Professional Documents
Culture Documents
Fierts Benchmark Model For Computor Simulation of Subsynchrnoud Resonance PDF
Fierts Benchmark Model For Computor Simulation of Subsynchrnoud Resonance PDF
ELECTRICAL NE"W
E x t e n s i v e SSR s t u d i e so f t h e N a v a j oP r o j e c t re-
v e a l e dt h a t a s i m p l er a d i a l RLC c i r c u i t ,p r o p e r l y
t u n e d ,c a np r o d u c eb o t ht r a n s i e n t and s e l f - e x c i t a t i o n
problems as s e v e r e as anyobserved intheanalysisof
t h ea c t u a ls y s t e m . The s i n g l el i n ed i a g r a m shown i n
F i g u r e 1 r e p r e s e n t ss u c h a s i m p l e c i r c u i t . T h e c i r c u i t
UNITY
VOLTAGE
1565
TABLE I TABLE I1
Rotor Circuit Parameters Generator Impedances and
For Transient SSR Studies Time Constants
I
Parameter D-axis Q-axis X = 1.79 pu Tdo = 4.3 s
d
3 7 7 4 5.3 0.53 I
Xd = 0.169 pu Tio = 0.032 s
0.326 0.062 Xf
377% 3.1 1.54 Xi = 0.135 pu T' = 0.85 s
k' 0.0055 0.095 qo
a' 1.66 1.58 XL = 0.13 pu T" = 0.05 s
0.13 XL 0.13 qo
X
q
= 1.71 pu 5 = 0.0
Table I shows theimpedances for the rotor circuit
expressed in per uniton the machineMVA base. Current, X' = 0.228 pu
9
voltage, torque and rotor speed are also expressed in
per unit. For transient representation, divide reac- X" = 0.200 pu
q
tance in ohms by 377 to obtain the inductance in
henries.
TABLE I11
The generator standard impedances and time con-
stants from which the circuit parameters of Table I Rotor Spring Mass Parameters
were derived are shown in Table 11. The term XL is
armatureleakagereactance.Circuitparametersare
obtained from the standard impedances and time con- Inertia
Spring
Constant
stants by an iterative process outlinea in the appen-
dix. Where the treatment of these standard machine -
Mass -Shaft H(seconds) K_o pu Torque/rad
impedances and time constants is not mathematically HP 0.092897
reducible to the rotor network definedin Figure 2 and HP-IP 7,277 19.303
Table I, the results will differ. IP 0.155589
13,168 IP-LPA 34.929
Table I11 shows the inertias and spring constants LPA 0.858670
for the spring mass model. Inertia is expressed in 19,618 LPA-LPB 52.038
terms of the inertia constant H based on rated kVA. LPB 0.884215
The base torque is that required at synchronous speed 70 26,713
LPB-GEN .E58
to deliver mechanical power in kilowatts equal to the GEN 0.868495
rated (base) kVA value. Base angle is 377 radians, the 1,064 GEN-EXC 2.822
angle of shaft rotation in one second (the.base time). EXC 0.0342165
The spring constantK is given in per unit where base
spring constant is defined as base torque divided by
base angle. The simple second order torque equation in
this system of units is: *D +
0
T(pu) = 2 H5 + D6 + K8 I
id
The spring constant is also given in per unit torque
ed
perradian. For transientstudies,themechanical
damping is assumed to be zero.
CircuitparametersforFigure3areshown in
Table rV. Note that the elements XL, Xad, and Xaq are
W
the same as for the transient model. The generator
rotor impedances at the torsional frequencies to which
elements of Figure 3 were fitted are shown in Table V. Fig. 3 Rotor Model for Self-ExcitationStudies
1566
Where t h et r e a t m e n to ft h e s er o t o ri m p e d a n c e s i s not turbinedampingsandshaftdampings are n o t o b t a i n a b l e .
m a t h e m a t i c a l l yr e d u c i b l et ot h er o t o rn e t w o r kd e f i n e d However,modaldampings as o b t a i n e d by test are p r o v i d -
i nF i g u r e3a n dT a b l e IV, the results w i l l differ. ed i n terms of t h ed e c r e m e n ft a c t o r un. S i n c et h e s e
v a l u e s are load dependent they are f u r n i s h e d a l o n g w i t h
The r o t o sr p r i n g mass
model
used f o tr r a n s i e n t t h e case d e s c r i p t i o n wherethey c a n b ea d j u s t e df o r
studies is alsousedforself-excitationstudiesexcept load. No l o am
d e c h a n i c adle c r e m e nf ta c t o rfsot rh e
f otrhien c l u s i o n of mechanical
damping.
Individual f i r s t f o u r modes a r e shown i n F i g u r e 4. F u l l l o a d v a l -
uesrangeuptotwentytimeslarger.
As a n a l t e r n a t i v e a n d f o r t h e c o n v e n i e n c e o f t h o s e
w i s h i n gt oc h e c kt h e i rc a l c u l a t i o n s byhand, a modal
TABLE IV mechanicalmodelhasbeenprovidedfor the f i r s t f o u r
modes (see F i g u r e4 ) . The mass Hn h a sb e e na d j u s t e dt o
Rotor C i r c u i t Per Unit Impedance s t o r et h e same mode energy as t h e sixlnass model when
a t 60 Hz f o r S e l f - E x c i t a t i o n S t u d i e s its v e l o c i t yd e v i a t i o nc o r r e s p o n d st ot h a t of t h eg e n -
e r a t o r mass. With n e g l i g i b l ee r r o r ,e l e c t r i c a lt o r q u e s
Mode 14 2 3 when a p p l i e d i n p h a s e w i t h t h e a n g u l a r d i s p l a c e m e n t a n d
32.28
25.55
20.21
Frequency
15.71
(hz) a n g u l a rv e l o c i t yo f mass Hn w i l l c h a n g et h e mode f r e -
quencyand mode dampingby t h e same amount a s i f t h e s e
XL 0.13 0.13 0.13 0.13 t o r q u e sw e r ea p p l i e dt ot h eg e n e r a t o r mass i n t h e six-
mass model.
Rrd PU 0.00587 0.00686 0.00764 0.00825
X r d p' 0.04786 0.04401 0.04080 0.03823 The per u n i ts y s t e m f o rt h er o t o r electrical and
mechanicalmodels i s i d e n t i c a lt ot h a td e s c r i b e df o r
xad PU 1.66 1.66 1.66 1.66
t r a n s i e n ts t u d i e s .C o n s t a n tf i e l dv o l t a g e is assumed.
Rrq PU 0.00884 0.00936 0.00998 0.01081 Electrical torque variations on the exciter are assumed
0.04469 to be zero.
Xrq Pu 0.04742 0.04648 0.04556
Xaq PU 1.58 1.58 1.58 1.58
TRANSIENT CASE DESCRIPTION
TABLE V I
T r a n s i e n t Case D e s c r i p t i o n
Case 1-T
Fault location B ( F i g u r e 1)
-
Mode -
Hn -
fn (hZ) %(no b d )
- Type o f f a u l t Simultaneous 3L-G
I 2.7 15.71 0.05 Prefault phase voltage va = 0
27.8 2 20.21 0.11
6.92 3 25.55 0.028 Clear 1st p h a s e , i = 0 . 0 7a5f ftsae ur l t
4 3.92 32.28 0.028 C l e a r 2nd phase n e x t current z e r o
Clear 3rd phase nextcurrentzero
C a praecaicttoarn c e Xc pu0.371
Fig. 4 Modal MIckonical Spring Mau Modd C a p abcyviupots(oalentrsdaosg
)t e
C a p a c i troeri n s e r t i vo on l t a(gneuoste d )
1567
SELF-EXCITATION CASE DESCRIPTION ator mass and the other member used the modal mechani-
cal model. Their study results are sunmarized below:
Table VI1 shows the infomation required to speci-
fy a self-excitation case. Note that each case per- Six-Mass Modal
tains to the investigation of a single torsional mode. nodel Model
Study results should include one or all of the follow-
ing : 1. Coupled electrical-mechanical
system decrement factor
1. The torsional mode decrement factor Uric
(reciprocal time constant), forthe entire A u (per
~ ~second) -1.525 .-1.503
coupled electrical-mechanical system.
2. Additional mechanical damping for
2. The additional mechanical damping required sustained oscillations
to just obtain sustained torsional oscil-
lation, expressed in terms of incremental A u (per
~ second) 2.743 2.722
decrement factor ADn or incremental b3(per unit) 75.9 15.3
damping factor 4. For consistency,
apply all dash pot damping on the genera- 3. Additional armature resistance for
tor mass of the six-mass model. sustained oscillations
3. The additional armature resistance AR1 ARl(per unit) 0.465
0.461
required to just obtain sustained torsion-
al oscillation. 4. Reduction in series compensation
for sustained oscillations
4. Reduction in network series compensation
AXc to just obtain sustained torsional AXctper unit) 0.0465
0.0470
oscillation.
Capacitor reactance
0.287
Xc pu The chairman wishes to acknowledge the contribu-
Filter none tion of Eli Katz who provided the major portion of the
effort in the-preparation of the benchmark model and
comparison of case studies.
RESULTS OF TRANSIENT CASE 1-T
APPENDIX
Transient response curves for case 1-T based on The rotor network parameters are obtained from the
the rotor model defined in Figure 2 and Table I were generator standard inpedances and time constants by
provided by three task force members using different solution of the following simultaneous equations which
computer programs and problem formulations. The results are based onthe material contained in reference [l] :
were overlayed and found to be in close correspondence.
From among the sixteen curves specified, five were se-
lected for presentation and these are shown in Figure5.
The solid line is a composite of the response curves
submitted by thesethree task force members.
o = +- 1 (6)
RESULTS OF SELF-EXCITATIONCASE 1-S w
Two members of the task force have provided re- o = 1
Xa-377T;i0qd
+- 1 (7)
sults for the self-excitation case. One member used xad
the six-mass model with a single dash pot on the gener-
1568
0
( a ) CapacitorVoltage, A Phase
in
5 0
92 50
3
a
Seconds.
0
( b ) Generator
Current, Phase
A
0 ( c )Torque
Electrical
Generator
0 ( dSl h a f Tt o r q u eL, P A L- P B
,
4
t
W
Seconds
0 GEN-EXC
Torque,
( e ) Shaft
Seconds
9
O_I
Fig. 5 Response
Curves
For
Transient
Case 1-T
1569
1 1 1
o = X -377T" R
+-
fgqfoq ' xkq-377T;0%q Xa q
1 1 + -'adI + -L'1
o = Xfd-377ThRfd ' X,d-377T!&d (9)
1 1
o = xkq-377T'S(
Xfq-377T'R
+ + 2 + I_ (10)
q q fq q 'aq xL
1 1 1 + -
+- 1 (11)
O = t
Xfd-377TxRfd Xkd-377T'&
'ad L
'
1 1 + 1+
o = x (12)
fq-377TiRfq X kq-377T"\q 'aq L
'
REFERENCES
Th = aT h ,
d
1570
Discussion electrical power, which in per-unit would be generator torque, is the
sum of the products of instantaneous phase current and voltage. If this
Yao-nanYu and M. D. Wvong (University of British Columbia, Van- torque is applied to a spring-mass model of the rotor, theresulting tran-
couvei, Canada): The authors are to be congratulated for the excellent sient waveforms are as shown below.
work they have done in developing this first benchmark model for com-
puter simulation of SSR. It is not only useful for comparing computer
programs, but also convenient for stabilizer design. We would appreciate .e
-
0
1571
Peak acceleration of the last LP turbine section following shearing chines such as the Navajo units which, because of their pole face damper
of the generator/turbine coupling bolts can be excessively large. In one windings, are nearly symmetrical. Less accuracy can be expected for
case, which has been examined, [A] the peak acceleration was found to machines not so equipped.
be 1600 rad@, the corresponding acceleration under non-shearing con- T o illustrate how the symmetrical and nonsymmetrical interaction
ditions being 750 rad@. The frequency of the vibration set up in the compare, we have remn the self-excitation case using the modal model
last LP stage by coupling failure was 103 Hz. The hgher acceleration but this timewith rotor circuitparameters based on manufacturer’s
which results on coupling failure may be acceptable provided the struc- data furnished for the same machine before the pole face damper was
tural limitations of the LP blading are not exceeded. It may therefore added. Table VI11 shows the parameters for the new rotor circuit and
be necessary to consider for these higher accelerations pulsating at high the parameters forthe symmetricalcomparison case which was o b
frequency the inertia stresses induced in the blades by movement of tained by averaging the direct and quadrature axis elements. The cir-
their roots. Similar phenomena m g h t be ’induced by subsynchronous cuit series capacitor was increased from .287 per unit to .3 1 per unit to
resonance which could lead to displaced resonant frequencies and addi- compensate forthe addedcircuitinductance andthereby maintain
tional material damping which would effect system response. maximum third mode interaction.
An up to date assessment of this phenomena in relation to sub- Table IX shows the change in third mode decrement factors for
synchronous resonance and related transient peak shaft stresses would these two cases as the line resistance is varied from .01 to .6 per unit.
be timely in this case. Using Mr. Ramey’s equation, the line resistance for zero damping was
REFERENCE calculated to be S 2 2 4 which corresponds closely to that shown for the
symmetrical rotor. For the nonsymmetrical rotor, the comparable line
[AI Hammons, T. J., “Effect of Three-phase Faults and Faulty Syn- resistance is .4429.
chronization on the Mechanical Stressing of Large TurbineGen-
erators”, Rev. Gen. Elect., Vol. 86, (749, July - August 1977, Table IX
pp 558-580. Third Torsional Mode Decrement Factors
Comparison of Symmetric and Nonsymmetric Machines
X,= .31 pu
Eli Katz: Messrs. Yu and Wvong have suggested correctly that exciter
torques are important when dealing with the effects of the power sys- Line Resistance Decrement Factor
tem stabilizer or when designing the exciter controls for damping rotor Per Unit Nonsymmetric Rotor
Symmetric
Rotor
oscillations. The important role played by the exciter in the self-excita- .o 1 - .83243 - .76104
tionproblem is illustrated by the need to remove the power system
stabilizers from service at the Navajo Plant, because of their destabilizing .02 - 1.21123 - 1.00821
tendency, until they could be equipped with filters to desensitize them
to torsional frequencies.
.03 - 1.93901 - 1.45696
As noted in the paper, self-excitation studies require arotor model .030355 - 1.97398
withaccurate rotor impedance atthemodal frequencies. When the .030356 + 1.83941
exciter is permitted to play arole, the modelmust a1.w accurately .036985 - 2.020 12
represent the variation in h d caused by variations in Efd at each tor-
sional frequency. These machine characteristics are not generally avail- .036990 + 1.8896 1
able, but they can be obtained by test or calculated by the machine de- .04 + 1.13938 + 1.63349
signer. When these characteristics are available, I believe that a simple .05 + .76016 + 1.08284
three parallel branch model might be satisfactorily fitted for the study
of one modeat a time. .1 + .23850 + .33285
The variation in field current caused by armature current varia- .2 + .07502 + .lo796
tions need not be precisely modeled so long as thtre is a damper wind- .4 + ,00655 + .01997
ing. Exciter torques produced by this field current are comparatively
inconsequential in both self-excitation and transient
studies. For .442905. 0
transient studies, the exciter torque caused by exciter voltage variation .5 - ,00699 + .00312
cannot be ignored unless the exciter is desensitized to those torsionai 324378 0
frequencies withlarge exciter motion. If this is not done, theexciter can
play a dominant role in shaft torque buildup. .6 - .01603 - .00080
D. G . Ramey presents some very useful approximations which im-
part considerable understanding of the SSR phenomena. His interaction Mr. Ramey’s transient response curves speak for themselves. Un-
equation gives the added generator mass dampbg required for the inter- fortunately they were not available in time to be included in Figure 5
action to just sustain oscillations. Although the equation is derived for along with those submitted by other members of the Task Force.
a machine with a symmetrical rotor, it is also quite accurate for ma- Mr. T. J . Hammons hasprovided an interesting review of the mech-
anical problems related to faulty synchronizing, an allied subject which
Table VI11 the Task Force has elected to exclude from its scope of activity. In his
Mode 3 Rotor Circuit Per Unit Impedances discussion, Mr. Hammons raises the hope of using shear bolts to limit
at 60hz for Self-Excitation Study shaft torque caused by subsynchronousresonance,an idea that also
Pole Face Damper Removed occurred to some of us in 1972 when we were up to our ears in this
problem. However, engineers who are familiar with turbine-generator
Nonsymmetric Rotor Symmetric Rotor design tell us that it is a very bad idea for the following reasons:
XL .13 .13 1. Shear bolt failure would subject the turbine buckets to large
bending stress because of the sudden release of shaft-stored energy.
Rrd BO7841 ,0325455 2. Sudden loss of generator load and inertia would result in ex-
xrd .050268 .135274 cessive turbine overspeed subjecting the longer blades to excessive radial
stress and possible contact with stationaryparts.
xad 1.66 1.62 3. It would be nearly impossible to design the shear bolts so they
would have sufficient strength to withstand normal full load torque and
Rrq .057250 .0325455 atthe same time provide reliable protection against the relatively
.135274 .220280 modest cyclical torques which are capable of causing shaft failure.
xrq Because of the serious consequences of Items 1 and 2 above, we could
xaq 1.58 1.62 not accept the risk of an occasional unnecessary failure of shear bolts.
On behalf of all the task force members, I wish to thank the dis-
Manuscript received April 18,1977. cussers for their thoughtful and informative comments.
1572