ML20216J263
| ML20216J263 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 09/28/1999 |
| From: | Raghavan L NRC (Affiliation Not Assigned) |
| To: | Ray H SOUTHERN CALIFORNIA EDISON CO. |
| References | |
| NUDOCS 9910040270 | |
| Download: ML20216J263 (21) | |
Text
,
7 yLL$ $\.0 September 28, 1999 Mr. Harold B. Ray Executive Vice President Southern California Edison Company San Onofre Nuclear Generating Station P.O. Box 128 -
San Clemente, CA 92674-0128
SUBJECT:
FINAL ACCIDENT SEQUENCE PRECURSOR ANALYSIS OF CONDITION AT SAN ONOFRE NUCLEAR GENERATING STATION, UNIT 2
Dear Mr. Ray:
Enclosed for your information is a copy of the final Accident Sequence Precursor (ASP) analysis of the operational event at the San Onofre Nuclear Generating Station, Unit 2, reported in Licensee Event Report (LER) No. 361/98-003. This final analysis (Enclosure 1) was prepared by our contractor at the Oak Ridge National Laboratory (ORNL), based on review and evaluation of your comments on the preliminary analysis and comments received from the NRC staff. Enclosure 2 contains our responses to your specific comments. Our review of your comments employed the criteria contained in the material which accompanied the preliminary analysis. The results of the final analysis indicate that this event is a precursor for 1998.
Please contact me at 301-415-1471 if you have any questions regarding the enclosures. We recognize and appreciate the effort expended by you and your staff in reviewing and providing comments on the preliminary analysis.
Sincerely, ORIG. SIGNED BY L. Raghavan, Senior Project Manager, Section 2 Project Directorate IV & Decommissioning ;
Division of Licensing Project Management Office of Nuclear Reactor Regulation Docket No. 50-361 i
Enclosures:
As stated (2) -
- cc w/encis: See next page j i
DISTRIBUTION Docket File SRichards, (clo) SMays, RES PUBLIC OGC PO'Reilly, RES
.l1 PDIV-2 Rdg ACRS LSmith, RIV To receive a copy of mis document, inaicate U in me box Q; '
i l V I OFFICE PDIV-2/PM C PDIV D/LA C PDIV-2/SC ,
NAME LRaghavahkCJamersonk SDembe[ '
DATE' 9 /L*) /99 7 /M /99 [] 9 /N[99 DOCUMENT NAME: G:\PDIV-2\ SONGS \ ASP.wpd OFFICIAL RECORD COPY PD k 1 S PDR
on
, g "%, 3 4 & UNITED STATES d )# E NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20A%-0001 Septenber 28, 1999 Mr. Harold B. Ray Executive Vice President Southern California Edison Company San Onofre Nuclear Generating Station P.O. Box 128 San Clemente, CA 92674-0128
SUBJECT:
FINAL ACCIDENT SEQUENCE PRECURSOR ANALYSIS OF CONDITION AT SAN ONOFRE NUCLEAR GENERATING STATION, UNIT 2
Dear Mr. Ray:
Enclosed for your information is a copy of the final Accident Sequence Precursor (ASP) analysis of the operational event at the San Onofre Nuclear Generating Station, Unit 2, reported in Licensee Event Report (LER) No. 361/96-003. This final analysis (Enclosure 1) was prepared by our contractor at the Oak Ridge National Laboratory (ORNL), based on review and evaluation of your comments on the preliminary analysis and comments received from the NRC staff. Enclosure 2 contains our responses to your specific comments. Our review of your comments employed the criteria contained in the material which accompanied the preliminary analysis. The results of the final analysis indicate that this event is a precursor for 1998.
Please contact me at 301-4151471 if you have any questions regarding the enclosures. We recognize and appreciate the effort expended by you and your staff in reviewing and providing comments on the preliminary analysis.
Sincerely,
?
i u- -,
L. Raghavan, enior Project Manager, Section 2 Project Directorate IV & Decommissioning Division of Licensing Project Management Office of Nuclear Reactor Regulation Docket No. 50-361
Enclosures:
As stated (2) cc w/encis: See next page i
Eg ,
I . .
San Onofre Nuclear Generating Station, Units 2 and 3 l !
cc:
Mr. R. W. Krieger, Vice President Mayor i Southern Califomia Edison Company ' City of San Clemente
. San Onofre Nuclear Generating Station 100 Avenida Presidio P. O. Box 128 - San Clemente, CA 92672 San Clemente, CA 92674-0128 i
/
l .
Mr. Dwight E. Nunn, Vice President i . Chairman, Board of Supervisors ;
Southern California Edison Company i i County of San Diego San Onofre Nuclear Generating Station l 1600 Pacific Highway, Room 335 P.O. Box 128 San Diego, CA 92101 San Clemente, CA 92674-0128
! Alan R. Watts, Esq.
Woodruff, Spradlin & Smart 701 S. Parker St. No. 7000 Orange, CA 92668-4720 Mr. Sherwin Harris Resource Project Manager ,
Public Utilities Department City of Riverside 3900 Main Street l Riverside, CA 92522 Regional Administrator, Region IV l U.S. Nuclear Regulatory Commission (
Harris Tower & Pavilion 611 Ryan Plaza Drive, Suite 400 Arlington, TX 76011-8064 i
Mr. Michael Olson San Onofre Liaison San Diego Gas & Electric Company P.O. Box 1831 San Diego, CA 92112-4150 .
Mr. Steve Hsu Radiologic Health Branch State Department of Health Services Post Office Box 942732 i Sacramento, CA 94234 I Resident inspector / San Onofre NPS c/o U.S. Nuclear Regulatory Commission Post Office Box 4329 San Clemente, CA 9267.4 June 1999 i
V ., .
l
! LER No. 361/98-003 l
LER No. 361/98-003 Event
Description:
Inoperable sump recirculation valve l
Date of Event: February 5,1998 Plant: San Onofre, Unit 2 Event Summary San Onofre, Unit 2, was in a mid-cycle outage when personnel discovered that the linestarter for the '
containment emergency sump outlet valve wasjammed because ofgrit in the sliding cam. The grit would have prevented the valve from opening on a recirculation actuation signal (RAS). This would result in one ,
inoperable train while in the recirculation mode of the Emergency Core Cooling System (ECCS) and the Containment Spray (CS) system. This condition existed for ~18 d until the unit shut down for a mid-cycle outage. The core damage probability (CDP) at San Onofre 2 increased during these 18 d because of the increased susceptibility that would result from any loss-of-coolant accident (LOCA) that progressed to the recirculation phase. The estimated increase in the CDP (i.e., the importance) for this event is 7.2 x 104 Event Description On February 5,1998, utility electricians were replacing Square D linestarters as part ofplanned maintenance.
The electricians discovered the mechanical interlock on the linestarter for the Train A containment emergency sump outlet valve (HV-9305) jammed. The sump outlet valve was in the closed position at the time the failure was discovered, fulfilling the containment isolation function of the valve (Fig.1). However, the as-found j condition ofthe linestarter would have prevented valve HV-9305 from opening. Consequently, the recirculation '
function for Train A of High Pressure Safety Injection (HPSI) and CS could not be fulfilled without some recovery action. He Train A containment emergency sump outlet valve was last cycled open and closed on January 6,1998. San Onofre, Unit 2, was shut down for the mid-cycle outage on January 24, 1998.
Herefore, from the nature of the failure, the licensee considered the Train A containment emergency sump outlet valve inoperable for approximately 18 d before it was no longer required by Technical Specifications.
Consequently, the ECCS Train A and CS Train A were inoperable for ~18 d.'
Additional Event-Related Information The licensee had been programmatically replacing all of the Square D linestarters - 60 of 86 !inestarters in !
Unit 2, and 61 of 86 linestarters in Unit 3 had already been replaced. All remaining old linestarters (26 at Unit 2 and 25 at Unit 3) were replaced; no additional failures were discovered.'
The grit that caused the linestarter for the Train A containment emergency sump outlet valve to jam was )
identified as Portland cement particles.2 No grit was discovered on or around other switchgear room components or in the ventilation ducts. However, some grit was found in other 480-V ac motor control center I ,
1 ENCLOSURE 1,
D.-
I
{
LER No. 361/98-003 buckets, but it had not affected the operation of the associated linestarters. The grit was assumed to have been introduced before plant startup and was known not to migrate after being deposited.'
The HPSI system has three centrifugal pumps divided among two trains (Fig.1). Pump P-017 is in Train A and pump P-019 is in Train B. The third pump, P-018, is a swing pump and can be aligned to either train on the suction or discharge side. P-018 is nonnally aligned to Train A. Because the HPSI pumps do not automatically stop in respense to an RAS signal, operators are directed to stop the pumps before the water level in the refueling water storage tank (RWST) decreases below 5%.5 While the recirculation phase of ECCS Train A was compromised between January 6,1998, and January 24, 1998, the opposite train - ECCS Train B - was inoperable six times during this same period. These six occasions were for
- 1. I h,43 min to perform an in-senice test of an HPSI pump (January 12,1998),
- 2. 27 h,5 min to repair a Component Cooling Water (CCW) heat exchanger tube leak (January 13, 1998)
(CCW is required to support ECCS.),
- 3. 6 h,36 min to perform heat treatment of the main condenser (January 16,1998). (This treatment process {
increases the heat load on the salt water cooling (SWC) system, which is required to support ECCS.), (
- 4. 19 min to swap the in-senice SWC pump to the opposite train (January 22,1998),
- 5. 5 h,45 min to perform maintenance work on the Train B RWST outlet valve's breaker-position indicating light replacement (January 23,1998), and
- 6. 5 h,31 min to perform an additional heat treatment of the main condenser (January 24,1998).
Modeling Assumptions I
This event was modeled as an 18-d (432-h) condition assmment with the Train A containment emergency I sump outlet valve failed (valve HV-9305). The CCW he2 t exchanger maintenance (27 h,5 min) was included in the modeling because of the time required to back out cithe maintenance. The maintenance period with the unavailable RWST outlet valve (5 h,45 min) was not included in the event model because the valve was l
deenergized in the open position, malmg Train B available da ring the injection phase of an accident. Time was considered available to manually close the RWST outlet vah c before recirculation or to discontinue repairs and make the valve available remotely. Likewise, the two periods invohing heat treatment of the main l condenser (6 h,36 min and 5 h,31 min) were not included in the model of this event because any heat treatment would likely be terminated quickly by the operator. Even if this were not done, a turbine trip initiated by a LOCA would self-limit any added heat loads on the SWC system. The in-senice test of the Train B HPSI pump (I h, 43 min) was not modeled because of operator staffing for the test, the ability to restore the normal lineup quickly, and the limited time the pump was unavailable The time required to swap pumps (19 min) was ;
i not modeled because of the limited time required to perform the task. Therefore, two distinct cases, totaling j 432 h (18 d), were modeled as part of this event.
i Case 1. 404 h, 55 min with only the Train A containment emergency sump outlet valve failed (valve HV-9305).
2 I
pw .
..,. q.-
L LER No. 361/98-003 Case 2. 27 h,5 min with the Train A containment emergency sump outlet valve failed (valve HV-9305) and CCW Train B unavailable.
i
' The CS pumps are not represented in the Integrated Reliability and Risk Analysis System (IRRAS) model for San Onofre. However, because Train B of the CS system and all of the containment emergency fan coolers were available throughout the 18-d event, no attempt was made to inco@ orate the unavailability of one train of CS into the IRRAS model for San Onofre. This is estimated to have an insignificant impact on the calculated importance of this event because CS impacts containment pressure and not core cooling.
'Ihe failed Train A containment emergency sump outlet valve was modeled by s:.tting basic event HPR-SMP-FC-UMPA (Containment Sump A Failure) failure probability from 6.1 x 10-2 to TRUE (i.e.,
probability = 1.0 that the valve would fail on demand). Because of multiple locations which were discovered with grit, previous operational success of the linestarters does not preclude the grit failure mechanism from simultaneously affecting more than one linestarter. In fact, multiple safety components are affected by this failure mechanism; however, most of the affected equipment is not identified. Therefore, only the sump isolation motor-operated valves were considered for adjustment ofthe common-cause treatment. No supporting evidence was available which would suggest that the failure ofthe Train A containment emergency sump outlet valve linestarter was unique tojust this one linestarter. Therefore, the associated common-cause failure basic 8
event (HPR-MOV-CF-SUMP) was adjusted from 1.1 x 10 to the p factor of the Multiple Greek Letter method used in the IRRAS models (8.8 x 10 2) based on the failure of the Train A containment emergency sump outlet valve.
It was assumed that the operators would correctly follow procedures and secure the HPSI pumps before the RWST level decreases below 5%. Therefore, this was not modeled in the analysis.
An evaluation of this event,d prepared by the licensee, estimated that if a small-break LOCA (SLOCA)
(%-2 in. pipe diameter) occurred,250 min would be available to recover a recirculation flow path before the onset ofcore damage. Operators would initiate recirculation flow about 118 min after an SLOCA occurred.
Although other CE plants consider depressurization an option, simulator exercises at San Onofre 2 shewed that operating crews would not attempt to cool down and depressurize the plant for a leak size in this range.
Conversely, it was not expected that small-small-break LOCAs (SSLOCAs) (<% in. pipe diameter) would :
proceed to the recirculation phase because sufficient time was assumed to be available to cool down and -
depressurize the primary system. This differentiation required the IRRAS model for San Onofre 2 to be adjusted to reflect the different operator responses expected following an SSLOCA and an SLOCA. Because I the importance of medium-break and large-break LOCAs calculated by the licensee using a methodology which parallels the IRRAS development was less than 1.0 x 104, these larger LOCAs were not specifically modeled (i.e., the contribution to the overall importance of the event from these events is small).
Recovery from the CCW heat exchanger maintenance could begin at the time a LOCA event was ecognized because the operating staffwas aware ofthe maintenance being performed from pre-shift briefings. By similar reasomng, planning for recovery from the RWST Train B outlet valve maintenance could also begin at the time a LOCA event was recognized. The RWST Train B outlet valve was open for the injection mode and would 3
1 j
m - t,
. o l l
LER No. 361/98-033 1
not be required to change position for 118 min. It was assumed that this was ample time to plan and exi a desired course of action for this RWST valve.
The recovery from the train B CCW heat exchanger maintenance to repair a tube leak was expec 200 min.' This assumes 15 min for operators to recognize that an SLOCA occurred and to order the restoration of the CCW heat exchanger,120 min for maintenance personnel to reassemble the CCW heat exchanger, 60 min for operators to realign the system valves correctly, and 5 min to restore power and stal the appropriate CCW pump. These time estimates made by the licensee are conservative, yet still leaveI additional 50 min before core damage would occur following an SLOCA. Performance shaping factors considered that the process would be govemed bya maintenance procedure and performed under stress outside the control room by a skilled crew.d Based on this, the licensee estimated a 60% probability of success in !
restoring the CCW heat exchanger within 250 min. In addition, one HPSI pump and one residual heat removal (RHR) pump were affected by the maintenance on the CCW heat exchanger. Because the CCW system is no directly modeled by the San Onofre IRRAS model, a basic event was added to several fault trees to repres the CCW system failure probability during the ~27 h maintenance period. The new basic event (CCW-TRNB-FAIL) was added such that a failure to return the train B CCW heat exchanger to sersice would cause the affected pumpF (HPI-MDP-FC-P019 and RHR-MDP-FC-P016) to be failed during the ~27 h CCW maintenance period. The probability of basic event CCW-TRNB-FAIL was adjusted to 0.4 for Case 2; for Case 1, the probability of this basic event occurring was zero.
- Two viable options exist to recover from the Train A containment emergency sump outlet valve failing closed.5 First, the failure of the valve could be traced to the breaker linestarter and replacement could be initiated.
Secondly, it is possible to cross-connect the HPS1 Train A suction to the Train B suction. In either case,132 min (250 - 118 min) would be available before the onset of core damage following an SLOCA. Because operator training and emergency operating procedures focus attention on the correct entry into the recirculation mode, it is assumed that the operators would quickly notice the failure of the train A sump valve to open.
Recognition and correction of the breaker failure were assumed to require 40 min.' This would allow an additional 92 min (132 - 40 min) to complete repairs before the onset of core damage. Performance shaping factors considered that the breaker repair process would not be governed by a maintenance procedure and would be performed under stress outside the control room by a skilled crew.' Based on this, the licensee estimated a 50% probability of success in restoring the linestarter and openmg the train A sump valve within 132 min of RAS. A new basic event (HPR SMPA-XHE-NRE) was added to the High Pressure Recirculation (HPR) fault tree to represent the probability (0.5) that electricians would fail to repair the breaker linestarter. !
Recognition ofthe failure and cross-connecting the HPSI pump suctions were assumed to require 20 min. This action would allow an additional 112 midl32 - 20 min) to complete realignment before the onset of core l damage. Performance shaping factors considered that the breaker repair process would be govemed by an !
operating procedure and performed under stress outside the control room by a skilled crew.' Based on this, the licensee estimated an 80% probability ofsuccess in cross-connecting the HPSI pump suction if there were an SLOCA. A new basic event (HPR-XCONN-XHE-NR) was added to the HPR fault tree to represent the probability (0.2) that operators fail to cross-connect the HPSI pump suctions within 132 min ofRAS. Because these two new events involve separate groups ofplant personnel (electricians and operators), the basic events l
i 4
)
LER No. 361/98-003 were considered to be independent. Independence was also assumed when these two new basic events were compared with the effort to restore the CCW heat exchanger, which would involve mechanics.
Analysis Results Determining the overall increase in the CDP required determining the increase in the CDP for the two different cases, and then summing the results. The cases are i
Case 1. 404 h, 55 min with only the Train A containment emergency sump outlet valve failed (valve IW-9305).
Case 2. 27 h,5 min with the Train A containment emergency sump outlet valve failed (valve HV 9305) and CCW Train B unavailable.
The combined increase in the CDP from this 432-h event (i.e., the importance)is 7.2 x 104 This increase is 4
above a base-case probability for the 432-h period (the CDP) of 3.9 x 10 and credits the possible recovery actions discussed in Ref. 2. The resulting conditional core damage probability (CCDP) for the 432-h period in which the linestarter was failed is 4.6 x 104 Most of the increase above the CDP (90%) is driven by Case
- 1. As expected, the common-cause failure of the containment sump valve shows up most often in the cut sets of the most significant sequences because it is driven by the initial sump alve failure. Potential recovery actions and the CCW train B failure are more conspicuous in Case 2. However, the dominant core damage sequence in both cases of this event (Sequence 2 on Fig. 2) involves
- an SLOCA, e
a successful reactor trip, a successful initiation of emergency feedwater, a successful initiation of high pressure injection, and a failure of high pressure recirculation.
The SLOCA sequences account for -80% of the overall increase in the CDP for this event. The nat most dominant sequence among both cases involves an SSLOCA with a failure to cool down the plant before requiring HPR (SSLOCA Sequence 3). This sequence contributes 5% to the overall importance of this event.
Definitions and probabilities for selected basic events are shown in Table 1. The conditional probabilities associated with the highest probability sequences are shown in Table 2. Table 3 lists the sequence logic associated with the sequences listed in Table 2. Table 4 describes the system names associated with the dominant sequences. Minimal cut sets associated with the dominant sequences are shown in Table 5.
Acronyms CCDP conditional core damage probability CCW component cooling water CDP core damage probability l 5
[ '.
h LER No. 361/98-003 l CS containment spray l ECCS emergency core cooling system HPR high-pressure recirculation HPSI high-pressure safety injection IRRAS Integrated Reliability and Risk Analysis System LOCA loss-of-coolant accident l LOOP loss ofoffsite power LPSI low-pressure safety injection MOV motor-operated valve RAS recirculation actuation signal RCS reactor coolant system RHR residual heat removal RWST refueling water storage tari SGTR steam generator tube rupture SLOCA small-break LOCA SSLOCA small-small-break LOCA SRV safety /reliefvalve SWC salt water cooling TRANS tranaent event References
- 1. LER No. 361/98-003, Rev.1," Inoperable Valve Due to Grit in Linestarter Mechanism," March 17,1998.
- 2. Letter from Dwight E. Nunn, Vice President, to U. S. Nuclear Regulatory Commissios " Response to NRC Inspection Repon 98 05 Regarding Linestaners San OnofreNuclear Generating Station, Units 2 and 3," June 22,1998.
- 3. San Onofre, FinalSafety Analysis ReporI (Updated Version).
l 4. Letter from Dwight E. Nunn, Vice President, San Onofre Nuclear Generating Station, to U. S. Nuclear Regulatory Commission, "Linestarter and AFW Supplemental Information," April 7,1998.
l l
I i
6 t
. . j l
J l
LER No. 361/98-003 )
I 1
l I ~
$s $e a 'g IE $ $ El4 Og Eg 4
i gE b 'O E /8 '
%e E 8 8 E
________7, l IXe1 l*i l *]: l +]:l*]:l
- ll *]:l ej: l *]: l llXs l Ze! I
_ _ _ _I l l l Ze
+];l - -
1- -
p .::1 ;
-E n f*
fg BE i I , XE i
,d k'
, ... g.,
- El eM l u
f we 1,,"g -
l B
.% 3 g ^8 l
- El *E l {l X M ,e B B Zgjg i
i l*i ggW ), I[----I g6 I Y'
Z8 l g .!l
- tl *:t l gl l 2 ,_ _* ]:l+]:l, i
l 1 l l
Fig.1 San Onofre High Pressure Injection System (source: San Onofre Nuclear Generating Station Units 2 and 3. Individual Plant Examination). [CCW is component cooling water system, CS is containment spray, LPSI is low-pressure safety injection, RCS is reactor coolant system, and RWST is refueling water ;
storage tank.] l 7
]
1 1
I
LER No. 361/98-003 8g 5 8 8 6 8 8 6 8 8 8 8 l $ - - -
l! 8 I
e f! !
o!
lI l I$
t ]g 8
ll1 p =
i Fig. 2 Dominant core damage sequence for LER No. 361/98-003.
8
P,
, , l LER No. 361/98-003 Table 1. Definitions and ProbaHlities for Selected Basic Events for LER No. 361/98-003 Modified Event Base Current for this name Description probability probability Type event 4 IE LOOP Initiating Event-loss of offsite 1.1 E 005 1.1 E-005 No power (LOOP)(includes the i Probability ofRecovering OfTsite Power in the Short Term) q IE-SGTR Initiating Event-Steam Generator 2.1 E-006 2.1 E-006 1
No {
Tube Rupture (SGTR)
IE-SLOCA Initiating Event-SLOCA 1.6 E 007 1.6 E-007 Yes IE-SSLOCA Initiating Event-SSLOCA 2.1 E-006 2.1 E-006 NEW Yes
[
IE-TRANS Initiating Event-Transient 6.2 E-004 6.2 E-004 No (TRANS)
CCW TRNB-FAIL Train B CCW lleat Exchanger is 0.0 E+000 4.0 E-001 NEW Yes not Retumed to Sersice (Case 2)
IIPR-MOV-CF-SUMP Common-Cause Failure of Sump 1.1E-003 8.8 E-002 Yes Isolation motor operated valves (MOVs) :
IIPR-SMP-FC-SUMPA Containment Sump Train A 6.1 E-003 1.0 E+000 TRUE Yes f Failure (Valve HV 9305 Stuck Closed)
HPR-XCONN XIIE-NR Operator Fails to Cross-Connect 2.0 E-001 2.0 E-001 NEW No !
IIPSISuction from Train B to Train A l l IIPR-X11E-NOREC Operstor Fails to Recover the 1.0 E+000 1.0 E+000 No IIPR System
^
f
\
IIPR XIIE-XM-IILEO Operator Fails to Initiate liot-Leg 1.0 E 003 1.0 E-003 No 1 Recirculation f
PCS-VCF-l!W Failure ofEquipment Required 1.0 E 003 1.0 E-003 No for Plant Cooldown PCS-XIIC-XM CDOWN Operator Fails to Initiate 1.0 E-003 1.0 E-003 No Cooldown 1 J
PPR-SRV-CO-TRAN Safety Relief Valves (SRVs) 2.0 E-002 2.0 E-002 No Open During a Transient
)
PPR-SRV-OO-1 SRV 1 Fails to Rescat 1.6 E-002 1.6 E-002 No PPR-SRV-OO 2 SRV 2 Fails to Rescat 1.6 E-002 1.6 E-002 No i
9 l l
l
l
. . l LER No. 361/98-003
< 1 Table 1. Definitions and Probabilities for Selected Basic Events for LER No. 361/98-003 (Continued)
Modified Event Base Current for this name Description probability probability Type event RiiR-MDP-CF-AB Common-Cause Failure of RIIR 5.6 E-004 5.6 E-004 No '
Motor-Driven Pumps RIIR-MOV CF-IIX Common-Cause Failure of R1IR 1.1 E-003 1.1 E-003 No IIcat ExchangerIsolation MOVs RIIR-MOV-CF-SUC Common-Cause Failure of RifR 1.3 E 003 1.3 E-003 No Suction MOVs R1IR-PSF-VF-BYP Flow Diverted From IIcat 9.0 E 003 9.0 E-003 No Exchangers or Reactor Vessel RIIR-XIIE NOREC Operator Fails to Recover the 3.4 E-001 3.4 E 001 No RIIR System RilR-XIIE-XM Operator Fails to Actuate the 1.0 E-003 1.0 E-003 No R1IR System j
10
n ,
LER No. 361/98-003 Table 2. Sequence Conditional Probabilities for LER No. 361/98-003 Conditional Event tree Sequence core damage Core damage Importance Percent name number probability probability (CCDP-CDP) contribution' (CCDP) (CDP)
SLOCA 02 6.0 E-006 1.9 E-007 5.8 E-006 89.4 SSLOCA 03 4.0 E-007 1.3 E-008 3.9 E-007 6.0 SSLOCA 05 1.6 E-007 4.8 E-009 1.5 E-007 2.3 TRANS 05 7.6 E-008 2.4 E-009 7.3 E-008 1.1 Sub+;tal Case 1 (shown)' 4.3 E-005 3.7 E-005 6.5 E-006 #f O Subtotal Case 76 3.2 E-006 2.4 E-006 >
7.0 E-007 '
Total (all sequences) x 4.6 E-095 3.9 E-005 7.2 E-006
(% 4 *}
qqgry
' Case I represents the increase in the CDP because of the long-tenn unavailability of the Train A containment emergency sump outlet valve HV-9305 (404.9 h).
b Case 2 represents the increase in the CDP because of maintenance being performed on the Train B CCW heat exchanger while the Train A containment emergency sump outlet valve HV.9305 was unavailable (27.1 h).
'Because case I presents the largest contribution to the total importance, the reported dominant sequences are ordered e: cording to the importance ofcase 1.
I1 1
- y*, .
LER No. 36I/98-003 Table 3. Sequence Logic for Dominant Sequences for LER No. 361/98-003 (Case 1 Only)
Event tree name Sequence Logic number SLOCA 02 /RT, /AFW, /HPI, HPR SSLOCA- 03 /RT,/AFW, /HPI, /COOLDOWN, RHR, HPR SSLOCA 05 /RT,/AFW, SIPI, COOLDOWN, HPR TRANS 05 /RT, /AFW, SRV, SRV-RES, SIPI,
/COOLDOWN, RHR, HPR Table 4. System Names for LER No. 361/98-003 (Case 1 Only)
System name Logic AFW No or Insumcient Auxiliary Feedwater System Flow COOLDOWN Reactor Coolant System Cooldown to RHR Decay Heat Removal Mode of Operation HPI No or Insumcient HPSI Flow HPR No or Insumcient HPR Flow RHR No or Insumcient RHR System Flow RT Reactor Fails to Trip SRV SRVs Open During a Transient SRV-RES SRVs Fail to Rescat 12
) , .,
LER No. 361/98-003 Table 5. Conditional Cut Sets for Higher Probability Sequences for LER No. 361/98-003 Cut set Percent number contribution CCDP' Cut setsd SLOCA Sequence 02 6.0 E-006
~
dMDM' ~ s;
- gi e 9[ '
I 96.6 5.7 E-006 IIPR-MOV CF-SUMP, HPR XIE-NOREC 2 1.1 6.5 E-008 IIPR SMP-FC-SUMPA,IIPR XIE XM-HLEO SSLOCA Sequence 03 4.0 E-07 E' N! ' 4 sd J l ,. f 1 57.2 2.3 E-007 RIEPSF-VF BYP,RHR XIE-NOREC,IIPR-MOV-CF SUMP, HPR-XIIE-NOREC 2 18.7 7.4 E-008 RiiR-XIIE-XM. IIPR-MOV-CF-SUMP. IIPR-XI E-NOREC 3 8.4 3.3 E-008 RI EMOV-CF-SUC, RIIR-XIENOREC, IIPR-MOV-CF-SUMP, liPR XIE-NOREC 4 6.7 2.8 E-008 RIIR MOV-CF-IIX,RIIR XIENOREC,IIPR-MOV-CF-SUMP, HPR-XIE-NOREC 5 3.6 1.4 E-008 RIEMDP-CF-AB,RIIR-X1E-NOREC,IIPR MOV-CF SUMP, IIPR XIE NOREC SSLOCA Sequence 05 **
1.6 E-007 %
1 48.1 7.4 E-008 PCS XIIE-XM<DOWN,IIPR-MOV-CF-SUMP, HPR-XI-E-NOREC 2 48.1 7.4 E-008 PCS-VCF-HW, HPR-MOV-CF-SUMP, IIPR-XI E-NOREC TRANS Sequence 05 7.6 E-008 m, ,1'
- 1 28.6 2.1 E-008 PPR SRV-CO-TRAN,PPR SRV-OO-1 RIEPSF-VF-BYP, RIIR-XFE-NOREC,IIPR-MOV CF-SUMP,IIPR-XIE NOREC 2 28.6 2.1 E-008 PPR SRV-CO-TRAN,PPR SRV OO-2,RIEPSF VF-BYP, RIIR-X1E-NOREC,IIPR MOV-CF-SUMP,IIPR-XIIE-NOREC 3 9.3 7.0 E-009 PPR SRV CO-TRAN,PPR-SRV-OO-1,RIIR XIE XM, liPR-MOV-CF-SUMP,IIPR-XIE-NOREC 4 9.3 7.0 E-009 PPR SRV-CO-TRAN, PPR SRV-OO-2, RIEXIE XM, HPR MOV-CF-SUMP. IIPR XIE NOREC 13 l
l
i LER No. 361/98-003 Cut set Percent number contribution CCDP' Cut setsd 5 4.2 3.1 E-009 PPR-SRV-CO-TRAN,PPR SRV-OO-1,RHR-MOV-CF-SUC, RHR-XIENOREC, HPR-MOV-CF-SUMP, HPR-XIENOREC Table 5. Conditional Cut Sets for Higher Probability Sequences for LER No. 361/98-003 (continued)
Cut set Percent number contribution CCDP" Cut setsd 6 4.2 3.1 E-009 PPR SRV CO TRAN, PPR-SRV OO-2, RHR MOV-CF-SUC, RIIR-XHE-NOREC,llPR-MOV-CF-SUMP,HPR XIE-NOREC 7 3.4 2.6 E-009 PPR-SRV-CO-TRAN, PPR-SRV-OO 1. RHR-MOV CF-11X, RIIR XHE-NOREC,HPR-MOV-CF-SUMP,HPR-XHE NOREC 8 3.4 2.6 E-009 PPR SRV-CO-TRAN,PPR-SRV-OO-2,RHR-MOV-CF-IIX, RHR-X1IE-NOREC, HPR-MOV-CF-SUMP, HPR-XHE-NOREC 9 1.8 1.3 E-009 PPR SRV CO-TRAN,PPR-SRV-OO 1,RHR-MDP-CF-AB, RHR-XHE-NOREC,IIPR MOV-CF SUMP,IIPR-XHE-NOREC 10 18 1.3 E-009 PPR-SRV-CO-TRAN,PPR SRV-OO-2,RHR-MDP-CF-AB, RHR XHE-NOREC, HPR-MOV-CF SUMP, HPR-XHE-NOREC Subtotal Case 16 4.2 E-005 / ,, st%
- Nl[ %
(shown above) _ l' ~.l 'd
- Subtotal Case 2' 3,2 E-006 i ' ";;f .
, U , e Total (all sequences) 4.6 E-005 , ,,
^ a: ,
"The change in conditional probability (importance)is determined by calculating the conditional probability for the period in which the condition existed, and subtracting the conditional probability for the same period but with plant equipment assumed to be operating nominally. The conditional probability for each cut set within a sequence is determined by multiplying the probability that the portion of the sequence that makes the precursor visible (e.g the system with a failure is demanded) will occur during the duration of the event by
, I the probabilibes of the remaining basic events in the minimal cut set. This can be approximated by 1 c*, where p is determined by l multiplying the expected number ofinitiators that occur during the duration of the event by the probabilities of the basic events in that minimal cut set. The expected number ofinitiators is given by it, where lis the frequency of the initiating event (given on a per-hour basis), and t is the duration time of the event. This approximation is conservative for precursors made visible by the initiating event. The frequencies ofinterest for this event are ArRANs = 6.2 x 109h,1w, = 1.1 x 10 th,1 m., = 1.6 x 10 9h,lasi.oc, = 2.1 x 10 th, and lam = 21 x 109h 14
LER No. 361/98-003 b
Case 1 represents the increase in the CDP because of the long-term unavailability of the Train A containment emergency sump outlet valve (404.9 h).
' Case 2 represents the increase in the CDP because of Train B CCW heat exchanger maintenance while the Train A containment emergency sump outlet valve was unavailable (27.1 h).
j
(
' M event HPR-SMP-FC-SUMPA is a TRUE type event which is not normally included in the output of fault tree reduction programs but has been added to aid in understanding the sequences to potential core damage associated with the event.
4 15
El,,
o.,
LER No. 361/98-003
' LER No. 361/98-003 Event
Description:
Inoperable sump recirculation valve Date of Event: February 5,1998
. Plant: San Onofre, Unit 2 Licensee Comments
Reference:
Letter from A. E. Scherer, Manager of Nuclear Regulatory Affairs, Southern California Edison, to U. S. Nuclear Regulatory Commission, " Comments on Preliminary Accident Sequence Precursor Analysis of Linestarter Failure, San Onofre Nuclear Generating Station, Units 2 and 3," April 27,1999.
Comment 1: Modeling Assumptions and Analysis Results: For Case 1 of the Accident Sequence Precursor (ASP) analysis [404 h,55 min with only the Train A containment emergency sump outlet valve failed (valve HV 9305)], the conunon-cause failure probability of " Sump Isolation Motor-Operated Valves" failing to open was increased from 1.1 x 10 5 to 8.8 x 10 2, based on the potential of the other sump isolation valves to fail also due to grit.
In response to NRC Special in?pection Report 50-36H98-05, $0-362 98-05 (Ref.1),
Southern California Edison (SCE) provided additional information relative to the estimated number of valve actuations (17,000) in the presence of the grit with only the one failure.2 Based on this plant-specific data, there was no evidence that the potential for common-cause failure was impacted by the grit. Therefore, SCE concludes that the increase in common-cause probability is not appropriate.
Response 1: LER No. 361/98-003 indicates that grit was found in numerous Motor Control Center (MCC) cubicles. Previous operational succes's of the Square D linestarters shows the durability of this equipment. However, because ofgrit being discovered in multiple locations, operational success did not preclude the grit failure mechanism from simu!taneously affecting more than one linestarter. In fact, multiple safety components were affected by this failure mechanism; however, a list of affected equipment was not available. ~Iherefore, only the sump isolation motor-operated valves were considered in this analysis. No supporting evidence was presented to suggest that the failure of the Train A containment emergency sump outlet valve linestarter was unique tojust this one linestarter. No change to the analysis methodology was made based on this comment; however, additional justification for increasing the common-cause failure probability from the base case has been provided in the Modeling Assumptions.
1 July 14,1999 ENCLOSURE 2
G)L .. .
" '4 ! . ,
l LER No. 361/98-003 i
Comment 2: Modeling Assumptions and Analysis Results: Case 3 ofthe ASP analysis [5 h,45 min with l the Train A containment emergency sump outlet valve failed (valve HV-9305) and the Refueling Water Storage Tank (RWST) Train B outlet valve unavailable because of ,
i maintenance (valve HV-9301)] models the unavailability of the RWST Train B outlet valve !
as a failure of that valve in the closed position thus preventing Train B injection. The RWST Train B outlet valve unavailability included in the LER was for replacement of the breaker-position indicating light. During this maintenance, the valve was open and available to allow injection. Therefore, from a probabilistic risk assessment (PRA) perspective, the 5 h and 45 min stated in the LER is not risk significant. This conclusion was included in Inspection Repon 98-05 (Ref. 2).
Response 2: Because the RWST Train B out' t valve was in the open position and available during the injection phase of an accident, Case 3 was removed from the analysis. In addition, item 5 in Additional Event-Related Information now indicates that the outlet valve was unavailable because of replacement of the breaker-position indicating light. His did not result in any change in the estimated overall imponance of this event because the importance estimated for 4 l Case 1 increased slightly (1.0 x 10 ), since its conditional time period was adjusted from 399 h,10 min to 404 h,55 min [i.e., the period with only the Train A containment em ergency sump outlet valve failed (HV-9305)J.
- Comment 3: Additional Event Related Information: The first paragraph infers that thelinestarter replacement had just beg n in February 1998. The linestartes replacement began in 1995 after the linestarter failure due to excessive wear of the sliding cams. The linestaner replacement was completed in 1998. It is' suggested that "just staned"in the first sentence be replaced with "been" to reflect the correct timing.
Response 3: This editorial change was made to reflect more accurately the timing of the linestaner replacement effort. The first sentence now reads "He licensee had been programmatically replacing all of the Square D linestarters - 60 of 86 linestarters in Unit 2 and 61 of 86 linestarters in Unit 3 had already been replaced."
Comment 4: Additional Event-Related Information: He second paragraph includes information from LER No. 361/98-003, which indicated that the source of the grit was most likely gunite 2 July I4,1999
O~
- * 'll . a LER No. 361/98-003 similar to what was used for hillside stabilization. In the response to NRC Inspection Report 98-05, additional infonnation indicated that the grit substance had the same elemental sisual and size characteristics as unmixed Portland cement. The response to NRC Inspection Report 98-05 further indicates that the deposition ofthe grit took place during construction actisities and that potential sources included batch plant operation and grouting and guniting actisities.
It is suggested that the reference to gunite used to stabilize the hillside in the first sentence of the second paragraph be deleted and replaced with " . identified as Portland cement particles
... " In addition, it is suggested that the response to NRC Inspection Report 98-05 be referenced at the end of the second paragraph and added to the reference section.
Response 4: This editorial change was made to reflect more accurately the nature of the grit source. The first sentence of the second paragraph in this section now reads, "The grit that caused the linestarter for the Train A containment emergency sump outlet valve to jam was identified as 2
Portland cement particles " The licensee's response to NRC Inspection Report 98-05 has been added to the list of references as Reference 2 in the analysis.
References l
L NRC Specialinspection Report 50-361/98-05; 50-362/98-05, May 21,1998.
(
- 2. Letter from Dwight E. Nunn, Vice President, Southern California Edison, to U. S. Nuclear Regulatory j Commission, " Response to NRC Inspection Report 98-05 Regarding Linestarters San Onofre Nuclear Generating Station, Units 2 and 3," June 22,1998.
I i
i i
l l
3 July 14,1999 l
l l
l l