced power electronics resolve the contradictory. generators adopting the constant potential E′ and the load
using constant impedance model, the system equations of

___________________________________
978-1-4244-9690-7/11/$26.00 ©2011 IEEE


motion are as follows relative to the system center of inertia protection action because of the large line current. This will
(COI). lead to compensation effect of TCSC being inadequate, make
the MOV protection overload and increase the accumulation of
energy. However, the TCSC operates in bypass mode properly
­δ = ωi [7,8], which can change the total impedance of TCSC and the
° ˄3˅
® M ω = P − P − M i P = f (δ ) phase angle, inhibit short-circuit current, reduce the burden of
° i i gi ei
M¦
COI i
¯ MOV protection and improve the effect of TCSC
compensation.
Where,
Ng
PCOI = ¦ ( Pgj − Pej ) M¦ = ¦ Mi
j =1 i∈Ng

Pei = E i G ii + ¦ [ E i E j Bij sin䯴 δ i − δ j 䯵 + E i E j G ij cos䯴 δ i − δ j 䯵]

2
j∈N g
j ≠i

Pei,Pgi indicate respectively electromagnetic power and

mechanical power for the i-th generator. Based on formula (3),
transient stability constraints are described as power angle Fig 2 The transient stability control sketch of TCSC
constraints which is to make COI as reference system and to
ensure each generator rotor angle relative to COI not Transient control of TCSC is shown in Fig 2. The damping
exceeding a certain limit as the criterion. control link adopts conventional PSS control. Combining the
discrete control theory [9], DCS is added. The control input
signal is active power of tie-line. DSC link is shown in Fig 3,
Hi = δ −δ it COIt
≤ δ lim 䯸 i ∈ ng 䯸 t ∈ T ˄4˅
in which XTCSCmax, XTCSCrated and XL indicate respectively the
where
ng ng
maximum capacitance value of the TCSC, the output
δ COIt
=䯴 ¦M δ 䯵 / ¦M
i =1
i it
i =1
i
capacitance value of damping control at tci time and small
inductance value of bypass mode; tci, tLi , TC i, TLi represent
respectively the first time of the ith maximum capacitive
III. TCSC MODEL value, the first time of bypass mode, the i-th continuous time
of XTCSCmax and the i-th continuous time of bypass mode
TCSC main circuit consists of a capacitor, bypass reactor considering delay.
in series reverse parallel thyristors and protection of metal
oxide MOV [6]. Three basic modes of TCSC are used: Closed
mode, adjustable capacitance mode and bypass mode.
Impedance of TCSC in relations with firing angle α is
expressed as formula (5).

1 A
X 䯴 α䯵 = − [䯴2 π − α 䯵 + sin 䯴2 π − α 䯵 ] +
TCSC
ωC πω C
4 A cos 䯴 π − α 䯵
2

[ β tan䯴 βπ − βα 䯵 − tan䯴 π − α 䯵 ] ˄5˅

β
2
πω C 䯴 − 1䯵
Fig 3 Discrete control link
ω 20 1 ω
where, A= 䯸ω = 䯸 β =
0
2

ω
0
ω 02 −ω 2
LC
Installation of TCSC equipment will significantly improve
TCSC steady-state control model is shown in Figure 1, the stability in power system [9], especially, the TCSC mode
error between active power of tie-line and the reference value conversion time play important role in dynamic stability.
can change the XTCSC, and further change admittance matrix Selection of the wrong time not only is defeated to improve
elements in the flow calculation. system stability but also make instability deteriorated. For
TCSC in the system to determine the mode transition time is
usually estimated ether using more experience or offline
testing. Lyapunov energy function as a online successful
method has been successfully applied to power system
switching control of TCSC [10,11]. However, the energy
function will increase the number of equations of complex
nonlinear multi-machine power systems. COI based on the
Fig 1 Stability model of TCSC with active power control amendment energy function significantly reduce the number of
equations to accelerate the calculation speed. The DSC is to
study control strategy of TCSC mode conversion time in the
Impedance of TCSC adopting the maximum capacitance multi-machine system based on the corrected energy function.
value in transient stability control process results in the MOV Multi-machine system is treated as two groups of generators


which are connected by tie-line. The corrected energy function iterations, compute flow. If stable, repeat the third step,
of tie-line can be calculated. The moments of its special values otherwise enter the fourth step. The fourth step is that the last
act as TCSC conversion control moment. In summary, the iteration of transfer power is transfer power limit satisfying the
system equation of motion (3) is modified as the following: transient stability constraints. Return to the second step. The
fifth step is to adjust λ ,γ of formula (1) and reduce transfer
power, plus a number of iterations , calculate power flow, if
­  = ω eq stable, back to the second step, otherwise, repeat the fifth step.
°δ eq
° M eq
°
 eq = M eq (ω A − ω B ) =
° M eq ω ¦ (P − P − IV. EXAMPLE ANALYSIS
°
M A i∈M A gi ei
® Mi M eq Mj
° P COI ) − ¦ ( P gj − P ej − P COI ) The first test system is IEEE 4 machines system shown in
° MΣ M B j ∈M B MΣ Fig 4. Three-phase short circuit fault occurs near the bus 7.
°
° M eq M eq
° = ¦ (P − P ) − ¦ (P − P ) = P Supposing that fault occurs at 0s and clears at 0.1s. Time step
¯ M A i∈ M A gi ei M B j ∈M B gj ej eq

adopts 0.01s. Transient simulation time is 6 seconds. The

˄6˅ simulation results are compared by the following three cases:
Where, tie-line without auxiliary equipment, tie-line with Fixed Series
Compensation (Compensation degree of FSC is 45%) and tie-
M = M M M Σ M = ¦M M = ¦M
eq A B A
i∈
MA
i B
i∈ MB
i
line with TCSC (that of FSC is 30% and that of TCSC is 15%
ω eq
= ω A − ω sB
ω = M ¦M ω
A
1
i i ω B
=
1
¦M ω i i
using DSC). β takes 2.799 and selection of other control
M A i∈ A
M B i∈ MB parameters refers to the method of the document [13].
According to the DCS, considering the trigger delay and the
MA, ωA,MB, ωB respresent respectively equivalent inertia effect of errors such time as discrete control of the TCSC is
time constants and equivalent angular velocity of tie-line two shown in Table 1.
ends. After revised sum of kinetic and potential energy along
the fault trajectory is conserved.
1 δ
+ = const
eq

M eq ω eq2 ³ p eq dδ eq
δ
2
sep
eq

The extended equal area criterion[12] certifies that

corrected kinetic energy is maximum when equivalent
generator angular velocity reaches maximum. At just the
moment impedance of TCSC should reach immediately the
Fig 4 The map of 4 generator and 10 bus system
maximum capacitance value in order to ensure the system
running in a stable region, reduce acceleration area and
increase the deceleration area and improve the transient Applying the time of discrete control of Table 1 and
stability. Furthermore, when the derivative of corrected control policy above mentioned, the active power limits of tie-
potential energy is zero at the just moment when the line considering transient stability constraints are shown in
equivalent generator angular velocity is less than zero, the Table 2. As can be seen from the table, the DSC of TCSC will
TCSC immediately operates in bypass mode in order to greatly enhance transfer capability satisfying transient
change the total impedance of TCSC and the phase angle, stability.
inhibit short-circuit current, reduce the burden of MOV
Table 1 Discrete control schedule
protection and improve the compensation effect of TCSC. In
summary, DSC is described as following: i 1 2 3 4
• the moment at which generator angular velocity tci(s) 0.02 1.4 2.4 3.2
reaches maximum acts as the start time of XTCSCmax.
Tci(ms) 120 80 40 40
• The moment at which the derivative of the corrected
potential energy is zero and the angular velocity of tLi(s) 0.96 1.7 2.75 3.6
modified equivalent generator is less than zero acts as
the start time of TCSC bypass mode. TLi(ms) 200 200 40 40

DCS full use of TCSC features that can quickly switch

between bypass mode and capacitive mode to enhance the
Table 2 Look-up table of transmission active power limit by tie line satisfying
system transient stability and improve the transfer power
considering transient stability constraints. The algorithm transient stability
follows as: The first step is that the results of power flow act
Tie-line Without additional With FSC With TCSC
as the initial value. The second step is transient stability test, if equipment
stable, enter the third step, otherwise go to the fifth step, if all
test cases are finished, end up. The third step is to adjust λ ,γ Plmax (MW)   
of the formula (1) , increase transfer power , plus a number of


  greatly improving transfer capability of multi-machine system
and strengthening transient stability.

REFERENCE

[1] Peng Wu, Haozhong Cheng, Jie Xing, et al. “A new transmission
network planning model considering power reserve”, Automation of
Electric Power Systems, 33(6):44-48, 2009
[2] Jing-jing Zhao, Wei Wei, Cheng-shan Wang, “Distributed transfer
capability computation of electric power system based on decomposed
system network”, Proceedings of the CSEE, 2008, 28(7): 1-6

[3] Jingqiang Sun, Dazhong Fang, “Total transfer capability with transient
Fig.5 Maximum D-value of power-angle curve stability constraints”. Automation of Electric Power Systems, 2005,
29(8): 21-25,70
[4] Ming Wei, Yuhong Wang, Chaobo Dai , “Experimental research on
To better illustrate the advantages of the DSC, the fig.5 control strategy of TCSC for UHVAC transmission system from Yimin
shows maximum Power angle difference curves of two control to Fengtun by real-time digital simulator”, Power System Technology,
method at the same 500.9MW of transmission power. 2009,33(4): 71-76
Obviously, the DSC ensure transient stability, yet the [5] Yanjuan Wu, Lin-chuan Li, “Transient stability constrained optimal
conventional PSS control has lost. power flow based on sensitivity of transient energy margin”. Power
System Technology,.2005,29(15): 28-33
The second test system is 10 machine 39 bus system in [6] Xiaoxin Zhou, Jianbo Guo,Jiming,et,al, “TCSC on power system”,
which single line of line16-19 is changed into double lines. Beijing, Science Process, 2009
Three-phase short circuit fault occurs at the near the bus 16, [7] Xu Zheng, Zhihong Pei, Shaorong Wang,et,al “Study on transient
yet other conditions with reference to 4 machines system. characteristics of TCSC with dynamic simulation”, Electric Power
Table 3 lists the discrete control time. Table 4 lists the active Automation Equipment, 2005,25(6):17-21
power limits of four cases considering transient stability [8] Jiming Lin, Jianchao Zheng, Chengqi Wu,et,al. “Digital simulation of
basic control and overvoltage protection for TCSC” Automation of
constraints. Improvement of transfer capability satisfying Electric Power System,2000, 24(8):26-28
transient stability evidently is seen because of adoption of the
[9] Nicklas P. Johansson, Lennart Angquist, Hans-Peter Nee, “Adaptive
DSC of TCSC. control of controlled series compensators for power system stability
improvement, 2007 IEEE Lausanne Powertech Proceedings, 2007: 355-
Table3 Discrete control schedule of 39-bus system 360.
i 1 2 3 4 [10] Padiyar,,K.R.,Uma,Rao,K., “Discrete control of series compensation for
stability improvement in power systems”, International Journal of
tci(s) 0.02 2.38 4.38 5.8
Electrical Power and Energy System, 1997,19(5): 311-319.
[11] BORTONI G, LUSVARGHI R, LACOSTE J, et al. Impact newly
Tci(ms) 80 40 20 20 conceived flexible AC transmission system control strategies on
the dynamic behaviour of power system[C],International Symposium
tLi(s) 0.96 3.0 5.0 6.38 on Electric Power Engineering, Stockholm,Sweden, 1995:89-94.
[12] Yusheng Xue, Van Custem,Thierry,Ribbens-Pavella, Mania, “Extended
TLi(ms) 40 40 20 20
equal area criterion justify cations,,generalizations,,applications”,IEEE
Transactions on Power Systems, v 4, n 1, p 44-52, Feb 1989
[13] Qiang Guo, Xiaoxin Zhou, “ Parameter optimization of auto disturbance
Table 4 Look-up table of transmission active power limit by tie line satisfying
rejection control of thyristor controlled series capacitor”, Power System
transient stability
Technology,1999,23(11): 28-30
With TCSC
Tie-line Without additional With FSC
equipment PSS DCS

Plmax (MW)    

ฎUGCONCLUSION

Combined with its own characteristics, the DSC of TCSC

installed in the tie-line of multi-machine power system is
proposed on the basis of corrected energy function theory. The
relation is put forward between DSC conversion time of TCSC
and corrected energy function. Taking 4-machine 10-bus
system and 10-machine 39-bus system for examples, the
validity of the DSC of TCSC installed in tie-line is verified


 Research on Design and Application Based on TCSC
Optimal Model for Congestion
Fei Li
College of Electrical Engineering and Information Science
China Three Gorges University
Yichang, China
Lifei_ctgu@163.com
Abstract-People pay more attention to the use of FACTS devices
to ease grid congestion problems. On the base of the traditional
optimal power flow (OPF), an optimal congestion dispatch model
with TCSC devices is proposed in this paper, expressing the
control of the TCSC over the power systems as additional power
equations. Instead of introducing a TCSC controller equation,
optimization can be achieved in the variable space including
TCSC controlled parameters. Non-linear interior point method
(Nonlinear IP method) is used to solve the optimal problems. The
results show that the control of TCSC can flexibly adjust the
network power flow, and effectively ease the network congestion
problems.
Keywords- Congestion; TCSC; Additional power; Nonlinear IP
method
I. INTRODUCTION
With the interconnection of the large regional power grid
and the establishment of the future electricity market, since the
grid power flow randomness and the probability of
transmission congestion occurred greatly increased,
transmission congestion has become more serious that affects
the safe and stability of the modern power grid [1], which is bad
for the efficient and economical operation of electricity
market. How to solve the transmission network congestion
problems have become forefront topics at the field of
electricity market.
In the traditional vertical management mode, once the
transmission power exceeds the limited case, the dispatch
center schedules independently according to certain principles,
which is different from the electricity market environment. The
electrical energy is circulated by electricity trading. The
economic interests of all market participants is involved in
eliminating the congestion process, meanwhile, eliminating
congestion requires not only the technical support, but also the
consideration of the justice and benefits, which makes the
congestion problem more complicated. Elimination of
congestion mainly uses two types of technical and economic
methods. The former mainly relies on the advanced and
reasonable scheduling means (e.g. FACTS controlling
technology [2]); the latter mainly uses the price mechanism [3].
978-1-4244-4813-5/10/$25.00 ©2010 IEEE
Xianshan Li
College of Electrical Engineering and Information Science
China Three Gorges University
Yichang, China
lixianshan@ctgu.edu.cn
As the special advantage of flexible AC transmission
device (FACTS) [4], such as quickly adjusting, convenient,
directional adjustment as needed and good effects, it plays an
increasingly important role in the network congestion
scheduling.
In the early time, transmission congestion can be eased by
improving the tie-line power flow capacity through power flow
control [5]. Based on DC power flow model, the control effects
of TCSC and TCPS are expressed as additional power flow of
nodes. Coordination and optimization for congestion
controlled problems of power system which contains a number
of phase shifter is deeply studied in reference [6], and
interior-point method is applied to the solution process.
Recently, the injection power of FACTS device is expressed as
independent variables [7], and introduced it to the global
optimal control mode of FACTS device, using non-linear
programming method to solve the problem for improving
system transmission capacity. Using FACTS device to
improve system security is presented in reference [8]. It gives a
new format for solving the optimization problem: It expresses
additional injected power and control parameters of FACTS
device as independent variables introduced to the optimization
problems, and successive quadratic programming algorithm is
used to the solution, but lack of the analysis of convergence
performance. Many scholars have considered that the
introduction of FACTS control parameters will increase the
system's non-linear and non-convexity, and in some cases, the
convergence will be affected by using numerical method
directly, even not converge. The introduction of UPFC control
parameters affecting the convergence of optimization
algorithm is proposed in reference [9, 10].
In this background, from a point of view of technology,
this paper focuses on the study of the optimization congestion
method including the FACTS devices (e.g. TCSC). The part 1
introduces FACTS congestion model, and the optimization
congestion model containing TCSC is proposed in part 2.
Using the system of IEEE5 to check the method is in part3,
and the part4 comes to the conclusions.
II. FACTS CONGESTION MODEL
FACTS devices using to active power regulation are as
follows: thyristor controlled series compensation (TCSC),
controlled phase shifter (TCPS), unified power flow controller
(UPFC).
There are mainly the physical model and the power
injection model for power flow control of FACTS devices. The
physical model is based on the physical characteristics. The
other is power injection model. It expresses the control of the
FACTS devices as injection power flow of nodes and can be
good to combine optimization power flow method. In this part,
the power injection model [11] of FACTS devices is introduced.
TCSC is taken as an example to illustrate the method proposed
here.
TCSC equivalent circuit model is shown in Figure 1.
i Zij=rij+jxij
-jxc
Fig.1 TCSC equivalent circuit model
i Zij=rij+jxij
△Si
Fig.2 TCSC power injection model
Fig.2 shows the power injection model of TCSC. The
injected power of lines at both ends can be expressed as:
⎧ΔPi = Vi2 ΔGij − VVi j (ΔGij cos θij + ΔBij sin θij )
⎪ the accuracy inevitably.
⎪ΔQi = −Vi2 ΔBij + VV i j (ΔBij cos θij − ΔGij sin θij )
⎪
⎨
⎪ΔPj = V j ΔGij − ViV j (ΔGij cos θij − ΔBij sin θij )
2 III.
⎪ MODEL ADDING TCSC
i j (ΔGij cos θij + ΔBij sin θij )
2
⎪⎩ΔQ j = −V j ΔBij + VV
KC X ij2 ( KC − 2)
ΔGij =
(rij2 + X ij2 ) ⎡⎣ rij2 + X ij2 (1 − K C ) 2 ⎤⎦
KC X ij ⎡⎣ X ij2 (1 − KC ) − rij2 ⎤⎦
ΔBij =
(rij2 + X ij2 ) ⎡⎣ rij2 + X ij2 (1 − K C ) 2 ⎤⎦
KC = X C / X ij
In which,
Additional active power of line caused by TCSC can be
expressed as:
ΔPij = Pij ′ − Pij = −Vi 2 ΔGij + ViV j (ΔGij cos θ ij + ΔBij sin θ ij )
KC is defined as the compensation degree of TCSC. When
KC >0, TCSC comprehensive reactance represents capacitive,
otherwise is perceptual. It considers as controlled parameters
of TCSC. The value range of TCSC is among the maximum of
the capacitive and perceptual compensation degree. In
practical engineering calculations, since the reactance is much
larger than resistance in high-voltage lines, it ignores the
resistance to simplify the model which is easy to resolve. It
supposes the resistance rij=0, and substitute it to equation (2):
ΔGij = 0 ΔBij = Kc / ⎡⎣ X ij (1 − K c ) ⎤⎦ = KTCSC / X ij
(4)
KTCSC = K C /(1 − K C )
There, , and：
KTCSC max = KC max /(1 − KC max )
K TCSC min = K C min /(1 − K C min )
j In the simplified equations, KTCSC is controlled parameter
of TCSC in steady-state analysis. The following analysis will
use simplified parameter of KTCSC.
Calculated by equation (1 ~ 3) were:
⎧ ΔPi = −ViV j sin θij KTCSC / X ij
⎪
⎪⎪ ΔQi = −Vi 2 KTCSC / X ij + ViV j cos θij KTCSC / X ij
j ⎨
⎪ ΔPj = ViV j sin θij KTCSC / X ij
⎪
⎪⎩ ΔQ j = −V j KTCSC / X ij + ViV j cos θij KTCSC / X ij
2
△Sj
ΔPij = ViV j sin θ ij K TCSC / X ij
(5)
The simplified equations ignore the resistance of line and
eliminate the nonlinear term, which helps to improve the
convergence of resolving the power flow, while it also affects
OPTIMIZATION CONGESTION MANAGEMENT
(1) In the system environment, the total cost of fuel to
generate electricity (6) and net loss can be chosen as the
objective function for optimal congestion model.
Fcos t = ∑
i∈S
( a 2 i PGi2 + a1i PGi + a 0 i ) (6)
G
Considering the equation and inequality constraints:
(2) A. Equality constraints are as follows:
⎧ n
⎪⎪ ΔPi = PGi − PDi + ΔPi − Vi ∑ V j (Gij cos θij + Bij sin θij ) = 0
'
j∈i
(7)
⎨ n
⎪ ΔQi = QGi − QDi + ΔQi' − Vi ∑ V j (Gij sin θij − Bij cos θ ij ) = 0
(3) ⎪⎩ j∈i
ΔPi ' , ΔQi' are the additional active power and reactive
power of the nodes connected with TCSC respectively.
B. Inequality constraint are as follows:
Upper and lower bounds constraints of generator’s active
power and reactive power are as follows:
 ⎧⎪ Pgimin ≤ Pgi ≤ Pgimax i ∈ mp IV. SIMULATION EXAMPLES
⎨ min max IEEE standard 5-node system (Fig.3) is used to explain
⎪⎩Qgi ≤ Qgi ≤ Qgi i ∈ mp
(8) congestion phenomenon. The active power of 1# generator is
6p.u. TCSC is installed at the branch (2-3), the boundary of
Upper and lower bounds constraints of node-voltage are as the active power of each branch (1-2,1-3,2-3,2-4,3-5) is
follows: respectively 2p.u.,0.65 p.u.,2 p.u.,6 p.u. and 5 p.u..
4 2 3 5
Vi min ≤ Vi ≤ Vi max i∈n 1# 2#
(9)
G G
The line flow constraints are as follows:

Plijmin ≤ Plij + ΔPlij ≤ Plijmax 2+j1 3.7+j1.3

i, j ∈ n, i ≠ j 1
1.6+j0.8
i j (Gij cos θij + Bij sin θij ) − Vi Gij
2
Plij = VV
(10) Fig.3 IEEE standard 5-node network topology

TCSC control parameters constraints are as follows: 2 0.1

10

K Cmin ≤ KC ≤ KCmax i ∈ mk (11)

0
10 0.05

-2
10 0
△ Plij is the additional active transmission power
generated in the line i-j due to the introduction of TCSC; KC -4
10 -0.05

is the controlled parameter of TCSC. Superscript “max” and -6

“min” indicate the maximum and minimum value constraints 10
0 10 20 30
-0.1
0 10 20 30
of the corresponding variables.
Fig.4 Convergence of the Gap Fig.5 Convergence of the KC
The original-dual interior point method [12] is used to solve
the optimization problem. The original-dual interior point 4 2 3 5
method is also known as tracking center locus method. The 1# 6 1.5251 2#
original-dual interior point algorithm is essentially a 2.12
combination of Lagrange function, Newton algorithm and the G G
logarithmic barrier function. While maintaining the feasibility 2
and viability of the original and the dual solutions, it finds an
5.7946 1.88 0.0998 1.6729
optimal solution along a primal-dual path. It can inherit the 2+j1 3.7+j1.3
advantages of Newton algorithm of OPF, and deal with 1.7946 0.058
functional-type inequality constraints and variable-type 1
inequality constraints well.
Power flow calculation 1.6+j0.8
When considering the controlled parameter variable Ki, Optimization congestion calculation adding
comparing with the OPF algorithm, it has the main change of TCSC located in tie-line (2-3)
adding the effects of additional power caused by TCSC in
power flow equation constraints of nodes connected with Fig.6 Simulation results of transmission power of each branch
TCSC and active inequality constraints related to TCSC (see
above section 2). It needs to consider the effects of Fig.4 and Fig.5 show the convergence of Gap and KC,
corresponding additional power in the solution of OPF Fig.6 shows the active power of all braches of the system.
calculation, including calculation of the initial value, Jacobi From these data above, by conventional power flow
and Hessen matrix which related to the additional power calculation, congestion phenomenon has occurred at the
equations of TCSC. To Jacobi and Hessen matrix, it only branch (2-3). After adding a TCSC device at the branch (2-3)
needs derivative calculation for additional power individually, and using optimization congestion model calculations,
and takes the results to the corresponding location of Jacobi congestion is eliminated. The model took full advantage of the
and Hessen matrix. Considering that the additional power physical characteristics of the network, increasing network
only partly related to the branch parameters and the node transmission capacity, which was conducive to eliminate the
variables connected with TCSC as well as the TCSC control congestion phenomenon.
parameter itself, so the modification of Jacobi and Hessen
V. CONCLUSIONS
matrix caused by the introduction of TCSC control parameters
and the additional power has a little calculation. This paper proposes an optimal congestion dispatch model
adding TCSC, expressing the regulation of the TCSC over the
power systems as additional power equations. Nonlinear IP
method is used to solve the optimal problems. Comparing
with the optimization power flow algorithm without TCSC, it
adds the effects of additional power equations among the
resolve process, which is easy to use the algorithm and
program of former OPF. Results obtained by the method
proposed here show that FACTS devices can effectively ease
the congestion of transmission lines to promote the healthy
development of power industry.
REFERENCES
[1] Zhang Yongping, Jiao Lianwei, Chen Shousun, et al. Electricity Market
Congestion Management Overview [J]. Power System Technology,
2003, 27 (8): 1-9.
[2] Zhang Yanping, Ge Chaoqiang, Tang Guoqing. FACTS Application in
Congestion Management [J]. Electricity Demand Management, 2002,
10:16-19.
[3] Harry Singh ， Shangyou Hao, Alex Papalexopoulos. Transmission
congestion management in competitive electricity markets[J]．IEEE
Transactions on Power Systems，1998，13(2)：672-680．
[4] Xie Xiaorong, Jiang Qirong. Flexible AC Transmission Systems Theory
and Application. Beijing: Tsinghua University Press, 2006.
[5] Taranto.G.N，Pinto.L.M.V.G，Pereira.M.V.F. Representation of FACTS
devices in power system economic dispatch．IEEE transactions on
Power Systems，1992，7(2)：572-576．
[6] James A.Momoh，Jizhong Z.Zhu，Garfield D.Boswell，et al．Power
system security enhancement by OPF with phase shifter ． IEEE
transactions on Power Systems，2001，16(2)：287-293．
[7] Ying Xiao， Y.H.Song， ChenChing Liu ， et al． Available transfer
capability enhancement using FACTS devices．IEEE transactions on
Power Systems，2003，18(1)：305-312
[8] Carsten Lehmkoster．Security constrained optimal power flow for an
economical operation of FACTS-devices in liberalized energy
markets[J]．IEEE transactions on Power Delivery，2002，17(2)：
895-897．
[9] Xu Yan, Li Naihu. Power Flow Control of Interconnected Power Systems
Based on FACTS Technology [J]. Electric Power Systems Automation,
1998, 22 (5): 37-40.
[10] Naihu Li，Yan Xu，Heng Chen．FACTS-Based Power Flow Control
in Interconnected Power System．IEEE transactions on Power Systems
，2000，15(1)：257-262．
[11] Zhao Jianjun, Guo Jianbo, Zhou Xiaoxin. Using the Method of Injecting
Current from Additional Nodes for Modeling and Simulation of SSSC
Power Flow Controller [J]. Proceeding of the CSEE, 2005, 25 (23) .1-6.
[12] Wang Xifan. Modern Power System Analysis [M]. Beijing: Science
Press, 2003.
 2013 International Conference on Power, Energy and Control (ICPEC)

Improving Power System Dynamics By Series

Connected FACTS Controllers (TCSC)
K. Balamurugan G. Suganya N. Manojkumar
Asst. Prof III in Department of EEE Asst. Prof in Department of EEE PG Student in Department of EEE
SASTRA University, Thanjavur PRCET, Thanjavur. SASTRA University, Thanjavur
INDIA INDIA INDIA
kbm@eee.sastra.edu rajkannan93@gmail.com manonmk999@gmail.com

Abstract— Nowadays, the power transmission network
is expanding and is getting more complicated day by day. The
problems of uncontrolled loop flows, overloading, excess of short
circuit current levels and system instabilities are causes of
concern for a power engineer. These serious problems have
necessitated new transmission technologies to be adopted such as
FACTS (Flexible AC Transmission System) devices which make
electrical energy transmission more efficient and reliable. Among
the FACTS controllers the TCSC (Thyristor Controlled Series
Compensators) plays a major role. TCSC is mainly used to
enhance line loadability and damp inter-area oscillations for
smooth synchronous operation of the interconnected regions.
Thus TCSC is used to enhance the effective performance of the
system. The main objective of this paper is to examine the
improvement of power system stability by introducing TCSC in
the power system model through the inspection of their
waveforms, where the design and simulation is done by using
PSCAD (Power System Computer Aided Design).
Keywords- FACTS, TCSC, PSCAD, Dynamic Response,
Three Phase Fault.
I. INTRODUCTION
In a power system the effective power generation,
transmission and utilization is the foremost goals for an
electrical engineer. Among these three modules of a power
system the transmission is more important to be considered as
it includes more losses and instabilities. Nowadays, due to
increase in population the power transmission is expanding
and is getting more complicated day by day. This leads to the
necessity of using new technologies to achieve effective power
transmission.
• Increase transfer capability of transmission lines.
• Control the power flows.
• Assurance of system reliability, by increasing various
stability margins.
• More importantly, increase the overall efficiency by
optimizing the available facilities.
All these can be done using a group of high speed power
electronics controllers called FACTS controllers. These
groups of controllers are expected to revolutionize the utilities
in many ways by rapidly and smoothly controlling power
system quantities. Flexible Alternating Current Transmission
System (FACTS) is static equipment used for the AC
transmission of electrical energy. It is meant to enhance
controllability and increase power the transfer capability of the
network. An efficient and reliable PSCAD stands for Power
System Computer Aided Design based simulation is used for
solving the power system oscillation damping problem is
presented. The proposed approach is carried out in simple two
bus system [1] and [7].
II. FACTS CONTROLLERS
A. FACTS
Alternating current transmission systems incorporating
power electronic-based and other static controllers to enhance
controllability and increase power transfer capability. A power
electronic-based system and other static equipment that
provide control of one or more AC transmission system
parameters [8].
B. Basic Types Of FACTS Controllers
In general, FACTS Controllers can be divided into
three categories.
• Series Controllers:
The series controller could be variable impedance, such
as capacitor, reactor, etc. or a combination to serve the desired
need. In principle, all series controllers inject voltage in series
with the line. Even variable impedance multiplied by the
current flow through it, represents an injected series voltage in
the line. As long as the voltage is in phase quadrature with the
line current, the series controller only supplies or consumes
variable reactive power. Any other phase relationship will
involve handling of real power as well.
• Shunt Controllers:
As in the case of series controllers, the shunt controllers
may be variable impedance, variable source, or a combination
of these. In principle, all shunt controllers inject current in to
the system at the point of connection. Even variable shunt
impedance connected to the line voltage causes a variable
current flow and hence represents injection of current in to the
line. As long as the injected current is in phase quadrature with
the line voltage, the shunt controller only supplies or

978-1-4673-6030-2/13/$31.00 ©2013 IEEE

98
 2013 International Conference on Power, Energy and Control (ICPEC)
consumes variable reactive power. Any other phase
relationship will involve handling of real power as well.
• Combined Controllers:
In principle, combined shunt and series controllers
inject current in to the system with the shunt part of the
controller and voltage in series in the line with the series part
of the controller. However, when the shunt and series
controllers are unified, there can be a real power exchange
between the series and shunt controllers via the power link.
C. Choice of Series Controllers
The shunt compensation is ineffective in controlling the
actual transmitted power which, at a defined transmission
voltage, is ultimately determined by the series line impedance
and the angle between the end voltages of line. It is always
being recognized that ac power transmission over long lines is
primarily limited by the series reactive impedance of the line.
Series capacitive compensation is introduced to cancel a
portion of the reactive line impedance and thereby increase the
transmittable power. The basic idea behind series capacitive
compensation is to decrease the overall effective series
transmission impedance from the sending end to the receiving
end. The effective transmission impedance Xeff with the series
capacitive compensation is given in (1)
Where ‘k’ is the degree of series compensation given in (3)
i.e.,
In series compensation, the FACTS are connected in
series with the power system. It works as a controllable
voltage source. Series inductance occurs in long transmission
lines, and when a large current flow it causes a large voltage
drop. To compensate, series capacitors are connected.
III. THYRISTOR CONTROLLER SERIES CAPACITOR (TCSC)
A capacitive reactance compensator which consists of a
series capacitor bank shunted by a thyristor controlled reactor
in order to provide a smoothly variable series capacitive
reactance. The Control Attributes of TCSC are listed below.
• Current control
• Damping Oscillations
• Transient and Dynamic stability
• Voltage Stability
• Fault current limiting.
A. TCSC Model
The following picture represents the TCSC model as
shown in Fig 1. It has two thyristors connected in anti-parallel,
an inductor and a capacitor [2].
Figure 1 TCSC Model.
The basic thyristor controlled series capacitor scheme
proposed in 1986 by Vithayathil as a method of “rapid
adjustment of network impedance” consists of the series
compensating capacitor shunted by a thyristor controlled
reactor. In a practical TCSC implementation, several such
basic compensators may be connected in series to obtain the
desired voltage rating and operating characteristics. It provides
a continuously variable capacitor by means of partially
canceling the effective compensating capacitance by the TCR.
TCSC is of a parallel LC circuit, consisting of a fixed
capacitive impedance Xc, and a variable inductive impedance
XL. The TCSC thus presents a tunable parallel LC circuit to
the line current. XL is varied from its maximum (infinity)
towards its minimum (ωL), the TCSC increases its capacitive
impedance. Decreasing XL further, the impedance of the
TCSC becomes inductive. Thus the TCSC has two operating
ranges: one is capacitive and the other is inductive [6].
IV. POWER SYSTEM COMPUTER AIDED DESIGN (PSCAD)
A. Introduction to PSCAD
PSCAD stands for Power System Computer Aided
Design. PSCAD is the Professional’s Tool for Power Systems
Simulation. It is a powerful Transients Simulation Software by
Manitoba HVDC Research Centre. PSCAD is a general-
purpose time domain simulation tool for studying behavior of
electrical networks. First developed in 1976, the EMTDC
simulation program has been constantly evolving in its scope
and capabilities. Presently, it is used as a Graphical User
Interface for the EMTDC transients’ instantaneous solution
engine. Users can now harness the power of EMTDC through
the user-friendly PSCAD interface on personal computers.
This seamlessly integrated visual environment supports all
aspects of conducting a simulation including circuit assembly,
real-time control, analysis and reporting. Simply put, PSCAD
is the professional’s tool for electromagnetic transient’s
simulation. PSCAD comes with a comprehensive library of
models encompassing all aspects of AC and DC power
systems and controls. If we need a model that is not in the
PSCAD library, we can create your own model and libraries
using the built-in graphical Component Workshop. PSCAD
supports models written in FORTRAN/C/C++, so constructing
models is easy. Engineers working with circuits consisting of
power electronic switches such as thyristors and GTOs will
99
 2013 International Conference on Power, Energy and Control (ICPEC)

benefit from PSCAD’s Interpolated Network Solution. This
permits study of circuits that would be impractical or almost
impossible to simulate using other tools, and ensures a highly
accurate solution in the least amount of time. The size of a
circuit to be simulated is limited only by the resources of the
computer. PSCAD automatically resizes to give the optimum
performance. PSCAD has many advanced features to boost
productivity, including intelligent data forms, interactive
control and feedback, up-to-date documentation of the circuit
along with plots and comments, context sensitive help,
hierarchical designs, and multiple levels of zooming. PSCAD
enables you to design circuits in a fraction of the time you
would otherwise spend with text-based or semi-graphical
tools. PSCAD is a fast, accurate and easy-to-use power system
simulator for the design and verification of all types of power
systems. For power quality studies, power electronic design,
distributed generation and transmission planning, PSCAD
improves our power system performance and reliability [9].
B. Applications
PSCAD becomes an indispensable tool for a variety
of power system designs and studies. It is a multi-purpose tool.
It is equally capable in the areas of power electronic design
and simulation, power quality analysis, and electrical utility
system planning studies.
PSCAD is a multi-purpose power system simulator and
can thus be used for any scenario where a detailed
understanding of the full time domain of analysis is beneficial.
This includes the design and modeling of virtually any
electrical power system. PSCAD in used in the planning,
design and operational phases of power systems. It is also very
prevalent in power system research around the world.
V. DESIGN AND SIMULATION OF POWER SYSTEM
Here the simple two bus system as shown in fig 2 is
considered for investigate the effectiveness of the proposed
approach [3] – [5]. The system with three phase fault with out
TCSC as shown in fig 3. The corresponding voltage waveform
and the current waveform are shown in fig 7, fig 8 and fig 9;
the maximum system voltage is 300V and damping distortion
time is 0.35. When one TCSC is connected to transmission
line of the two bus system as shown in fig 10, the system
voltage is reduced to 250V and the damping distortion time
reduced to 0.32sec. The voltage and current waveforms are
shown in fig 14, fig 15and fig 16. When two TCSC is applied
the two bus system as shown in fig 17, the system voltage is
reduced to 240V and the damping distortion time reduced to
0.31sec. The voltage and current waveforms are shown in fig
21 and fig 22.

As electrical power and power electronic systems Z1 = 52.9 [ohm] /_ 80.0 [°]
become more prevalent in electric vehicles, ships, trains, and 100.0 [MVA]
V Ph 230.0 [kV], 60.0 [Hz]
distributed generation system, the need for easy-to-use and
accurate simulation and modeling tools becomes ever more SUBSTATION 1 100.0 km SUBSTATION 2
important. It is easier and much less expensive to design and TLine
optimize electrical devices and systems prior to prototyping or 230.0 [kV], 60.0 [Hz] Ph V
manufacturing. Thus, PSCAD is becoming a true Power 100.0 [MVA]
V1 V2
System Computer Aided Design tool for a variety of industry Z1 = 52.9 [ohm] /_ 80.0 [°]
applications. PSCAD users include engineers and Figure 2 Design of simple power system PSCAD model
technologists from energy utilities, electrical equipment
manufacturers, engineering consulting firms, research and
Z1 = 52.9 [ohm] /_ 80.0 [°]
academic institutions. The applications of PSCAD are listed 5000.0 [MVA]
below. V Ph 230.0 [kV], 60.0 [Hz]
SUBSTATION 1 100.0 km SUBSTATION 2
TLine
• To investigate power system instabilities created by
harmonic resonance or control system interactions. 230.0 [kV], 60.0 [Hz] Ph V
5000.0 [MVA]
• To perform insulation co-ordination studies. Z1 = 52.9 [ohm] /_ 80.0 [°]
V1 V2

• To design and simulate variable speed drives of many Timed

Fault
types including cyclo-converters and electric vehicle ABC->G Logic
and ship propulsion system drives.
• To design industrial systems, such as compensation Figure 3 System with fault
controllers, power electronic drives, electric furnaces V1
and filters. V2
• To study the transient and harmonic impact of V1A
1 1 V2A
distributed generation systems such as wind and
V1a V2a
micro-turbine systems on the power grid.
2 V1B 2 V2B
• To study and mitigate capacitor switching transients.
V1b V2b
• To study the system impact of transmission line
3 V1C 3 V2C
imbalances during contingency periods.
V1c V2c

100
 2013 International Conference on Power, Energy and Control (ICPEC)

Figure 4 Figure 5 Z1 = 52.9 [ohm] /_ 80.0 [°]

5000.0 [MVA]
V Ph 230.0 [kV], 60.0 [Hz]
SUBSTATION 1 100.0 km SUBSTATION 2
TLine
10000.0
230.0 [kV], 60.0 [Hz] Ph V
5000.0 [MVA]
V1 V2

1 .0
Z1 = 52.9 [ohm] /_ 80.0 [°]

Timed Com.
Fault Bus
Figure 6 ABC->G Logic AM
VOLTAGE 1 GM
V1a V1b V1c
400
300
200
AO
100
KB
0 6 Pulse
y

-100 Bridge
-200
-300 TIME
0.160 0.180 0.200 0.220 0.240 0.260 0.280 0.300 0.320 0.340 0.360 .
.
.
Figure 10 Systems with TCSC
Figure 7 Voltages at Bus Bar 1 V1 V2

VOLTAGE 2
1 V1A 1 V2A
400
V2a V2b V2c V1a V2a
300 2 V1B 2 V2B
200 V1b V2b
100
3 V1C 3 V2C
0
y

-100 V1c V2c

-200 Figure 11 Figure 12
-300
0.150 0.175 0.200 0.225 0.250 0.275 0.300 0.325 0.350 0.375 .
.
.

Figure 8 Voltages at Bus Bar 2

FAULT CURRENT
Ia Ib Ic
10.0
8.0 Figure 13
6.0
4.0
2.0
0.0
y

-2.0 VOLTAGE 1
-4.0
-6.0 V1a V1b V1c
300
-8.0
200
0.180 0.200 0.220 0.240 0.260 0.280 0.300 0.320 0.340 .
. 100
.
0
y

Figure 9 Fault Current -100

-200
-300
0.150 0.175 0.200 0.225 0.250 0.275 0.300 0.325 0.350 .
.
.

Figure 14 Voltages at Bus Bar 1

101
 2013 International Conference on Power, Energy and Control (ICPEC)

VOLTAGE 2 V1
V2a V2b V2c V2
250
200
150 1 V1A
100 1 V2A
50 V1a
0 V2a
-50 2 V1B
y

-100 2 V2B
-150 V1b V2b
-200
-250 3 V1C 3 V2C
0.150 0.175 0.200 0.225 0.250 0.275 0.300 0.325 0.350 0.375 V1c V2c
Figure 18 Figure 19

Figure 15 Voltages at Bus Bar 2

FAULT CURRENT
Ia Ib Ic
10.0
8.0
6.0
4.0
2.0
0.0
y

-2.0
-4.0 Figure 20
-6.0
-8.0
0.180 0.200 0.220 0.240 0.260 0.280 0.300 0.320 0.340 VOLTAGE 1
V1a V1b V1c
300
Figure 16 Fault Current 200
Z1 = 52.9 [ohm] /_ 80.0 [°] 100
5000.0 [MVA] 0
V Ph 230.0 [kV], 60.0 [Hz]
y

SUBSTATION 1 SUBSTATION 2 -100

100.0 km
TLine -200
10000.0 10000.0
230.0 [kV], 60.0 [Hz] V1 V2 Ph V -300
1 .0

1 .0

5000.0 [MVA] Com. 0.150 0.175 0.200 0.225 0.250 0.275 0.300 0.325 0.350 .
Z1 = 52.9 [ohm] /_ 80.0 [°] Bus .
AM .
GM
Timed Figure 21 Voltages at Bus Bar 2
Fault
ABC->G Logic

AO VOLTAGE 2
Com. KB V2a V2b V2c
Bus 6 Pulse 300
AM Bridge
200
GM
100
TIME
0
y

-100
AO -200
KB -300
6 Pulse
Bridge 0.160 0.180 0.200 0.220 0.240 0.260 0.280 0.300 0.320 0.340 .
.
.
TIME
Figure 22 Voltages at Bus Bar 2

Figure 17 Systems with Two Series TCSC Module

ANALYSIS:
Three phase fault
Time to apply fault = 0.2sec
Duration of fault = 0.1 sec

WITHOUT TCSC:
Maximum Voltage V1 about 300 V
Maximum Voltage V2 about 300 V
Damping of Distortions at 0.35 sec

102
 2013 International Conference on Power, Energy and Control (ICPEC)

VI. CONCLUSION
WITH TCSC: In this paper, an efficient PSCAD based simulation
Maximum Voltage V1 about 250 V using TCSC has been solved the power oscillation damping
Maximum Voltage V2 about 250 V problem. TCSC is very necessary when National Grid is
Damping of Distortions at 0.32 sec formed, to channel the power properly with firing angle
control mechanism. By controlling the firing angle of the
WITH TWO SERIES TCSC: thyristor valve from 180 degree to lower values the effective
Maximum Voltage V1 about 240 V impedance of the TCSC can be varied from capacitor
Maximum Voltage V2 about 240 V impedance to 3 times the capacitor impedance. During power
Damping of Distortions at 0.31 sec oscillations the impedance can be varied very fast between the
capacitor impedance and the maximum impedance.
vca

A A A A A The TCSC consists of a number of series connected

1.0 1.0 1.0 0.1 1.0
vcb

B B B B B modules, each having capacitor bank with a parallel thyristor

1.0 1.0 0.1 1.0
controlled inductor that circulates current pulses which add in
vcc

C C C C C
g1
1.0 1.0 0.1 1.0
phase with the line current. If the additional voltage created by
BRK

the circulating current pulses is controlled to be proportional

T1
C
A

B
B

A
Power

Power

to the line current, the transmission system will perceive the

Q

Q
P

TCSC as having a virtually increased reactance beyond the

1 .0

1 .0

1 .0

T2

g2 g3
physical reactance of the capacitor. Thus TCSC is an effective
of impedance control to transmit more power. Also it is a
T3

powerful means of damping power swings at the transient and

dynamic basis. The proposed approach is tested in simple two
bus system and results are shown.
T4

g4 g5

VII. REFERENCES
T5
T6

[1] Chi-Shan Yu; Chih Wen Liu " A practical design of TCSC controllers
g6 for the inter-area transient stability control using real-time
measurements”, Power Engineering Society Winter Meeting, IEEE
Figure 23 Electric Systems with a Circuit Breaker Using TCSC Volume: 1 , Page(s): 661 - 666 vol.1 , 1999.
ACTIVE POWER [2] Dubey G. K, Doradla S. R, Joshi A and Sinha R. M. K, “Thyristor
2.0k
P1 Power Controllers”, New Age International Publishers, 2005.
0.0 [3] Fan L, Feliachi A and Schoder K, “Selection and design of A TCSC
-2.0k control signal in damping power system inter-area oscillations for
-4.0k multiple operating conditions.” Electrical Power & Energy Systs., vol-
-6.0k 62, pp. 127-137, 2002.
y

-8.0k [4] Jaewon Chang , Troy NY, Chow J.H."Time-optimal series capacitor
-10.0k control for damping interarea modes in interconnected power systems”,
-12.0k IEEE Power & Energy Society, Volume: 12 , Issue: 1 Page(s): 215 –
-14.0k
221,1997.
0.00 0.10 0.20 0.30 0.40 0.50 .
.
.
[5] Klein M, Rogers G. J and Kundur P, “A Fundamental Study of Inter-
Area Oscillations in Power Systems,” IEEE Transactions on Power
Figure 24 Active Power Systems, Vol. 6, No. 3, pp. 914-921, 1991.
[6] Mathur R. M, Verma R. K, "Thyristor based FACTS controllers for
REACTIVE POWER
electrical transmission systems", IEEE Press, 2002, pp 277-288.
0.2k
Q1 [7] Misra S. C, Prashar V. K, Jha I. S, Seema Gupta and Thiagarajan V,
“TCSC – Major role to play synchronous interconnections between two
0.0
regions”, 11th NPSC, India, 2000.
-0.2k
[8] Narain G. Hingorani and Laszlo Gyugyi,, “Understanding FACTS,”
-0.4k Standard Publishers Distributors, 2001.
KW

-0.6k [9] Olimpo Anaya-Lara and Acha E, “Modelling and Analysis of Custom
-0.8k Power Systems by PSCAD/EMTDC”, IEEE Transaction on power
-1.0k delivery, Vol 12, No. 1, pp. 266-272, Jan 2002..
-1.2k
0.00 0.10 0.20 0.30 0.40 0.50 .
.
.

Figure 25 Reactive Power

103
 Selection of TCSC Parameters: Capacitor and
Inductor
S. Meikandasivam1, Rajesh Kumar Nema2, and Shailendra Kumar Jain3
Abstract-- Thyristor controlled series compensator (TCSC)
device is a series compensator to govern the power flow by
compensating the reactance of the transmission line. Both
capacitive and inductive reactance compensations are possible by
proper selection of capacitor and inductor values of the TCSC
device which can be realized through reactance equation. This
paper presents a conceptual study to adopt an apt value for the
capacitor and inductor of a TCSC device. Degree of Series
compensation (K) brings an idea of selecting the TCSC capacitor.
The Reactance characteristic curve is analyzed with resonant
point for different values of ‘’ which defines the square root of
ratio between capacitive and inductive reactance. The study is
helpful in selecting an appropriate value for TCSC inductor
which has influence on multi resonance point in TCSC device.
Additionally, power flows in the transmission line are discussed
with tabulation to realize the operating span of the TCSC for
various values of ‘’. While designing, TCSC parameters can
also be selected as per requirement on end side demands and for
safe operating span of a TCSC.
Index Terms-- TCSC device, Degree of series compensation,
Reactance Characteristics Curve, Resonance point, Power
Transfers.
I. INTRODUCTION
S ERIES Capacitive Compensation is an old and economic
technique to increase the power transfer capability of a
long transmission line. Since 1950, fixed series capacitors
were installing in long alternating-current transmission lines to
cancel a part of the inherent inductive reactance [1]. If 70%
line inductive reactance is canceled, then line is said to have
70% of series compensation and degree of series
compensation (K) is 70%. 100% compensation will raise a
problem of series resonance in the system. Usually, up to 70%
of compensation is chosen for transmission line compensation.
Moreover, series capacitive compensation technique may
inadvertently increase the risk of subsynchronous resonance
(SSR) problem [1-3].
Because of fixed series compensation, two series SSR
problem’s were occurred in the year 1970s, at the Mojave in
southern Nevada, experienced shaft damage that required
1
Research Scholar – meikandasivams@gmail.com
2
Associate Professor – rk_nema@yahoo.com
3
Associate Professor – sjain68@gmail.com
Department of Electrical Engineering, Maulana Azad National Institute of
Technology, Bhopal, INDIA
978-1-4244-7882-8/11/$26.00 ©2011 IEEE
several months of repairs [4]. In 1981 N.G. Hingorani
proposed a NGH damping scheme for series compensators,
named as thyristor controlled series capacitor (TCSC) which
has been proven to provide effective SSR mitigation. The
world’s first TCSC was commissioned in 1999 for SSR
mitigation in 400 kV grid in Sweden, [2, 3].
Apart from series resonance and SSR problem, there is a
possibility of parallel resonance due to presence of capacitor
and inductor in TCSC device. Hence optimal selection of
TCSC capacitor and inductor is very important.
Guowen H U referred a factor ‘’, in reactance equation,
should be less than 3 [5]; there will be only one resonance
point in the work region between 90o to 180o of firing angle.
Also it was explained that for larger/smaller values of
inductance how the capacitance operating region shifts in the
characteristics curve. For further investigation, this paper
starts with basic discussion of TCSC device and its reactance
characteristic curve. Securely pointing out the measures for
selecting the TCSC parameters and explains about shifting of
resonance region for different values of ‘’. Considering
Kanpur – Ballabhgarh TCSC project as a case study, possible
range of power flows are tabulated for various values of ‘’.
The main purpose of this paper is to optimally choose the
capacitor and inductor for a TCSC device. Section II, brings
out the operation of TCSC along with numerical equations.
Section III gives a reactance characteristics curve of a TCSC
device and specifies the range of inductance and capacitance
region. Section IV, V and VI discusses the condition for
selecting the TCSC parameters with single & multiple
resonance points by simulation. Section VII analyzes the
power transfers on transmission line for different values of ‘’
under various firing angles of TCSC.
II. OPERATION OF TCSC
Fig. 1 shows a TCSC which consists of a series
compensating capacitor (C) shunted by a thyristor controlled
reactor (TCR). TCR is a variable inductive reactor (XL())
tuned at firing angle , as shown in Fig. 2. The variation of XL
with respect to  can be given as [2, 3] –

X L ()=X L (1)
-2-sin2
1
XC = (2)
2fC
 C Thus, reactance characteristics of TCSC shows, operation
in both capacitive and inductive regions through variation of
T1
firing angle () as shown below –
Range of firing angle () Region
L 90 <  < Llim Inductive region
Clim <  < 180 Capacitive region
T2
Llim <  < Clim Resonance region
Fig.1 A schematic diagram of TCSC device.

T1
XL

Capacitive region
L L
Clim to 180
Varies from XL to infinity
T2

Firing ang Resonance region Firing ang

Fig: 2 Equivalent circuit of TCR
90º 180º
For the variation of  from 0 to 90, XL() varies from
actual reactance (XL) to infinity. This controlled reactor is Inductive region
connected across the series capacitor, so that the variable 90 to Llim
capacitive reactance, as shown in Fig. 3, is possible across the
TCSC which modify the transmission line impedance.
XC
Effective TCSC reactance XTCSC with respect to alpha () can
be given as [6-8] -
Fig: 4 Reactance Vs firing angle characteristic curve.
X TCSC ()=-X C +C1 (2(-)+sin(2(-)))
-C2 cos 2 (-)(tan((-))- (3) IV. SELECTION OF CAPACITOR AND INDUCTOR VALUES
tan(-)) An appropriate value for capacitor and inductor of a TCSC
Where device is based on the net reactance of the transmission line
X C +X LC and expected power demands in future.
C1 = (4) Capacitor value is chosen by a degree of series

compensation (K). Normally up to 70% of line reactance is
X 2LC chosen for compensation. Sub section ‘A’ gives an idea of
C2 = 4 (5)
XL selecting the degree of series compensation.
XC X L Choice of Inductor depends on the length of operating area
X LC = (6) required for inductive and capacitive region. It is perfectly
X C -X L
decided by a factor ‘’, given in (7) by shifting the position of
XC resonance region. In further sections, more about the effect of
= (7)
XL variation of ‘’ and multi resonances are analyzed.
For selecting a capacitor and inductor for a TCSC device,
C C
three main significant measures have to be considered,
T1 C A. Degree of series Compensation (K)
Degree of series compensation (K) is a ratio of effective
L L
Varies from XL to infinity
reactance of TCSC [XTCSC()] to net reactance of transmission
T2 line [XTL].
Fig: 3 Equivalent circuit of TCSC X TCSC (α )
K= (0 < K < 1) (8)
X TL
III. REACTANCE CHARACTERISTIC
Fig. 4 shows the reactance characteristics curve of a TCSC While choosing K, 100% compensation should not be
device. It is drawn between effective reactance of TCSC and provided to avoid series resonance in transmission line.
firing angle . The effective reactance (XTCSC()) of TCSC Practically up to 70% of series compensation is chosen for
operates in three region, inductive region, capacitive region line reactance compensation [1, 2].
and resonance region. Inductive region starts increasing from B. Inductive reactance, XL should be sufficiently smaller than
TCR reactance XL||XC value to infinity (parallel resonance capacitive reactance, XC [2]
condition, XL() = XC), and decreasing from infinity to
For selecting the TCSC inductor value, XL should be
capacitive reactance XC for capacitive region. Between the
sufficiently smaller than the XC to get both effective inductive
two regions, resonance occurs.
and capacitive reactance region in the TCSC device.
 Consider an example of shunt network, XC smaller than XL section IV.B. Thus ‘’ ≤ 1 are not permissible to get
and from the principle of circuit theory the effective reactance combined effect of inductive and capacitive region in TCSC.
(Xeff) of the shunt combination follows the smaller value of
the two reactances as shown in Fig. 5a and (-) sign shows Xeff TABLE I
is in capacitive nature. Hence only one capacitive region is SHIFT IN RESONANCE REGION FOR VARIOUS VALUES OF OMEGA
possible between 90o to 180o of firing angle in the reactance
characteristics curve. Resonance Region CAPACITIVE REGION
BETWEEN THE FIRING
ω = sqrt( X C / X L ) between the Firing
XC= -10
Angles (degrees) ANGLES (DEGREES)

1.4 115-116 116-180

1.6 123-124 124-180
Xeff= -20
1.8 129-131 131-180
XL=20  (a) 2.0 134-136 136-180
XC= -10 2.2 139-140 140-180
2.4 142-143 143-180
2.6 145-146 146-180
Xeff=  2.8 147-148 148-180
(b) 3.0 149-151 151-180
XL=10 
XC= -10
Further analyzing the factor ‘’ from 1.4 (slightly greater
than 1) to 3, resonance region shifts from its position and span
Xeff=2.5 
of capacitive region; shrinks. Table I shows occurrence of
XL=2 
(c)
resonance and capacitive regions between 90o to 180o of firing
Fig: 5 Reactance Variation in shunt network.
angles on the TCSC for different values of ‘’. Because of
If XL is equal to XC value as shown in Fig. 5b; then shifts in resonance region, operating area of inductive region
resonance occurs that result in infinite reactance which is an and capacitive region varies. For  = 1.4, capacitance region
unacceptable condition. is larger i.e., from 180o to 116o of firing angle and for ‘’ = 3,
Finally by choosing XL smaller than XC; makes effective capacitance region shrinks to small i.e., from 180o to 151o of
reactance has inductive reactance as like Fig. 5c. Therefore firing angle. As mentioned in [5] i.e.,  > 3, multi resonance
XTCSC() varies from inductive region to capacitive region points occur in the reactance characteristics curves which
with respect to 90o to 180o of firing angle. reduces the operating span of TCSC device.
While varying the XTCSC(), the condition should not be
allowed to occur ‘XL() = XC’, a resonance condition . VI. SIMULATIONS FOR SINGLE RESONANT POINT REACTANCE
CHARACTERISTIC CURVE.
C. Occurrence of multi resonance condition 50
TCSC Reactance Characteristic Curve for w =1.4 to 3.0
w=1.4
Between the inductive and capacitive region, a resonance 40
w=1.6
w=1.8

region occurs. Occurrence of resonance in TCSC device is 30

w=2.0
w=2.2
w=2.4

unavoidable. However, only one resonant region, explicitly 20

w=2.6
w=2.8
w=3.0
one capacitive range and one inductive range, is allowable.
TCSC Reactance in %

10

Multiple resonant points will reduce the operating range of the 0

TCSC. Thus, while selecting TCSC parameters, some -10

measures have to be taken to ensure only one resonant point -20

between 90o to 180o of . -30

Apart from these three factors, thermal loading limit of a -40

transmission line, maximum allowable current, etc., are also -50

considered while choosing capacitor and Inductor. But the -60

90 100 110 120 130 140 150 160 170 180

study concentrates only on the above three factors. Firing angle in Deg.

Fig: 6 Single Resonance Reactance characteristic curve for  = 1.4 to 3

V. ANALYSIS OF RESONANCE POINT
From (7), it is clear that factor ‘’ decides the range of
inductive and capacitive region in TCSC and numerous
resonance points which decrease the operating range of TCSC.
From [5], author referred that factor ‘’ should be less than 3.
XC
ω= <3 (8)
XL
If both XC and XL are same, then factor ‘’ becomes 1 and
meets resonance condition. When it is less than 1, XC is lesser
than XL, only capacitance region is possible as mentioned in
Once the capacitive compensation for a transmission line is
specified through degree of series compensation (K), the main
factor influencing Inductor value and resonance in the TCSC
is ‘’. To verify the above theoretical analysis, the simulation
has been carried out by considering Kanpur-Ballabhgarh
TCSC’s Capacitor; C = 306F and for different values of ‘’.
For =1.4 to 3, creates Single resonant point in Reactance
characteristic curves as shown in Fig. 6. As per the
requirement of degree of series compensation or power flows
in the line, optimum value of ‘’ is selected from 1.4 to 3 for
choosing the TCSC inductor value.
Fig. 7 shows only the capacitive region for ‘’ is less than
1. Reactance varies from -28.9 to -10.4 ohms. Negative sign
indicates for capacitive reactance.
Fig. 8 shows multi resonance points in reactance curve, for K 27% of FC B
‘’ is greater than 3. Between the firing angles 90o to 180o,
two resonance points occurred which reduces the span of the 8% - TCSC XTL
TCSC operation.

TCSC Reactance Curve for w < 1

-8 Fig. 9 Transmission line between Kanpur –Ballabhgarh with TCSC.
-10
Only Capac itive Re gion TABLE II
POWER FLOW PKB AT TRANSMISSION ANGLE 90O WITH FC
-12

Net Transmission Net Reactance

Line Reactance PKB (MW) including FC PKB (MW)
TCSC Reactance in %

-14

(ohms) (ohms)
-16
131.1929 406.5 96.0981 555.0
-18
KB Transmission Line Power Angle Curves With & Without FC
600
Without FC
-20 With FC
X: 90
Y: 552.2
500
-22

X: 90
Y: 405.1
-24 400
90 100 110 120 130 140 150 160 170 180
Firing angle in Deg.

Power in MW
Fig: 7 Reactance characteristic curve for  < 1(only capacitive region)
300

w>3 - Multi Resonance Point-TCSC Reactance Characteristics

100 200

O c c uranc e of Two
Res onanc e Points 100
50

X: 105
TCSC Reactance in %

Y: 0.05824
0
0 20 40 60 80 100 120 140 160 180
0 Delta in Deg.
X: 104
Y: -0.07175

Fig. 10 Power angle curves of KB transmission line with and without FC.
-50 O c c uranc e of c ross ove r be twee n
Capac itanc e to Induc tanc e Power angle curves for Kanpur-Ballabhgarh transmission
line and along with FC are shown in Fig. 11. Table III shows
-100
the possible power flows for various values of ‘’ from 1.4 to
3 and also Kanpur-Ballabhgarh line power flows are also
-150
90 100 110 120 130 140
Firing angle in Deg.
150 160 170 180 mentioned at ‘’ = 2.7432. Data which are highlighted in
Fig: 8 Multi Resonance Reactance characteristic curve for  > 3 Table III are in capacitive mode of TCSC and rests are in
inductive mode. From the analysis, capacitive mode power
VII. CASE STUDY ON KANPUR - BALLABHGARH TCSC flows are greater than 555 MW (power at which line along
PROJECT with FC) and inductive modes are lesser, also shown in same
Fig. 10. Data specified in bold are power flows nearer to
First TCSC project commissioned in India is between
resonance region which are not to trigger on those firing
Kanpur-Ballabhgarh 3-phase, 400 kV, 400 km long
angles.
transmission line to increase the power flow, approximately
from 400 MW to 600 MW[10]. Fig. 9 shows the single phase
equivalent to Kanpur-Ballabhgarh transmission line. The TABLE III.
POWER FLOW PKB AT TRANSMISSION ANGLE 90O WITH FC +TCSC
transmission line has line reactance of 1.044 mH/km, fixed
Power Transfer PKB at transmission angle 90o with FC
series compensation of 27% (Capacitor of 90.7μF) and TCSC and TCSC (MW)
of 8 to 20% of compensation (Capacitor is 306 μF and =
Firing Firing Firing Firing Firing Firing
Inductor is 4.4 mH) was chosen [12] with factor ‘’ is 2.7432. XC/XL angle angle angle angle angle angle
Considering Kanpur-Ballabhgarh transmission line for 90o 108o 126o 144o 162o 180o
1.4 498.7 396.0 818.0 644.0 624.4 622.4
TCSC characteristics study, Table II shows the net reactance 1.6 519.0 479.8 -363.2 655.8 625.2 622.4
and power flow PKB between the transmission line with and 1.8 529.4 508.7 36.7 675.8 626.0 622.4
without fixed capacitor [9]. Maximum Power transfer on the 2.0 535.7 523.2 469.8 715.6 627.0 622.4
line without any compensation is 406.5 MW (approximately) 2.2 539.8 531.9 505.0 831.2 628.3 622.4
2.4 542.7 537.6 522.7 441.6 629.7 622.4
and along with fixed series compensation (FC) is 555 MW. 2.6 544.8 541.7 533.3 31.4 631.5 622.4
Also shown through power angle curves in Fig. 10 2.743 546.3 544.9 540.4 458.0 633.7 622.4
2.8 546.3 544.9 540.4 458.0 633.7 622.4
3 547.6 547.5 545.5 502.9 636.3 622.4
 Table III shows maximum power flows between 90o to thyristor controlled series compensation”, 30th Annual Conference
IEEE Industrial Electronics Society, vol. 2, pp. 1128- 1132, 2004.
180o of firing angles for various values of ‘’. For  = 1.4, [6] C. R. Fuerte-Esquivel, E. Acha, and H. Ambriz-PBrez, “A Thyristor
possible variation of capacitive mode power flows are from Controlled Series Compensator Model for the Power Flow Solution of
622.4 MW to 818 MW between 180o to 126o of firing angle Practical Power Networks”, TEEE Transactions on Power Systems. vol.
15, no. 1, pp. 58- 64, February 2000.
and change in power with respect to firing angle is smaller. At
[7] Geng juncheng, Tong luyuan, and Wang Zhonghong, and Ge Jun,
 =3, variation is from 622.4 MW to 636.6 MW for 180o to “Mathematical Model for describing characteristics of TCSC”, IEEE
162o of firing angle and change in power is greater. From the 2002 PP- 14981-502
analysis, fine tuning for improving the power flow is possible [8] L. F. W. de Souza, E. H. Watanabe, J . E. R. Alves, and L. A. S.
Pilotto, “Thyristor and Gate Controlled Series Capacitors Comparison of
at ‘’ is small, but effect of harmonics has to be considered. Components Rating”, IEEE, 2003, pp: 2542-2547.
Also firing angle at 90o, inductive mode power flow is varied [9] Bijoy Kumar Sahoo and Narendra Kumar, “Power Flow Control Using
from 498.7 MW to 547.6 MW for various values of ‘’; 1.4 to TCSC”, M.Tech dissertation, Dept. of Electrical Engineering, Delhi
college of Engineering, 2005-2006.
3. From the discussion, depends on the requirement of end [10] Arunachalam M, Ghamandi Lal, Rajiv C G, BHEL, Bangalore Babu
side demands, ‘’ has been selected. Narayanan MM, CPRI, Bangalore, India, performance verification of
2200
Power Angle Curves with TCSC K=8%, w=1.4 TCSC control and protection equipment using RTDS, 15th PSCC, Liege,
X: 90
Y: 2169
2005,pp 22-26.
2000

1800
X. BIOGRAPHIES
1600 Powe r c urve s with
& without FC
Power in MW

1400

1200

Induc tive Mode Capac itive Mode

1000
Powe r Curve s Power Curves

800

X: 90
Y: 552.2
600
X: 90
Y: 405.1
400

X: 90
200 Y: 333

0
0 20 40 60
Fig. 11 Power angle curves at firing angle 108 and 162 for various ‘’
VIII. CONCLUSION
This paper discusses about the operation, reactance
characteristic and resonance condition of TCSC. It
investigates the condition of single and multi resonance points
for different values of ‘’. From the report of resonance
behavior, it is concluded that ‘’ should be optimum between
1.4 to 3 to select an appropriate value of inductor and
capacitor.
For the selected value of ‘’, study has been conducted to
simulate power transfers over Kanpur–Ballabhgarh
transmission line. Possible power flows on the Transmission
line are tabulated in both capacitive and inductive modes of
TCSC for various values of ‘’.
This focus an idea for selecting the TCSC parameters:
inductor and capacitor as per the requirement of power
demands on end side. Above study is based on consideration
of degree of series compensation and ‘’, but not considering
the thermal loading of the transmission line, maximum
allowable current limit and TCSC loop current limit etc.
80 100 120 140 160 180
Delta in Deg.
o
S. Meikandasivam was received the Bachelor Degree in Electrical and
Electronics Engineering in 2002 and Master Degree (Power Systems) in 2005.
He also worked as a Lecturer in EEE dept. of Sri Chandrasekharendra
Saraswati Viswa Mahavidyalaya (Deemed University), Kanchipuram. Now he
is pursuing the PhD Degree at Maulana Azad National Institute of
Technology (MANIT), Bhopal, India. Currently he is working as Assistant
Professor in VIT University, Vellore.
Rajesh Kumar Nema was received his PhD degree in Electrical Engineering
from Barkatullah University, Bhopal, India in 2004. He is currently AP at the
Depart. of Electrical Engg., MANIT, Bhopal, India. His current research
interest include power conditioning unit for Renewable Energy storage system
particularly Solar Energy, Hybrid Energy Systems, Grid Interconnection of
Renewable Energy sources.
Shailendra Kumar Jain received his B.E.(Elect.), M.E.(Power Elex), Ph.D.
degree in 1990, 1994 and 2003 respectively and his PDF at UWO London,
ON, Canada in 2007. He is currently working as Assistant Professor at the
Department of Electrical Engineering, NIT, Bhopal, India. He has been
awarded “Career Award for Young Teachers” by AICTE New Delhi, India for
the year 2003-2004. His research interests include power electronics and
electric drives, power quality improvement, active power filters, high-power
factor converters and fuel cell based distributed generation.

IX. REFERENCES
[1] Arthur R. Bergen Vijay Vittal, “Power System Analysis”, Second
Edition, Pearson Education Asia, pp 528-529.
[2] N. G. Hingorani and L. Gyugyi, Understanding FACTS Concepts and
Technology of Flexible AC Transmission Systems, IEEE Press,
[3] R. M. Mathur and R.K. Varma, Thyristor based FACTS controllers for
Electrical transmission systems, John Wiley & Sons Inc.,
[4] http://www.answers.com/topic/subsynchronous-
resonance?cat=technology
[5] Guowen H U, Ming CHENG, and Guilong CAI, “Relation between
Fundamental Frequency Equivalent Impedance and Resonant Point for

Mitigation of Subsynchronous Oscillations in a
Series Compensated Wind Farm with
Thyristor Controlled Series Capacitor (TCSC)
Rajiv. K. Varma, Member, IEEE, Ysni Semsedini, Non Member and Soubhik Auddy, Student
Member, IEEE
Abstract— Wind power penetration is rapidly growing all over
the world as the alternative, renewable and environment friendly
resource of energy production. With this rapid growth of wind
power, the power systems of future will likely see the integration
of large wind farms with electrical networks that are series
compensated for ensuring stable transmission of bulk power. This
may potentially lead to subsynchronous resonance (SSR) issues.
Although SSR is a well-understood phenomenon that can be
mitigated with FACTS devices, scant information is available on
the SSR problem in a series compensated wind farm. This paper
reports the occurrence and mitigation of SSR caused by induction
generator (IG) effect as well as torsional interactions (TI), in a
series compensated wind farm. SSR suppression is achieved as an
added advantage of a Thyristor Controlled Series Capacitor
(TCSC) actually installed to increase the power transfer
capability of the transmission line. In this study, a wind farm
employing self-excited induction generator (SEIG) is connected to
the grid through a series compensated line. A Thyristor
Controlled Series Capacitor (TCSC) is shown to damp
subsynchronous oscillations when provided with closed loop
current control. Extensive simulations have been carried out
using EMTDC/PSCAD to validate the performance of TCSC in
damping SSR.
Index Terms—Wind power systems, FACTS, Self-excited
Induction Generator (SEIG), Series Compensation,
Subsynchronous Resonance (SSR), Thyristor Controlled
Series Capacitors (TCSC), Open Loop Control, Closed
loop Current Control.
I. INTRODUCTION
W ind Energy Conversion System (WECS) is an emerging
means of energy production throughout the globe. Wind
generating capacities have increased from negligible levels in
the early nineties to over 50 GW today. This shift to wind
energy will inevitably lead to large wind turbine generators
(WTG) being integrated into electric power grids.
Rajiv K. Varma and Soubhik Auddy are Associate Professor and Ph.D.
student, respectively, in Electrical & Computer Engineering Dept., University
of Western Ontario, London, ON, N6A5B9, CANADA
rkvarma@uwo.ca, sauddy@uwo.ca).
Y. Semsedini is in London Hydro, London, ON (e-mail:
Semsediy@londonhydro.com )
1
It will be further necessary to transmit the generated power
through transmission networks that can sustain large power
flows. It is well known that series compensation is an effective
means of increasing power transfer capability of an existing
transmission network.
However, series compensation is shown to cause a highly
detrimental phenomenon called Subsynchronous Resonance
(SSR) in electrical networks. SSR manifests itself in two ways:
the induction generator (IG) effect and torsional interaction
(TI). The induction generator effect is caused by the interplay
of the series compensated network and the generator stator
circuit. Meanwhile, torsional interaction is caused by the
interaction of the mechanical/torsional system and the
electrical network. Flexible AC Transmission Systems
(FACTS) can provide an effective solution to mitigate SSR [1-
3].
A vast body of literature is available on the mitigation of
subsynchronous resonance in a conventional thermal power
plant [1]. However, very little information is presented on the
subsynchronous resonance issues in a series compensated wind
farm [2]. The major difference between WECS and
conventional power plants is that WECS uses induction
generators as opposed to synchronous generators. In this
paper, it is shown that induction generators are also prone to
SSR interplay. Growing subsynchronous oscillations both due
to induction generator (IG) effect as well as torsional
interaction (TI) are shown to be the cause of system instability
in a series compensated wind farm.
A thyristor controlled series capacitor (TCSC) is employed
in place of fixed series capacitors to mitigate the
subsynchronous oscillations, in addition to providing stability
enhancement. It is reported in [3, 4] that a TCSC with an open
loop firing angle control when employed in a transmission
system supplied with conventional steam turbine-driven
synchronous generators, offers more resistance at increased
levels of series compensation at subsynchronous frequencies
while at fundamental frequency the same TCSC offers
capacitive reactance enhancing the power flow through the
transmission line. This shows that TCSC has an inherent
(e-mails:
capability of damping SSR while increasing the power transfer
capability of a transmission line.
This paper reports the effectiveness of a TCSC with a
closed loop current control mode to damp the SSR in a wind

farm. The wind farm is connected to the grid through a series
compensated transmission network. The study system is
derived from the IEEE First Benchmark Model (FBM) and the
wind farm is modeled by a group of coherent induction
generators with shunt capacitance. The entire mechanical
system, comprising the gear train, rotor blades and tower, of
the wind turbine is modeled by a single mass, as their
individual natural resonant frequencies are very close.
Nonlinear time domain simulations of fault studies are carried
out using EMTDC/PSCAD.
The organization of the paper is as follows. The two
mechanisms of SSR - induction generator (IG) effect and
torsional interaction (TI) are briefly described in section II.
Section III outlines the study system configuration; section IV
proves the potential occurrence of SSR both due to IG effect
and TI interactions in a series compensated wind farm.
Section V covers the selection of TCSC parameters. The
performance of the TCSC in mitigating SSR is shown in
section VI; finally section VII concludes the paper.
II. SUBSYNCHRONOUS RESONANCE
Subsynchronous resonance occurs in a power system
network when the mechanical system of the generator
exchanges energy with the electrical network [1, 5]. Series
compensation in the line results in excitation of
subsynchronous currents at electrical frequency
XC
fe = f 0 (1)
XL
Where XC = reactance of the series compensator; XL =
reactance of the line including generator and transformer; and
f0 = the nominal frequency of the power system.
These currents result in rotor torques and currents at the
complementary frequency:
fr = f 0 − fe (2)
These rotor currents result in subsynchronous armature voltage
components which may enhance subsynchronous armature
currents to produce SSR. There are two aspects of the SSR:
(a) Self excitation involving both Induction generator
effect and Torsional Interaction
(b) Transient torque (also called as transient SSR)
A. Induction Generator Effect
Self-excitation of the electrical system alone is caused by
induction generator effect. As the rotating MMF produced by
the subsynchronous frequency armature currents is moving
slower than the speed of the rotor, the resistance of the rotor
(at the subsynchronous frequency viewed from the armature
terminals) is negative as the slip s of the induction generator is
negative, as depicted in the equivalent circuit shown in Fig. 1.
2
Ns − N r (3)
s=
Ns
When the magnitude of this resistance exceeds the sum of the
armature and network resistances at a resonant frequency,
there will be self-excitation.
The induction generator effect can be explained in case of a
wind farm comprising induction generator from its equivalent
circuit drawn in Fig. 1.
R1 jX1 jX2
Rc jXm R2/s
Fig. 1 Equivalent circuit diagram of an induction machine
R1 = stator resistance; X1 = stator leakage reactance; R2 = rotor resistance
referred to stator; X2 = rotor leakage reactance referred to stator; Rc = core-
loss resistance; Xm = magnetizing reactance;
B. Torsional Interactions
This form of self excitation involves both the electrical and
mechanical dynamics. This may occur when the electrical
resonant frequency fe is near the complement of a torsional
resonant frequency fn of the turbine-generator shaft system [1,
5]. The torques at rotor torsional frequencies may then get
amplified and potentially lead to shaft failure.
C. Transient SSR
Transient SSR generally refers to transient torques on
segments of the T-G shaft resulting from subsynchronous
oscillating currents in the network caused by faults or
switching operations. This usually occurs when the
complement of the electrical network resonant frequency gets
closely aligned with one of the torsional natural frequencies.
III. SYSTEM CONFIGURATION
The study system is derived essentially from the IEEE first
benchmark model of SSR studies [10]. A TCSC is replaced
with the fixed series compensator in the system shown in Fig.
2. The capacitive reactance of the TCSC is the same as
variable line series compensation and the inductive reactance
is chosen as to avoid multiple resonant peaks or infinite
impedance offered by TCSC in its impedance-firing angle
characteristics [3]. Section V describes it in more detail. The
detailed system data can be found in Appendix.
 3

Machine Terminal Voltage (p.u.)

0.9

0.8

0.7

0.6
Fig. 2 Wind generator study system with TCSC
0.5

0.4
IV. SSR IN SERIES COMPENSATED WIND FARM
In this section both the induction generator self excitation 4 4.5 5 5.5 6
and torsional interaction effects are studied in the study system Time (s)
Fig. 4 Generator terminal voltage for 100 MW power transfer for 90% series
depicted in Fig. 2. The firing angle control of the TCSC is
compensation
blocked so that the possibility of SSR can be examined in a
non-FACTS equipped series compensated wind farm. A three
phase to ground fault is implemented to study potential of both 1.07

self excitation and transient torque SSR. The power flow 1.06

Generator Rotor Speed (p.u.)

studies are conducted using the DSA Power Tools software
1.05
[11] and the electromagnetic time domain simulations are
performed with the PSCAD/EMTDC software [12] 1.04

1.03
A. Induction Generator Self-excitation Effect
1.02

Studies reported in this section show that there are two 1.01

factors which influence the self-excitation phenomenon of an

1
induction generator. These are: 4 4.5 5
Time (s)
5.5 6

• Power transfer level.
• Level of series compensation.
These factors are addressed separately.
For a power transfer of 100 MW, the induction generator
self excitation effect is not seen to be so prominent. Even
though oscillations are visible in the electromagnetic torque,
machine terminal voltage and rotor speed, they are seen to
decay with time. This is depicted in Fig. 3 - Fig. 5. If the line
resistance is decreased, these oscillations become larger and
continue for a longer duration which is expected of induction
generator self excitation oscillations. However, the line
resistance reduction is not a realistic option and is hence not
reported here.
1
0.5
Electromagnetic Torque (p.u.)
Fig. 5 Generator rotor speed for 100 MW power transfer for 90% series
compensation
The generator power output is now increased to 500 MW.
At this high power level, the induction generator is operated at
a much higher speed over the synchronous speed. This
increased power transfer makes the apparent negative rotor
resistance exceed the sum of total armature and network
resistance. Thus giving rise to subsynchronous oscillations.
The dominant electrical mode in this case is 20.54 Hz.
Next, the series compensation level is varied for 500 MW
generator power output. It is observed that with increasing
series compensation levels the oscillations due to induction
generator self excitation get enhanced making the system
eventually unstable. The electromagnetic torque is depicted for
50%, 65% and 90% in Fig. 6, 7 and 8, respectively. The
electrical frequencies fe in the electromagnetic torque increases

0
according to (1) and consequently, the rotor torque frequency
fr decreases as in (2).
-0.5

-1

-1.5

-2

-2.5
4 4.5 5 5.5 6
Time (s)
Fig. 3 Electromagnetic torque for 100 MW power transfer for 90% series
compensation
 4

1
It is observed that with increasing series compensation
0.5
levels the oscillations due to torsional interaction also get
Electromagnetic torque (p.u.)
0 enhanced making the system eventually unstable. The
-0.5
mechanical torque between Mass 1 and Mass 2 is depicted for
50%, 65% and 90% in Fig. 9, 10 and 11 respectively.
-1

-1.5 2

-2

Mechanical torque between

Mass 1 and Mass 2 (p.u.)
1.5
-2.5
4 4.5 5 5.5 6
Time (s)
1
Fig. 6 Electromagnetic torque for 50% series compensation
1 0.5

0.5
Electromagnetic torque (p.u.)

0
0

-0.5
-0.5
0 5 10 15 20
-1 Time (s)
-1.5 Fig. 9 Mechanical torque between mass 1 and 2 (T12) for 50% series
compensation and fr = 0.8 Hz
-2
3
-2.5

Mechanical torque between

-3

Mass 1 and Mass 2 (p.u.)

2
4 4.5 5 5.5 6
Time (s)
Fig. 7 Electromagnetic torque for 65% series compensation 1

4
0
Electromagnetic torque (p.u.)

2
-1

0
-2
5 10 15 20
-2 Time (s)
Fig. 10 Mechanical torque between mass 1 and 2 (T12) for 65% series
compensation and fn = 0.8 Hz
-4
4
-6
4 4.5 5 5.5 6 3
Mechanical Torque between
Mass 1 and Mass 2 (p.u.)

Time (s)
Fig. 8 Electromagnetic torque for 90% series compensation 2

B. Torsional Interactions 0

A very high level of series compensation is typically needed -1

as indicated by (1) and (2) to generate a rotor torque at a
-2
frequency in the close vicinity of the resonant frequency of the
wind turbine mechanical system (comprising rotor blades and -3
5 10 15 20
gear train). There are in fact two typical resonant frequencies Time (s)
[7-8]: Fig. 11 Mechanical torque between mass 1 and 2 (T12) for 90% series
• 1.1 Hz corresponding to the tower side ways compensation and fn = 0.8 Hz
oscillations.
• 2.5 Hz relating to mechanical system.
V. SELECTION OF TCSC PARAMETERS
A two-mass torsional system model is added to the The thyristor controlled series capacitor (TCSC) increases
induction generator model. Since both the frequencies are very the power transfer capability of a transmission network in
close, only one torsional frequency is considered to represent addition to several other functions [3]. It provides a rapid,
the effect of both these natural frequencies. Hence, Mass 1 continuous control of the transmission line series
models the combined tower and wind turbine mechanical compensation level thus dynamically controlling the power
system and Mass 2 represents the inertia of the generator. flow in the line.

In this application, the TCSC is assumed to be primarily
employed in the network for controlling line reactance and
hence control of power flow. The SSR damping function is
added through constant current control for this study. It is
reported [4] that a TCSC when operating at fundamental
frequency offers a pure capacitive reactance to increase the
power transfer capability of the network. On the other hand the
same TCSC offers resistive impedance at subsynchronous
frequencies which need to be damped. The resistive impedance
of the TCSC increases with the increased boost factor which is
the ratio of the capacitive reactance offered by the TCSC and
the total line reactance. Hence, the TCSC also provides a
resistive damping to the subsynchronous oscillations.
The general configuration of a TCSC is shown below:
Fig. 12 General configuration of a TCSC
For this configuration the equivalent TCSC reactance is
computed as per the following equation:
X C2 2β + sin2β 4X C2 cos 2 β ktankβ − tanβ
X TCSC = X C − +
XC − XL π XC − XL k 2 − 1 π shown for the worst possible operating condition of 500 MW
(4)
where β = angle of advance (before the forward voltage
becomes zero) = π-α; α is the firing angle of the thyristors. It
is noted from (4) that a parallel resonance is created between
XC and XL at the fundamental frequency, corresponding to the
values of firing angle αres, given by:
πω
α res = π - (2m - 1) (5)
2ω r
The different resonances can be reduced to one by proper
choice of k = ωr/ω = √ XC/ XL in the range of 900 < α < 1800.
As XC is known, XL could eventually be calculated once a
proper value of k is chosen to ensure a single resonant peak.
Once the XC and XL values of the TCSC are calculated, a
closed loop current control scheme is employed for proposed
application. The functional block diagram of TCSC closed
loop current control is depicted in Fig. 13.
5
α
Fig. 13 TCSC Constant Current Controller
Here Iref is the pre-fault or pre-contingency current
calculated from EMTDC/PSCAD line-current phasor. The
main current controller is a PI Controller. The parameters of
this controller are also adjusted systematically through
electromagnetic transient simulation studies by hit-trial to get
the minimum settling time or fastest damping criterion.
VI. PERFORMANCE OF TCSC IN THE MITIGATION OF SSR
A TCSC is already installed to improve the power transfer
capability of the transmission network of the study system
shown in Fig 2. The damping performance of this TCSC in
mitigating SSR is examined in this section. Here the TCSC is
operating in a closed-loop current control mode.
A. Damping performance of TCSC for IG effect
In this case also, the damping performance of the TCSC is
power flow and 90% series compensation. The signals
reported to show the damping performance of TCSC are:

Electromagnetic torque of the generator (Te)

Generator rotor speed (Wr)

Generator terminal voltage (Vt)
Fig. 14, 15 and 16 display the generator electromagnetic-
torque, generator rotor speed, machine terminal voltage,
respectively.
3
Without TCSC
2 With TCSC CC
Electromagnetic torque (p.u.)
1
0
-1
-2
-3
-4
-5
4 5 6 7 8 9 10
Time(s)
Fig. 14 Damping of SSO in the electromagnetic torque
 6

1.08 1.6
Without TCSC Without TCSC
With TCSC CC 1.5 With TCSC CC

Generator Rotor Speed (p.u.)

Generator rotor speed (p.u.) 1.06

1.4
1.04
1.3
1.02
1.2
1
1.1

0.98 1

0.96 0.9
4 5 6 7 8 9 10 5 10 15 20
Time(s) Time(s)
Fig. 15 Damping of SSO in the generator speed Fig. 18 Damping of SSO in the rotor speed by TCSC
1.6
Without TCSC
Generator terminal voltage (p.u.)

380
1.4 With TCSC CC

1.2
375
1

Xtcsc (Ohm)
0.8
370
0.6

0.4 365

0.2
4 5 6 7 8 9 10
Time(s) 360
5 10 15 20
Fig. 16 Damping of SSO in the generator terminal voltage Time(s)
Fig. 19 Variation of TCSC reactance in damping of SSO
B. Damping performance of TCSC for TI

Finally the damping of TCSC in mitigating torsional VII. CONCLUSION

interaction is investigated. The signals examined are:

Mechanical torque between Mass 1 and Mass 2. With the rapid growth of wind power penetration into the
power system grid, wind farms will likely be evacuating bulk

Generator rotor speed (Wr)
power through series compensated networks. This will render

TCSC reactance (XTCSC)
the power system vulnerable to SSR. In this paper a thyristor
controlled series capacitor is applied to damp SSR in such a
Fig 17, 18 and 19 demonstrate mechanical torque between series compensated wind farm. The following conclusions are
Mass 1 and 2, generator rotor speed and TCSC susceptance drawn from extensive electromagnetic transient simulation
respectively. It is apparent from the figures that TCSC closed- studies over widely varying levels of series compensation:
loop current control effectively damps the SSO due to TI
effect.
SSR is a potential threat in series compensated wind
4 farms even at realistic levels of series compensation.
Without FACTS
With TCSC CC

A TCSC which is originally employed to improve
3
power transfer capability of the transmission line is
Mechanical Torque between
Mass 1 and Mass 2 (p.u.)

2 sufficient to damp SSR provided it is equipped with

1
proper control scheme.

Closed loop current control of TCSC is found to be
0 effective in damping subsynchronous oscillations both
-1 due to torsional interaction and induction generator
effects, even when the system is subjected to a severe
-2
fault.
-3
5 10 15 20
Time (s)
Fig. 17 Damping of SSO in the mechanical torque between Mass 1 and
Mass 2 by TCSC
 7

Rajiv K. Varma (Member 1996) obtained B.Tech. and

Ph.D. degrees in Electrical Engineering from Indian
Institute of Technology (IIT), Kanpur, India, in 1980
and 1988, respectively. He is currently an Associate
VIII. APPENDIX Professor and Associate Chair-Graduate at the
University of Western Ontario (UWO), Canada. Prior to
A. Generator Data: this position, he was a faculty member in the Electrical
Engineering Department at IIT Kanpur, India, from
Power Rating = 1000 HP; VLL = 26.0 KV; 1989-2001. While in India, he was awarded the
Rs = 0.015 p. u; XLs = 0.091p.u; Rr1 = 0.0507 p. u; Government of India BOYSCAST Young Scientist Fellowship in 1992-93 to
Rr2 = 0.0095 p. u; XL1 = 0.0 p. u; XL2 = 0.0539 p. u; conduct research on Flexible AC Transmission System (FACTS) at the
University of Western Ontario (UWO). He also received the Fulbright grant
Xm12 = 0.1418; HG = 0.5 of the U.S. Educational Foundation in India, to conduct research in FACTS at
B. Torsional System Data: Bonneville Power Administration (B.P.A.), Portland, Oregon, USA, during
May-Aug. 1998. He is the Chair of IEEE Working Group on "FACTS and
HVDC Bibliography" and is active on a number of other IEEE working
Power= 100MW; HT = 12.5 pu; HG = 0.5 pu; KGT = 0.15 pu groups. He has received several Teaching Excellence awards at the Faculty of
Engineering and University level at UWO. His research interests include
FACTS, Power systems stability, Wind power generation systems and
C. TCSC Data: Distribution Automation.

K = Percentage series compensation XC/X∑ =90%, 65%, 50%; Ysni Semsedini received his B. Sc and M. Sc degrees in
XL = Calculated from (4) to avoid multiple resonant peaks in Electrical and Computer Engineering from the University
of Western Ontario, London, Ontario, Canada in 2003 and
TCSC reactance-firing angle characteristics. TCSC Current 2006 respectively. He was a distribution engineer in
Controller: KP = 0.0; KI = 200; Kitchener – Wilmot Hydro, Kitchener, Ontario, Canada
from 2003 to 2004. Currently he is working in London
REFERENCES Hydro, London, ON. His research interests include
modeling and performance studies of wind farms and
FACTS.
[1] IEEE Working Committee Report, “Third Supplement to a Bibliography
for the Study of Subsynchronous Resonance between Rotating
Machines and Power Systems,” IEEE Trans. on Power Systems, vol. 6,
Soubhik Auddy received the B.E. degree in Electrical
no. 2, pp. 830-834, May 1991
Engineering from Jadavpur University, Calcutta, India in
[2] P. Pourbeik, R. J. Koessler, D. L. Dickmander, and W. Wong,
2000 and the M.Tech. degree in Electrical Engineering
“Integration of Large Wind Farms into Utility Grids (Part 2 –
from the Indian Institute of Technology Kanpur, India in
Performance Issues),” in Proc. 2003 IEEE PES General Meeting Conf,
2003. He is currently working towards the Ph.D. degree at
pp. 1520-1525.
the University of Western Ontario, Ontario, Canada. His
[3] R.M. Mathur and R.K. Varma, Thyristor-Based FACTS Controllers for
research interests include FACTS, power system dynamics
Electrical Transmission Systems IEEE Press and Wiley Interscience,
and stability.
New York, USA, Feb. 2002.
[4] W. Zhu, R. Spee, R.R. Mohler, G.C. Alexander, W.A. Mittelstadt, D.
Maratukulam, “An EMTP Study of SSR Mitigation Using the Thyristor
Controlled Series Capacitor”, IEEE Transactions on Power Delivery,
Vol. 10, No. 3, pp. 1479-1485, July 95.
[5] IEEE Committee Report, “Reader’s Guide to Subsynchronous
Resonance,” IEEE Trans. on Power Systems, vol. 7, no. 1, pp. 150-157,
February 1992.
[6] J.F. Hauer, W.A. Mittelstadt, R.J. Piwko, B.L. Damsky, and J.D. Eden,
“Modulation and SSR Tests Performed on the BPA 500kV Thyristor
Controlled Series Capacitor Unit at Slatt Substation”, IEEE Trans.
Power Systems, vol. 11, No.2 May 1996, pp. 801-06
[7] T. Thiringer and J. A. Dahlberg, “Periodic Pulsations from a Three-
Bladed Wind Turbine,” IEEE Trans. on Energy Conversion, vol. 16, no.
2, pp. 128-133, June 2001.
[8] E. N. Henrichsen and P. J. Nolan, “Dynamics and Stability of Wind
Turbine Generators,” IEEE Trans. on Power Apparatus and Systems,
vol. PAS-101, no. 8, pp. 2640-2648, Aug. 1982
[9] R. K. Varma and S. Auddy, “Mitigation of subsynchronous resonance in
a series compensated wind farm using static var compensator,”
presented at the IEEE General Meeting, Montréal, Québec, June 18–22,
2006, Paper 06 GM1272.
[10] IEEE Committee Report, “First Benchmark Model for Computer
Simulation of Subsynchronous Resonance,” IEEE Trans. on Power
Apparatus and Systems ,vol. PAS-96, no. 5, pp. 1565-1572, Sept. / Oct.
1977
[11] DSA Power Tools User manual, Power Tech Lab, Surrey, BC, 2005
[12] EMTDC PSCAD User Manual, HVDC Research Center, Manitoba,
2003.
[13] P. Kundur, Power Systems Stability and Control: New McGraw-Hill,
1994