ML20079K188
| ML20079K188 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 12/15/1982 |
| From: | Thadani A Office of Nuclear Reactor Regulation |
| To: | Parr O Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20079F081 | List: |
| References | |
| FOIA-83-388 NUDOCS 8212290480 | |
| Download: ML20079K188 (8) | |
Text
.
December 15, 1982 MEMORANDUM FOR:
- 0. Parr, Chief Auxiliary Systems Branch Division of Systems Integration FROM:
A. Thadani, Chief Reliability and Risk Assessment Branch Division of Safety Technology
SUBJECT:
EVALUATION OF THE RELIABILITY STUDY FOR SEABROOK STATION AUXILIARY FEEDWATER SYSTEM Enclosed i's our evaluation of the reliability s'tudy for the Seabrook Auxiliary Feedwater System.
This evaluation is based on the review of the applicant's reliability analysis of the Seabrook Auxiliary Feedwater System performed by Brookhaven National Laboratory.
v If you have any questions, please contact' John Tsao (x27422) of RRAB.
A. Thadani. Chief Reliability and Risk Assessment Branch Division of Safety Technology
Enclosure:
'7 As Stated
}
cc:
S. Hanauer Distribution M. Ernst R. Lobel Central file goN '9pb R. Anand RMB RM
~
J Tsao
'y Tsao.CHRON DOgo R Frahm W
I A Thadani W(,09 v[
Q'
/
})N
/
g l
cmeo i..B.8 A.B.; DST l RRAB: DST
..sAB : D,,
,,,,,,,,,,,,,,,,l,,,,,,,,,,,,,,,,,,,,
i
! mamuok.i.u.q. %...e.1.... Era.he.W::f..A,i,ha,da ni,,,,,
am Q:.12L.Ls..LB2...
.12L..).3..La2....12L.Vp.82...
- rcr;u s:s ite.s:3 sacu c24o OFFICIAL RECORD COPY umam-um
Evaluation of The Reliability Study Seabrook Station Auxiliary Feedwater System 1.0 Introduction 1
As a result of the Three Mile Island accident, the NRC issued a letter dated March 10,1981, requiring the applicants to perform a reliability study on the auxiliary feedwater system (AFWS). The applicant, Publi.c Servige Company of New' Hampshire, responded by submitting a reliability study for the Seabrook Station on June 22, 1982.
Our contractor, Brookhaven National Laboratory {BNL), reviewed the licensee analysis;-and performed an independent analysis.
The review addressed the following three transient conditions, as required in NUREG-0611', for determing the reliability of the AFWS:
1.
LMFW:
Loss of Main Feedwater System.
2.
LOOP:
Loss of Off-site Power / Loss of Main Feedwater System.
3.
LOAC:
Loss of All AC Power / Loss of Main Feedwater System.
Following the guidelines presented in NUREG-0611, the review evaluates host AFWS component failures but excludes the effects of high energy pipe breaks ecand the external events, such as earthquakes, fires, and floods.
The system boundary includos the water sources, condensate storage tanks, and the AFWS nozzles on the steam generators, excluding the AFWS ring header inside of the steam generators.
Systems that support the AFWS such as the service water system are examined qualitatively.
2.0 Results The AFWS of Seabrook Station, a Westinghouse design, consists of two systems.
The first system consists of a motor-driven, non-safety startup feed pump (SUFP) and its associated piping and valves that are used for normal plant operation such as startup, hot standb, and hot shutdown.
The second system is the safety class emergency feedwater system (EFWS) which consists of a motor driven pump (MDP) and a turbine driven pump (TDP).
The SUFP is powered from off-site power; therefore, in a LOOP transient the operator should have sufficient time to perform this manual connection since the AFWS will not be needed for about 30 minutes.
Table 1 presents the BNL and applicant assessment of the Seabrook AFWS unavailability.
l l
v~
TABLE 1 AFWS Unavailability for Three Transient Cases Applicant Assessment BNL Estimates, Per Demand Per Demand Modified Design Original Desian Original Desian
-5
-5
-5 1.
LMFW 1.95x10 4.5x10 2.1x10
-5
-4
-5 2.
LOOP 8.5x10 1.8x10 5.2x10
-2
-2
-2 3.
LOAC 2.3x10 2.3x10 2.1x10 The difference in AFWS hardware between the modified design and original design is; in the original design the SUFP feeds the steam generators through the main feedwater piping, whereas in the modified design the SUFP discharge is revised so that it is also cross-tied to the emergency feedwater piping, creating a redundant flow path.
The difference between the BNL and applicant's AFWS unavailability for the original design case is because the applicant used component failure rates other than those presented in NUREG-0611.
cc --
2.1 AFWS DOMINANT CONTRIBUTORS
~
Table 2 compares the BNL and applicant estimated probabilities for the dominant contributors to AFWS unavailability in the LMFW, LOOP, and LOAC transients.
In the LMFW case the failures of the MDP/TDP and valve V125_
due to either maintenance or hardware malfunction are the dominant. con,
tributors. The failure of all three AFWS pumps is also dominant.
In the LOOP case, in addition to the dominant failure modes in LMFW case, the diesel generator failures also dominate the AFWS unavailability.
One
? --
dominant contributorg s the diesel generator 1B hardware failure or MDP i
hardware failure coupled with valve V125 failure due to maintenance error.
Another dominant failure mode is.the operator error in connecting the SUFP from a non-safety to safety bus.
In the LOAC case the dominant failure modes are the maintenance or hard-ware failures of the TDP train, and failure of dne of the steam admission valves (V127 or V128).
~
2.2 COMMON CAUSE/ SINGLE POINT FAILURES The BNL' analysis included common cause failures in the area of common location, environment, human interactions, and support systems.
No significant common cause failure or single point failure existed in the AFWS in the present design.
s.
W T
3-
. i
3.0 CONCLUSION
S The AFWS unavailability for the LMFW and LOOP cases meet the criterion of
.the Standard Review Plan, Section 10.4.9; and the AFWS unavailability for the LOAC case is acceptable when compared to the plants analyzed in the NUREG-0611.
e e
IC-
?
s e
e 3.hg* h 6
e 1
e e
y
.g
4
.I 9
REFERENCES 1.
" Actions required from Operating License Applicants of Nuclear Steam Supply Systems designated by Westinghouse and Combuction Engineering resulting from the NRC Bulletins and other task force review regarding the Three Mile Island Unit 2 Accident," from D. Ross to all pending Operating License Applicants, dated March 10, 1980.
~
2.
Brogan, B. A., et al., " Reliability Analysis of the Emergency Feedwater System at the Seabrook Nuclear Power S.tation," WLA-1-R-82-02, Wood-Leaver and Associates,'Inc., June 1982.
3.
Fresco, A., et al.,' " Review of the Seabrook Units 1 & _2 Emergency Feedwater' System Reliability Analysis," Brookhave National Laboratory, October'1982.
4.
" Generic Evaluation of Feedwater Transient and Small Break Loss-of-Coslant Accidents in Westinghouse Designed Operating Pla'5f.s,~" U.'S. Nuclear Regu:atory Commission, NUREG-0611, January 1980.
e e
'9 9
4 4
f
.o 4
O 4
Y Ag h
f,-
a-.
?
W y
.BW w
-g
(
6 TABLE 2 - DOMINANT FAILURE C0flTRIBU10RS TO SEABROOK AFWS UNAVAILABILITY
. Loss of Main Feedwater (LMFW), per demand f
3 DNL*
Applicant Contributor
'5.8 x 10'0 7.0 x 10-0 liardwareFailuresof-MDPandValheV125failur'cduetomaintenanceerror.
6.3 x 10 3.0 x 10-6 Maintenance outane of MOP. strain and failure of valhe V125.
~0
-6 5.0 x'10-*
2.0 x:'10 Failure of MDP and Valve V125 due to human error in maintenance (components left disabled).
s 4.5 i 10-7 1.9 x 10 liardware failure of both MDP and TDP and operator error in aligning SUFP.
-6 5.8 x 10-6 9.0 x 10 Equipmentfailure(e.g.halvefailures)disablingMDPtrainandfailure of Valve V125.
6.6 x 10-6 7.3 x 10-7
~
liardware failure of all three AFWS pump trains.
5.2xlb-7 5.9 x 10-7 f
liardware of..one EFW train whild other EFW train out of service coupled with SUFP unavailability.'
I
~
- Estimates extracted from the BNL's analysis of the orginal, des.ign.
s
.t t
a e
TABLE 2 - C0tiTIrluED l
'a
~.
Loss of Offiste Power (LOOP) per demand BNL
~
. Applicant Contributors Dieselgenerator1BorMDPfailurecoupledwithfa',lureofhalveV125
-5 3.0 x 10-5 2.1 x 10 5
due to maintenance error.
.!i,2 x 10-6 5.8 x 10-6 liardware failure of MDP nd TDP and operator error.jn aliani.nq SUFP.
~
/ 3,x10-5 5.0x'10I6[
Maint$enance culage of MDi( and Valve V125.
,3
'A 4.3 x 10 7.0 x 10-g ' '_
liardware failure offall three pumps.
~
-5
/
/
j' ~
5.8'x 10 2.0 x 10'6
'liardware failure of f1DP and Valve V125. ~
-6
~
c-1
[. -
3.0 x 10-5 1.8 x 10-6 Failure of TDP due to hunian error in maintenance and hardware failure
~
l of MDP and SUFP.
9.0 x 10-0 3.3x310-7 TDP failure due to maintenance tror failure of diesel generator IB e;-
and operator error in aligning the SUFP.
e :
.c
,l?
g'! ;
,s.. d
./
...r a
'-['/g i
.4 f) >*
-r=,
9,,j; /
p
?
+
.,A
~
a-
/
i s
J I
t
x TABLE 2 - C0tlTIlluED o.
o f:
Loss of All AC Power (LOAC), per demand-BNL Applicant Contributors
-3 5.7 x 10 1.4 x 10 liardware failure of TOP..
~
-3 5.x 10-3 4.1 x l'0 Failure of TDP-due to human error during maintenance act (lef t the
~
TOP disabled).
6.x 10 2.5 x 10-3 Scheduled maintenance outage of TDP.
1.1 x 10-3 [','.7.0x10-Failures of valves in the TDP train such as steam admission' valves.
-5 8.5 x 10 Miscellaneous single valve failures.
.I
- Not available in the 15NL's' analysis.
- t u.
s
?
s
- * * ~