ML20133H496
| ML20133H496 | |
| Person / Time | |
|---|---|
| Site: | Byron  |
| Issue date: | 07/12/1985 |
| From: | Merz J, Sharp D WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | Anastasia L COMMONWEALTH EDISON CO. |
| References | |
| NUDOCS 8508090367 | |
| Download: ML20133H496 (12) | |
Text
PD P-Westinghouse Water Reactor Bcx 355 Electric Corporation Divisions Pittsburgh Pennsy!vania 15230 0355 Mr. Lou Anastasia July 12,1985 Ccamonwealth Edison Company NS-RAT-PRRA-85-152 35th Floor FN West Chicago, Illinois 60690 Re: Follow-up to NRC/BNDCECod Heetiug on Byron LCORP g g gg DeIr Lou:
__ Attached is a stnmary of our responses to the questions raised by Brookhaven at tha review meeting. Action items have been identified which will be addressed in the next several days. Per our conversation on July 10, the asterisked items are best addressed by CECO due to the availability of infor1 nation. Could you please send us a. copy of everything that is sent to BNL to insure that our r;f6rences are identical to those being used by the review team?
Feel free to call me if you need any further information. My new phone ntaber la (412) 236-6470.
Very truly yours, 1
[
A on F. Her, Senior Engineer Approved:
Plant Risk Analysis D. S. Sharp, Manager Product Risk Analysis JFM/bbp cc: Mr. Nam Cho - Brookhaven National Labs l
Mr. Al Spano, Nuclear Regulatory Cennission Phillips Building
{
l l
4 i
8508090367 850712 PDR ADOCK 05000454
/ Td.3IGRA*
CRIGINAL s.
P PDR
/)
-~5, Certifica B:-
-[/.
NS-RAT-PRRA-85-149 July 11,1985 Response to Questions on Byron LOOR raised at the June 19 meeting General Topics o
Methodology - it is felt that the support state approach yields conservative results. It will be instructive to compare the approach to the SETS methodology.
o Mean ACT Usage - while components may be removed from service for maintenance for times up to (and possibly exceeding) the A0T, the frequency of such events should be much lower than the frequency of maintenance events of shorter duration. We feel that this inverse relationship is properly modeled by use of mean outage times.
o Ele;Leic Power - DC power has not been analyzed in this study, since no change in the LCO applicable to the batteries and inverters is being sought. It was considered to be beyond the scope of the study to verify proper design separation of Class 1 E equipment AC power has, for the most part, been modeled in the Support State model. 4160 V bus availability is explicitly modeled. 480 V bus availability is implicity included in the support state model. The failure rate of the 4160 - 430 V transformer and its breaker is small compared to the failure rates of both the offsite power source and the diesel generator associated with each bus. Thus, 480 V bus availability was assumed to be parallel with its associated 4160 V bus, thus a part of the support systems. Each component powered by 480 V AC includes the failure of its associated Motor Control Center, each of which is fed by that train's 480 V AC ESF bus. There are 5 MCC on each ESF bus.
- Modeling 1.
See the discussion under Electric Power, above. The six support states modeled do not include states where one bus is powered and service water is available. the loss of one bus is clearly dominated by a loss of offsite power (to both buses) and the failure of one diesel generator. In this mode, if service water is lost, the diesel generator will fail, leading to support state 6.
See, e.g., Figure 4.1.2-2 et seq.
5 Action: A copy of ESF 480 VAC Distribution Diagrams.
2.
DC Power loss has not been considered as an initiating event. See the discussion under Electric Power, above.
)#. Action: Copy of D.C. ESF Loading Tables.
3.
The loss of ESW may be a more benign event than at Zion, since Byron loads the makeup and HHSI ptraps on ESW, not CCW. A loss of ESW will 4
l l
I 1
3*
cause makeup pump trip, but will not cause an innediate loss of RCP thennal barrier cooling due to the heat capacity of the CCWS. Both i
the time to heat up the CCWS and the cooling of the RCS by Auxiliary Feedwater System operation will decrease the chance of seal damage.
Actions:
A.
List of CCW and ESW loads (& sizes).
j B.
CCWS heat capacity (following loss of ESW).
j C.
Recent Seal LOCA developments (See attachment).
D.
Annual frequency of loss of ESW.
i 4.
Average values applied to system operational lineups should obviate l
the need to analyze each nonnally operating system in each potential j
operating mode.
Action:
None 5.
The A W pumps are tested monthly on a staggered schedule. Thus, the i
.two trains are always separated by 1/2 month in their test period.
The 3/4 AT unavailability was applied to the more unreliable of the two AW trains, and is considered to be conservative. In addition, this time period is only applied to open valves, which do not comprise a significant contribution to AW system unreliability. Refer to Tables 3.5-9 and 10.
1 Action:
None 6.
The support state approach conservatively models the simultaneous maintenance phenomenon between front-line systems and support systems. The unavailability of a given support system train includes that train's maintenance. Thus, in a given support state, a contribution to that support state is the maintenance of a support system component. For example, Support State 3, Bus 141 unavailable, includes the maintenance of Diesel Generator A.
During quantification of front-line system unavailability, however, only Train B is included in, e.g., Low Pressure Injection. A contribution to LPI unavailability is the maintenance of the "B" RHR ptsop. This event, however, is precluded by the tech specs. Thus, the system unreliability is over-estimated by the product of the unavailabilities of the two systems due to maintenance.
Action: None 7.
(1) RHR valves 8811A,B are not tested quarterly, they both are stroke l
tested and automatically actuated at 18 month intervals (refueling).
. Action:
Correct Table 3.7-1 (2) At present there is no flow test of the CCW valv,es at the RHR HX.
There is a capability to test these valves by measuring design flow rates with the system properly configured. Also, during cold shutdown, flow conditions will be verified by proper reactor decay heat removal operation.
1 Action:
None i
I
(3) SI valves 8926, 8806, 8923A,B cannot be isolated during power operation, it is a violation of tech specs to do so. The unit would be shut down.
Action:
None (4) SI valves 8921A,B and CV8481A,B are not tested in the ISI program.' The check valves will be tested at refueling (during vessel injection) and whenever the centrifugal charging pumps are utilized for makeup. These valves will not contribute significantly to the unavailability of the HHSI systems.
See Table 3.7-23 Action:
None (5) SI8835 receives an S signal. Failure to restore after test should be ANDed with failure to open on demand. This error is conservative.
Action:
Potenti. ally correct fault trees (6) High pressure recirculation path valves 8804A,B, 8807A,B and 8924 are stroke tested quarterly.
Action:
None (7) High pressure recirculation fault tree, figure 3.7-4 should have gate 181 input to gate 185.
Action:
Verify and change fault tree (8) Agree that our treatment of switchover operation was non<onservative in the model of high pressure recirculation. The operator failures on each of the several MOV movements should be interrelated.
Action:
Possibly change fault tree (9) It was assumed that the NPSH available to the RHR peps from the containment sep upon opening the sep isolation values would be sufficient to close the RWST suction check valve. Under this assumption, manual closure of the RWST suction MOV would not be necessary to prevent the RHR pumps from draining the RSWT empty and cavitating. If the opposite were true, than the automatic actuation of recirculation operation by opening the sump valves would be defeated.
Action:
V'erify the NPSH (containment) exceeds NPSH-(RWST) upon the actuat, ion of recirculation of low RWST level.
(10) Accmulation discharge valves are electrically locked out in the open position, are alamed for improper position above about 600 psig RCS pressure, and are verified open at regular surveillance intervals (once per 31 days) See T.S. 3.5.1.
5 Action:
Correct the fault tree (12) Operator failure to restore valves 8809, 8716, and 8812, which are alarmed on wrong position, is a conservative treatment. The inclusion of this failure mode does not impact the results.
Action:
None 8.
The mean times to repair components in this study were derived assuming: 3-day LCO - lognonnal distribution with 5th and 95th percentiles of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> and 60 hours6.944444e-4 days <br />0.0167 hours <br />9.920635e-5 weeks <br />2.283e-5 months <br />, respectively; and 7-day LCO -
lognonnal dist. with 5th and 95_th percentiles of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> and 120 hours0.00139 days <br />0.0333 hours <br />1.984127e-4 weeks <br />4.566e-5 months <br />, respectively. The prior distribution was utilized throughout the study, since keeping the duration times consistent was the most comprehensible means for measuring the MITR as a function of LCO. The updated values for different components at Zion were representative of day LCO, and it was felt that it would not be meaningful to try to derive 3-day LCO durations for comparison to the updated values. No data existed (at the time of the study) to substantiate MITR on 3-day LCO plants, nor on the Byron facility.
Action:
Possible sensitivity analysis 9.
Diesel generator failure was modeled in two modes: demand failure to start, for which the data comprising the failure probabilitiy inherently includes a standard test period of about 1 month; and hourly failure to run, which is a failure rate derived from operating experience.
Action:
None
- 10. Connon cause failures in a nonnally operating system were modeled as follows:
a.
One pap running and one standby - connon cause failure to run; b.
Running pump trips and both must start (loss of offsite AC) -
connon cause failure to start and failure to run.
Action:
None
- 11. Connon cause linking of the two dissimilar AFW peps was included since the two peps are identical, although the drivers are different.
For this model, Atwood's beta factor for diesel driven pe ps was assmed to be entirely embodied in the pep, and thus used to represent the failure of both peps.
Action:
None
- 12. Typicalquanyficat195 cutoff probabilities were as follows:
WAMBAM - 10 to 10-WAMCUT - roughly 3 orders of magnitude below the WAMBAM result.
Action:
None
- 13. Agree that coefficients in comon cause equations are wrong, see pages 3.8-3,5.
These coefficients are applied to higher order failures, and should not impact the results.
Action:
Check that proper equations were used, correct report tables
- 14. AW diesel driven pump maintenance frequency was taken from the Torrey Pines 0611 Report.
gAction:
Look into additional data on Zico CS pump and Byron AW pep
- 15. The startup feedwater system was not credited in the LCOR analysis.
Action:
None
- 16. The provision for crosstying the A diesel generators has not been credited in the analysis. Thus, the study assumes that 2A DG is available when 1A is taken out of service. In actuality, if 2A were unavailable, the units would be shutdown within two hours.
Action:
None
- 17. Sequence information is available and will be provided.
Action:
Attach Adam output
- 18. Maintenance procedures are extreely volminous and probably of little value.
g Action:
CECO to identify system engineers at Byron who may help with review questions as they arise Additional Questions 1.
Reasons for the requested changes is A0Ts include:
A.
fewer unnecessary plant shutdowns due to T.S.;
B.
more time to diagnose failures, seek out the root causes; C.
less chance of haan error since rush situations will be avoided; D.
allow procurement of spare / replacement parts; and E.
there are limits on the number of hours any one worker can work on safety-related equipment, which impacts the actual time available for the performance of the repair effort.
2.
There are not substantive differences in the Zion and Byron maintenance policies, since the Byron staff has largely been derived from the Zion staff. The major constraint on the Byron staff is the 3-day LCO, which restricts the time available for preparation and trouble shooting.
3.
Refer to Question 3 under Modeling, above.
4.
Unit 2 Diesel Generator A is available.
5.
There are no room coolers for the AW pe ps. Auxiliary building
ventilation is available for' general area cooling.
6.
Emergency operating procedures for feed and bleed are used.
- 1ZFR-H.1, which follows the Westinghouse ERGS, calls for feed and bleed when AFW flow falls below 485 gpm.
Other Materials Will RHR discharge valves (cont. Isol.) open on S when system in test and valves closed?
Need copy of latest revisions of system P&ID.
g Do ESW valves to HX 0 open on S signal?
Fan cooler operability tech spec modeled incorrectly. Assess the impact of the current LCO.
O O
e 4
0
h"[ 5 5%#?M W Y'.Y ?..N!:k$$~f!?Y~ l' 5
^
,h,f' mv. 2p($.
.,.a
-n.- -
@1RCP) 1 concerned with Westinghouse reactor coolant pump e tenue seal performance in a postulated loss of all seal i dooling scenario..The postulated occurrence of less-of-all-AC power (station blackout) was evaluated generically by Westinghouse and found to be the dominant contributor leading to loss of all seal cooling.
This scenario involves loss of l
' c11 of f site power and f ailure of all-the diesel generators -to l
Ctcrt on demand..The fGtC has designated this postulated event co USI A-44.
This condition is beyond the current licensing j
boato and beyond the Westinghouse design basis for nu* clear power plants.
Westinghouse Owners Group uponsored analysi,s hoc thown that the seals will continue their sealing function and limit the reactor conlant system leakage to minimal levels if a ctation blackout occurs.
When the probabilities of a loss-of-all-AC event are f actored into the low estimated lockcge, the frequency of core uncovery from these events is Iower than the current preliminary NRC maf ety goal.
Therefore, Westinghouse does not consider plant modifications necest:ary to address the issue.
- nFQ~
Th9 cnalysis of RCP seal performance was submitted to the 45tC For comment in April, 1984 an' WCAP-10541)(Westinghouse s
3roprictary), " Westinghouse Owners Group Report, Reactor 7,oolcnt Pump Seal Performance Following A Loss Of All AC Sower."
The NRC contracted Energy Technology Engineering 7, enter (ETEC) to perform a review of the WCAP and to perform independant audit calculations.
The ETEC review, which was
- omplcted in December, 1984, concluded that the Westinghouse nothods were conservative while the audit calculations found
.Gckcg2 rates that evere 7 percent to 20 percent lower than the dedtinghouse results.
NRC comments have been forwarded to the testinghouse Owners Group in a letter dated April 17, 1985.
- ha NRC letter concluded that substantial additional testing,
.ncluding a f ull scale demonstration test, was required due to
.he complexity of a failed seal flow path.
Many of the fWtC
- omments are already being addressed by the most current lestinghouse Owners Group program.
locondcry sealing materials testing at the Chalk River lational Laboratory of Atomic Energy-of Canada, Ltd. (AECL) ndicicted that the currently used 0-ring mater ~ial would
~
robcbly not survive the expected loss of seal cooling
' cG--
j on' itions for the required duration of the loss of all AC ower cvent.
An alternate O-ring material subsequently tested y AECL did demonstrate a much higher resistance to f ailure n-J maintained it's integrity at conditions which are more xtreme than expected dur ing the event.
Based upon successful ormal operation testing, the Westinghouse Owners Group and estinghouse concluded that it is appropriate to change out ha 0-ring material to the alternate 0-ring material during ormally scheduled pump main'tainence outages.
We1Ptinghouse ill b3 able to begin supplying the' alternate D-ring material n April, 1986.
'Tho Westinghouca Dwnsro Group 01co fundsd tonto ct AECL of tho taflon boccd chcnn21 cool matcrici,cubjtet to the cxtramco of proccure and temperature identified in the analysis.
Tooto of
,unirradiated channel seals were successful, showing very little or no extrusion.
However, tests of highly irradiated chcnnel seals indicated that extrusion may result, pcrticularly if the channel seals are exposed to oxygen af ter 1
cxtcnded irradiation.
The NRC has expressed concern that cxtrusion of the channel seal material may result in the RCP' coc10 being forced open due to thermal growth of the RCP shaft end housing daring the loss of all seal cooling event.
-Moro--
Additional testing will be required to demonstrate accept'able cxtrusion limits for better estimate irradiation and oxygen cxposure, or an alternate channel seal material will require qualification.
Tha Westinghouse Owners Group participated, through a Wactinghouse three-party agreement, in a loss of all seal cooling test on a 7-inch RCP seal system in a static RCP m:ckup at the Montereau Power station in France on May 29 1985.
The test was conducted by Electricite de France (EDF) with joint participation by Framatome and Jesumont-Schneider cnd demonstrated acceptable leakage rates during the loss of all seal cooling event.
The Westinghouse Dwners Group pcrticipation made the test more representative of the loss of all AC power event.
Th3 response of the 7-inch seal system to the loss of all seal cooling was unknown prior to the test since there are oignificant dimensional design differences between the 9-inch RCP seal systems which were analyzed and the 7-inch RCP seal cyotem.
A less detaed evaluation of the design indicated
-Marc--
that higher leakage rates were expected in the 7-inch seal cyctem design.
However, the 7-inch RCP seal preliminary test rocults indicate that the leakage rates were approximately 40 percent lower than the analysis calculations for the 8-inch i
RCP seal system.
The 7-inch RCP seal system was well behaved cnd stable throughout the 20 hour2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> duration of the test which ciculated the expected reactor coolant system conditions at i
tha inlet of the RCP seal system during a loss of all AC power ovcnt.
Th9 test results indicate that resolution of the channel semi cxtrusion concern and substitution of the alternate 0-ring actsrial should be nufficient to close out tne twsue.
Hrwever, the NRC may require analysis of the 7-inch RCP seal cyctem as well as a full scale demonstration test of the C-inch RCP seal system to resolve the issue.
The Westinghouse Owncre Group will be meeting with the NRC in July, 1985 to diccuss the remaining NRC concerns and methods for resolving th3 issue.
Westinghouse and the Westinghouse Owners Group will petition the NRC to review the EdF 7-inch RCP seal full cccle test results for applicability to the Westinghouse dacigbefore embarking on a full scale demonstration test of thn B-inch RCP seal.
--Mor o--
t
'~G--
$ditionally, the Atomic Industrial Forum has proposed a skuntcry utility program that would supply the NRC with
- implified deterministic plant-specific analyses in order to
- stchlich the tolerable duration times for station blackout.
,smmento on this program have only recently been solicited.
9 l
i
ty y
OFf0041NANT ACCIDENT SEQU5NCES LIST 11t!!__!!!N{,,,y C[,}{ C[,(({Q;,, (({N{_S{g __ ((p((0 S{S((hS q SLO 2 SLFC 1.6 7 50E-0 5 01111' 211011 R 1 - k6f00 FD 7%%I s 0f.G1 I.OJ TT 17 52/5EZ 6.40015-0o 0022222222 's J 7.0 2 LMcw 44 52/5EC 4.?000E-Oo 102222222222 f)FCC/EC (WF-7 14.02 WT 44 Sc/3EC 4.SGCOE-Os 902222222222 12.03 LO5P 5e SJ 2.7750E-Oo 0100 20 2 2 2 2 d 12.00 LOSP 23 SJC 9.7 300 E-07 0100 210 2 311 7% Lor /~at 11.01 TT 3 SLFC 9.5772E-07 0111010111 >>7.01 LMcW 3 SLFC 7.15290-07 011112010111 gg7pg,yyggyd,/ 14.01 R1 3 SLFC 7.132?E-07 011112010111 33.02 SLO 59 52C 0.8259E-07 011161201011 O f Ac n,+ < 04.02 SGTR 48 VZE c.4535E-07 001110002 01.01 LLO 19 AdFC
- o. 2 789E-07 0111011122 92.01
'4 L O 37 ASFC 6.27E9E-07 01110111122 ffggycf A4;-g' 39.02 LRCS 44 52/SEC 5.750GE-07 002222222222 11.01 IT ? TEFC 4.1709E-07 0111002111 JS.02 MSIV 45 52/ SCC 4.0500E-07 0022222222222 b--][ 12.05 Losp 20 T-2.46d6E-07 G1100022222 14.01 RT 9 T5FC 3.143 2 E-0 7 011112 30 2111 l J7.01 LM.W 9 TIFC 3.14 3 2 E -0 7 01111200 2111 8 J1.01 LLO 2 ALFC 2. 3 413 E -0 7 11 11 111110 l / <d J2.01 MLO 2 ALFC 2.3413E-0? 01111111110 ) f c I 11.02 TT 12 TEL 2.1715E-f? G111002011 f i s 15.01 ISL 2V 2.1JJ)L-C7 00 12.03 LOSP 3 SLFC
- 2. 0 a F,3 E -0 7 0111010111
-d 11.01 TT 17 52/5L.C 1.900dE-0 7 00 22 22 22 22 4-2 l 34.01 SGTR 20 VZL 1.8273E-07 0120J1102 14.02 RT 12 TEC 1.o236E-07 011112002011 ? 'I 97.02 LMFW 12 TfC 1.62PoE-07 011112002011 37.01 LMFW 44 52/5L;C 1.5000E-07 002222222222 P-2 i 14.01 PT 44 52/5LFC 1.5000E-07 002222222222 R-2 ( 13.02 SSI so 52/5EC 1.2:01E-07 00 2 2 2222 22 22 13.C2 55! $ SLC 1.233?E-07 011112100011 12.05 LOSP 23 T5FC 1.2163c-07 01100102111 ( COREMELT FREQ total 4.60 E-05 d4.01 SGTR 4 V2L 7.183 sE-03 012111020 J4.01 SGTR 16 V2L
- 1. 3 34 6 E-0 8 0121000 20 is
\\ J4.01 SGTR 43 V2E 1.33136-03 001110002 24.01 SGTR 8 V2L 1.2461E-Od 012110021 >4.0c 5b7R 60 V2E 7.1424E-03 001100002 J4.Ud Sb7R c1 v2E 2.5160E-03 J01022222 '8 V2E 1.22 59E-03 00111000 2 24.00 SGTR +
- 3.00 SLO s1 55
- 1. 3141 E-0 3 011101201002
- J7.0 0 LMFw 44 52/CE 3.3000E-08 002222222222 14.Os RT
+4 52/5E 3.T 300E-0 3 00 22 22 22 22 22 12.0S LOSP ac 51 1.021 ?E-0 3 n100 20 22 ?2 2 12.00 LOSF 29 li 0.4350E-09 01103022222 11.0o T1 17 s2/SE 4.400dE-03 0022222222 J3.02 SM 77 SEC 3.3953E-02 2.4025E-03 011100201011 J3.01 SM 13 SCC 2.3915h-07 1.02300-03 011021221111 ud.01 C 45 52/3LFC 1.20006-03 0C22222222222 R-2 Ot.01 C 3 SLFC 2.3043E-07 6.0336E-0? 011121 ?C 10111 J8.01 C 9 TE5C 1.1477E-07 2.e*03E-06 0111212072111
6.02 CL 12 ThC 1.39o55-12 1.3674 -09 0111212002011 l 1.70005-09 902222222222 R-2 89.01 Lc 64 52/3LFC 9.Cl LO 3 SLFC 2.3943E-07 P.$716E-03 011112010111 1.04'70-07 3.7503E-08 311112002111 v.01 to 9 TEFC 9.C2 Lc 12 TEC
- 1. 60 $5 5-02 1.95105-0 5 011112 302011 2.01 Le 3 SLFC 2.3043d-07 1.3504E-03 3111010111 i.Ld LO 17 52/56C 3.30305-01 3.7500E-03 00 2222 2222 2.C 3 LG 9 TEFC i.1S?45-05 2.9 35 3E-03 011100 2111
>2.03 Lu 26 TEC 4.41135-03 3.3025E-0$ 01100102011 L3.01 SP 10 SLFC 2.3V375-07 1.9150E-03 011112010111
- 2. 24 51 E-0 7 1.7 9 61 E-O d 011112302111 3.01
>P 1e TEFC 3.01 .5P 3 SLFC d.15e23-97 1.7234E-OS 011112100111 c.01 ' 4T W 44 S E. C .74323-04 1. 3 02 3E-O s 010 2 22 20 22 2 2111 2 sw AbG 94 fbci si~ 2_lb> Cr ~ C 2tG + sw O e W}}