ML20043B379
| ML20043B379 | |
| Person / Time | |
|---|---|
| Site: | LaSalle  |
| Issue date: | 05/23/1989 |
| From: | Cheng H, Mallen A, Wulff W BROOKHAVEN NATIONAL LABORATORY |
| To: | |
| Shared Package | |
| ML20042D069 | List: |
| References | |
| FOIA-90-13 BNL-NUREG-42342, NUDOCS 9005290234 | |
| Download: ML20043B379 (18) | |
Text
- - . . . . _ . _ . . . . . . . .
g f
' #To be presented at The Society for Computer Simulation 1989 Eastern Multiconference, Tampa, Florida, March 28-31, 1989 BNL-NUREG-42342 SlHULATIONS OF THE RECENT LASALLE-2 INCIDENT WITH THE BNL PLANT ANALYEER H. S. Cheng, A. N. Hallen, and W. Vulf f Department of Nuclear Energy Brookhaven National Laboratory Upton, New York USA !!973 ABSTRACT EVENT DESCRIPTION This paper presents the resulta of sisulations of Af ter the control systema had been tuned out as the recent power oscillation incident at the LaSa11e-2 the cause of the instability by using the BPA. it was I;Jclear Power Plant using the BNL Plant Analyset. The postulated that the observed growing power oscillation
- cauces of the oscillation were investigated and the was caused by a nuclear thermal hydraulic instability sensitivity of the oscillation to key parameters was brought about by the recirculation pump trip and the studied. It is concluded that the observed power porttal feedwater heater loss plus power peaking.
oscillation was caused by boiling instability (i.e.,
density wave oscillation) reinforced by the reactivity feedback in neutron kinetics, and that the density the transient was initiated from the 851 power wave oscillation resulted f rom flow reduction due to and 75% flow condition by a recirculation pump trip rectreulation pump trip and feedwater teeperature followed by a patttal f eedvater heater loss. Table I reduction due to partial loss of feedwater hosting summarises the sequence of events for the transient.
capab(1ty as well as power peaking.
INTRODUCTION Table l On March 9,1988, an Instrument Maintenance (IH) Sequence nf Events technician at LaSalle Unit 2, while perf orming a func-tional test on a dif ferential pressure switch, caused Time both rectreulation pumps to trip off due to a valving Event / Action (a) error (Diederich 1988). Because of the large and rapid power reduction, feedwater heater high level 1. Steady state et 851 power and 75% flow -5.0 alares caused a partial isolation of feedwater hest-ers, resulting in a reduction of 31 *T in feedwater 2. Rectreulation pumps tripped 0.0 temperature. Approntantely 5 minutes into the event, the local power range sonttor (LPRM) up- and down- 3. Reactor power dropped to 37% 0.4 scale alarms began annunciating and the average power range monitors (APRM) were observed to be oscillating 4. Core flow teached natural ctreulation (29%) 0.5 with an 2.3 s period. Realtning the unit's unfa-vorable location on the power / flow map, the operating 5. Feedwater heaters partially isoisted 1.0 staff was preparing to scree the reactor manually, when an automatic scrae occutted on high-flux trip 6. Reactor power reached 45% & " beat" (118% trip on APRH). Prior to the scree, the opera- phenomena begon 2.0
. tors attempted to tenedy the situation by trying to testart the recirculation pumps, but failed. 7. Modulated limit cycle oscillations continued 5.2 The growing power oscillation observed in this 8. Enhanced itait cycle oscillations began 5.9 incident raises concerns about the stability of BWRs.
The taputtant questions are: Why did the oscillation 9. Crowing oscillations started 7.3 occurt Was there a possibility for divergent oscilla-tionst What if the operator did restart the recircu- 10. Reactor power reached 118%, and reactor 8.1 lation pumpst What if the Main Stese Isolation Valves tripped (MSIV) were inadvertently closed right after the pump restart? Brookhaven Nations! Laboratory (BNL) was 11. End of transient 9.0 asked by the USNRC to sinalate the LaSalle-2 tvent with the BNL Plant Analyser (BPA) (Wulff 1984 and Cheng 1986). This paper reports the results of the BPA staulations as well as the important findings of the present analysis. THE BPA S1HULATION The proper initial conditions are essentist for 9005290234 pOlA 900320 the analysis of this event. A steady-state run was pg g 77c/O-13 first made to obtain the desired inttial condtttons.
The initial conditions obtained by the BNL Plant
- Work perf ormed under the auspices of tM U.S. Nuclear Ar;alyser are sume rized as follows:
Regulatory Commission.
14
\ I .
- 1. Reactot Power 2808 Wt 1. The best-estimate conditions for LaSalle-2 led to Ilmit cycle oscillations only.
- 2. Core Inlet Flow Rate 10.210 kg/s (81 Klb/h)
- 2. The La S411e-2 condtttons within the uncertainty
- 3. System Pressure 69.5 MPa (1007 pota) envelope produced r. roving oscilla-tons le,Jlne i..
aut oma t ic strae as actually observed in the
- 4. Steam Flow Rat e 1424 kg/s (11.3 Mlb/h) event.
- 5. Feedwater Flow Rate 1424 kg/s (11.3 Klb/h) 3. The cause of the LaSalle-2 event was the coupled nuclear thermal-hydraulic instability ortgtnated
- 6. Recirculation Drive Flow 3641 kg/s (28.9 M1b/h) by the density wave oscillations and reinforced Rate by the reactivity f eedbeck. The instability was brought about by the combination of:
- 7. Rectreulation Pump Speed 1590 rpe
. Power peaking (especially radial peaking) ,
- 8. Core Average Void Fraction 41.7 1
- Flow reduction due to the recirculation pump
- 9. Core Average Fuel Temp. 60l*C (Ill3*F) trip.
- 10. Core Average Coolant Te mp . 285'c ( $45*F) . Feedwater temperature reduction d ue to the partial feedwater heater loss,
- 11. Core Inlet Subcooling li'C ( 20*F) 4 The amplitude of the power oscillation r e ma in s
- 5. Reactor should be scrammed when LaSalle-t y pe Reactivity feedback plays an important role in a power oscillations occur.
BWR for both the steady-state and transient analyses.
- 6. These s t ud ie s reinforce the importance of The feedback coefficients used in the SPA simlation continued monitoring and controlling of peaking were obtained from the earlier work on the void feed-back (Cheng 1977) and the Doppler feedback (Cheng factors.
1978). A.x t a l power distribution is known to affect the core instability of a BWR (Yokomiso 1987). The antal power profile used in this work to a typical bot ton-peaked power shape with an axial peak of 1.38.
The results of the SPA simulations are presented in Figures I through 9. Figure 1 presents the simu-lated power oscillation which shows a r e ma r k a ble resemblance to the actuai TARM traces (Kaufman 1988).
The aoomed display of the power oscillation just prior to the au t oma t i c scram is shown in Figure 2, which shows a period of oscillation of -2.8 e as compa r ed to -2.3 e as observed in the actual event. T.at the 1.2 power oscillation is the result of the density wave oscillation is clearly demonstrated by the oscillatory void behavior as shown in Figure 3. ra w 1.0 -- .-
2 Figure 4 presents the steulated core flow re- o g, '"
p.
sponse and Figure 5 shows the soomed display of the core flow oscillation with the s a me period of -2.8 s as the power oscillation. The core flow quickly w 0.6 - -
> j reaches the natural circulation condition within a _dA '3 sinute due to the rectreulation pump trip as shown in W'W 1-Figure 6 for the recirculation pump speed and in s- g 3 ,
Figure 7 for the rectreulation drive flow. C J
w 0.2- -
Figure 8 shows the feedwater flow response and su Ftgure 9 presents the feedwater temperature response k along with the actual plant data. The temperature 0.0 i i i i i
, , i 4 6 7 8 g response shows good ag r eeme n t with the plant data 0 1 2 3 5 except for the value at 1 min. ggggg gggy)
LaSalle-2 F.umt ML Plant esslyzer 814EiB 13:28 St?tStARY AND CONCLt!SIONS Figure l Simulated Reactor Power Oscillation The BPA has been used to simulate the recent for the LaSalle-2 Event.
LaSalle-2 power oset tiation incident. Extensive sen-sitivity studies were pe r f or med . The samlation results support the following conclusions:
t -r 10 1.2 a m
w 1.0 -
) J 8-- '-
o
E
- a. 8 8 "' , 6- --
, k l i l j 1
, j __ . :- :__----_. _
a ,
n g., <-
w s.2- -
o a w L '
C.0 0 , , .
6.2 6.6 7.0 7.4 1.8 82 0 1 2 3 4 5 6 7 8 9 flMt (MIN) flME (MIN) 14 bile-2 twent M Plant emelyse, 814E-E 13:3 145alle-2 turnt M Flant Nlyset 814C 513:N Figure 4 e re Response Figure 2 Zoneed Display of the Power Oset11ation.
50 m t.0 s
- ~
7 "
z a 3.5-> <-
- =
[ 3%h\hN$hMf f' I'I '
g., ..
3
- s 40 , i i > >
' 2.0 , ,
0 1 2 3 4 5 6 7 8 9 6.2 6.6 7.0 7.4 1.0 0.2 1]PE (MlN) 9gpg (n[y)
Lahlte-2 twent M Flint analyse, B14E-E 13:3 145alle-2 turnt M Plant NIyter 814E-E 13:3 Figure 3 Oset11 sting Sebertor of Co're Average Figure 5 Zoomed Display of the Core Flow Oscillation.
Veld Fraction.
l g'p.
=
3 15 m
~
x
., gg -. g j o m 10 - ~
a a E
x 2- -
I E 4 g a 5- l m5-- '- O a
4 0 , , , , , , 0 . , 1 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 1 8
, TIM (MIN) 112 (MIN) tabile2 twent ML Plant finalyse. 814K413:J8 145alle-2 turnt M Plant melyurt B14E-6B 13:28 Figure 6 Recirculation Pump Speed Response. Figure 8 Steulated Feedwater Flow Response f or the La-Salle-2 Event.
3- fiD n
,,, w 0 Plant Sala
^ 3" w - SP6 x
N g, . w gi 9
- m d I ::3 E g. .
p cm
- 10 - w IID -
s ,
o g 4 g w g. .-
w a-0 i i > 2 -
0 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 1 e 1
T!E IMIM) IIE IMIH)
E Plant Analyser 914E413:;8 1.aSalle-2 twat M Plant Analyte, 814E-te 13:28 ta h llr2 twent Figure 7 Rectreulation Drive Flow Response. Figure 9 Feedwater Temperature Response for the LaSalle-2 Evesit REFEPINCES LaSalle County Station. Lettet to N. Ka l t via nak t s .
Cheng. H. S. et al. (1977), "A Space-Time Analysis of March 11. 1988.
Vold Reactivity Feedback in SWRs." Stookhaven National Laboratory, BNL-NUREC-23501. Kauf man, J. and Lantk. C. (1988). " Private Communica.
tion," Otvision of Analysis and Evaluation of operat.
Cheng. H. S. et al. (1978), "BWR Doppler Feedback - ing Data The U.S. Nuclear Regulatory Commission, May Effect of voids. Control Blades, Cadolinta, and Expo- 6, 1988.
sure." Stookhaven National Labotstory, BNL-NUREG.
24433. Wulff, W. et al. (1984). "The RWR Flant Analy z e r ,"
Final Report. Brookhaven National Laboratory, NUREC /
Cheng H. S. (1986). "Botling Water Reactor Plant Ana- CR-394 3. BNL-NUREC-5 t B l2.
lyser Development at Brookhaven National Laboraroty."
Nuct. Set. Eng., 92, 144-156 January 1986. Yokontan, O. et al. (1987). " Space-De pendent Analysis of BWR Core Nuclear Thermal Hydraulle Stahitity." Pro-Diede t t ch . C. J. (1988), "Potentially Significant ceedings of ANS Topical Me e t ing on Anticipated and Eventt Unit 2 Scree initiated by Valving Ettor." Abnormal Transients in Nuclear Power Plants. Vnt. 1 V 11-9. At lanta, r.cu rnia . Aprit 12-15 1987.
' ' --_._..__.-m-___m_____._ _ _ _ _ _ _ _ _ _ . ~ . _ _ _ _ _ _ . _ . , , _ _ _ _ _ . , _ _ . _ . . _ , , . _ _ _ _ _ _ . _ _ . _ _
s en a m .a awe - - - -mms ere + m 4M Jt e 4+4-* .am- 8 .= .h - -4 d- J A -,. ~ ,.wam.-a A a# -A.-E
< _ , . ..i g , A ', i l; *; ii
... B-4 jR :
i l
OUTLINE l
Limit cycles l
. Why is there a limit cycle ?
- Is there an average power increase during large limit cycle oscillations ?
- Bifurcations ???
L l
. ATWS ,
- What happens when the operator reduces water level at natural circulation ?
4 2
. ( ) M-L: ACRS 23-MAY 89]
1,.<
AVERAGE POWER INCREASE DURING BWR LIMIT CYCLES We will show that during limit cycle oscillations :
. Time varying portion of feedback reactivity is essentially an undisturbed sine wave (i.e., little higher harmonic contamination)
. If the reactivity oscillates as a sine wave, and the power oscillations remain bounded by a limit cycle, then the reactor must be subcritical (i.e., there is an average or DC component in the feedback reactivity)
==>Thus, there is a power increase during limit cycle oscillations.
- Reactor-independent correlations relate the average reactivity increase with the amplitude of oscillation.
( '
( ) M-L ACRS 23-MAY 89)
+s 4 TYPICAL BWR TRANSFER FUNCTIONS 4
REACTIVITY " POWER y
.g . xi
- .. .., .i i s. i.
I as de l
m
't., .., ., , i. i.
/ L
( J M U ACRS 23.MAY 89)
9 4
l~
r 7 BWR OPEN LOOP TRANSFER FENCTIONS 50 0 g
l t
i
?
O --
90 P G ;
H A A I -
S N i E D j D B -50 - - - - - -
-180 E G
,g ........,..i.....,......,.............'270 -
0.001 0.01 0.1 1 10 100 FREQUENCY (Hz)
[ ( ) M-L: ACR$ 23 MAY 89)
I
r ,
REACTIVITY FEEDBACK DURING LIMIT CYCLES IS SINUSOIDAL Limit cycles have been observed (numerically and experimentally)
Observed limit cycles are stationary signals. Amplitude is bound and signal is periodic (before bifurcation regime)
, => Power can be expanded in Fourier series n(t) = I Au sin (kwt)
Thermohydraulics are an excellent filter.
(1/10 at 0.5 Hz, 1/1000 at 1 Hz)
=> p(t) = I H Au i sin (kwt)
~
F,A+HAi o 3 sin (wt) + e
< 2
( NU ACr4 2>WY 8Q
c.
TYPICAL LIMIT CYCLE 600i 6 i
i l l
1 400: -4 I R e-a P -2 C 200 -
t W I -
( e r
0
\ A A A i 0 t
% y
-200 -
-2 r
-400 - -4 0 2 4 6 8 10 Time (s) .
c
- ouucrn 2 war.eo
4'
.o .
l .
SINUSOIDAL REACTIVITY With zero average, Power increases 160, t
l Reactivity (S) 140 -
l j
120 -
l e
1 ,
a ..
l 300 _
i .
- k. V- 80 -
Time (s)
P 60 -
o I w
e @ -
r i bbb 0 10 20 30 40 50 Time (s)
() M-L ACRS 23-MAY 89) a
.- . .. . . . ...s . . . . . . ., . . . . . . . .
.s: >
SINUSOIDAL REACTIVITY With too much average, ,
Power d.ecays ;
2
, Reactiv!ty (S) l .
es .
R e
1.5 -
g g g ,
1 ,[A a '
l '
1 t
i ,, ,
l V' } b l 3J. : 4 e e i.
P '!i h e(5)
I o i w
e o.5 -
i>\>I r V g bbU00-bhhhhg(
0 O 10 20 30 40 50 Time (s) 1
( J M-L: ACRS 23 MAY 89] (
l l
l
_ _ - . _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ - _ - _ - - - - _ _ _ _ _ _ _ _ _ _ _ _ - _ . _ - _ _ _ _ _ _ _ _ . .~
[
F 7 t
SINUSOIDAL REACTIVITY With the right average, l Limit cycle is established
, Reseunts ($) g
- ~
R e
1 <
a ,, ,
t 3 -
I "
v Tvi m .
e 2 ! i l l l !
p i f !
O W
e l r 1
, Oh0hhhh00h hhOhhh00 I
)
0 O 10 20 30 40 50 Time (s) !
L "
( ()M L ACR$ 23 MAY 89) l l
4 1
_ .._____ , . . . ~ . _
l
- THE MODEL
- i 6 , P d n gc, . L dt A ^
POINT KINETICS d.f. s ('t. n AC '
de ^
Ndr s K ( m fW) - %T FUEL '
44t U , ni p s. T ENERGY AND CONTINUITY d&
dt l
l MODEL PARAMETERS ARE OBTAINED BY FITTING THE RESULT OF A DETAILED NEUTRONIC-THERMAL-HYDRAULIC CALCULATION i i gg i g i l illiiq ilillij i l i stiij i I llilig i I s isis l s ilg
^
18 8 r 1 5 5 3
~
. ,4 .. .
10-1 f 1-I -
13-2 i i iiiiig ,_ i uting , , j i,iiil i i iiiiid__2,3 i,iint i i iii,u
~
10-3 yg-2 yg-1 yg 8 3g1 gg2 yg3 FREQUENCY (H7.)
J M.L ACRS 23$MAY 89L I ' #
A :t - /
i AVG. REACTIVITY DECREASE DUE TO OSCILLATIDNS
?
0 .
l
-0.5 'r !
R ,
e j
a li '
C i
t 1
1.5' -
y i
t y -2 - ,
S 2.5 I
3 500 1000 1500 2000 2500 3000 3500 0 i Peak Oscillation (%)
l I < (;u.e Acres 23.uAv.e9) 1 l
L--______-_ , - . - . - - - - - - - - . -
4 AVERAGE POWER INCREASE DUE TO OSCILLATIONS 20
/ .
i i
i !
i 15 '- :
i !
1 P ;
O i W
e 10 - ,
r Sr i
!i
' i i + 1 0 500 1000 1500 2000 2500 3000 3500 Peak Oscillation (%)
I "
( m.tc Acas 2>MAY tQ 4
t
e r !
POWER INCREASE
SUMMARY
l Average reactivity decrease stabilizes (i.e., I bounds) limit cycle oscillations
- Govemed by general correlations -
Approximately .3 cents /%_ steady _ state l
. Average power increase must exist i
. Depends on particular power feedback :
i coefficient
. Typically, power increase is 1.5% of peak oscillation amplitude '
a
.l L
( J M L: ACR$ 23 MAY 89)
l l
l l 1
I -
Et+ECT OF ATWS WATER LEVEL REDUCTION ON STABILITY DECAY RATIO FOR AXIAL SIIAPE ,
POWER FLOW BOTTOM-PEAKED SINE UNIFORM Ol'flMAL' (MW) (Mib/h) __
1.35 0.99 1.70 0.63 2000 36 t 1.12 1.03 1.47 0.93 1500 27 0.86 1.07 1.13 0.97 1000 18 fs 1.02 0.24 I.53 1
)c 500 9 0.37 --
)
i O
S m
Y
' Density scactivity coefficie:W redesced by a factor of M so Iwing decay rat' ,nso reasawial>Ie values
=
k>
L J l
. .- . . .- . . . . -. - ._ - - - - - _ ,