ML20217G674
| ML20217G674 | |
| Person / Time | |
|---|---|
| Site: | Prairie Island  |
| Issue date: | 10/09/1997 |
| From: | Wetzel B NRC (Affiliation Not Assigned) |
| To: | Wadley M NORTHERN STATES POWER CO. |
| References | |
| NUDOCS 9710140133 | |
| Download: ML20217G674 (17) | |
Text
..
October 9, 1997 Mr. M.D. Wadley Vice President, Nuclear Generation Northern States Power Company 414 Nicollet Mall Minneapohs, MN 55401
SUBJECT:
RLViEW OF PRELIMINARY ACCIDENT SEQUENCE PRECURSOR ANALYSIS OF EVENT AT THE PRAIRIE ISLAND NUCLEAR GENERATING PLANT
Dear Mr. Wadley:
Enclosed for your information is a copy af the final Accident Sequence Precursor (ASP) analysis of the operational event at the Prairie Island Nuclear Generating Plant reported in
" Licensee Event Report (LER) No. 282/96-002. 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 proliminary analysis and comments received from the NRC staff and from our independent contractor, Sandia National Laboratories (SNL). contains our responses to your specific comments. Our review of your comments employed the criteria contained in the material that accompanied the preliminary analysis. The results of the final analysis indicate that this event is a precursor for 1996.
Please contact me at 301415-1355 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.
NI)kSIGNEDBY Beth A. Wetzel, Senior Project Manager Project Directorate 111-1 hD Db!k050 282 Division of Reactor Projects - lil/IV s
PDR Office of Nuclear Reactor Regulation
- Docket Nos. 50-282 and 50-306
Enclosures:
As stated cc/w encia: See next page OFOI DISTRIBUTION w/encis:
ADocket FW PUBLIC Al PD# 3-1 Reading EAdensam (EGA1)
OGC ACRS JMcCormick-Barger, Rlli w/o enc!s:
PO'Reilly SMays DOCUMENT NAME: G:\\WPDOCS\\ PRAIRIE \\PIASP.LTR To receive a copy of this document, indicate in the box C= Copy w/o attachment / enclosure E= Copy with attachment / enclosure N = No copy OFFICE PM:PD31 E
LA:PD31 C
D:PD31 INAME BWetzel:dbh CJamerson h JHannok ll' lf l' llllllll ll DATE
/0/ 8 /97 to / 6 /97 h e /1/97
' ~
OFFICIAL RECORD COPY dbO1 k
kh[ hh s
Mr. M. D. Wadley, Vice President Prairie Island Nuclear Generating Northem States Power Company Plant
~
cc:
(
J. E. Silberg, Esquire Site : : censing Shaw, Pittman, Potts and Trowbridge Prairie Island Nuclear Generating 2300 N Street, N. W.
Plant
- Washington DC 20037 Northem States Power Company 1717 Wakonado Drive East Plant Manager Welch, Minnesota 55019 Prairie Island Nuclear Generating Plant Tribal Council Northem States Power Company Prairie Island Indian Community 1717 Wakonade Drive East ATTN: Environmental Department Welch, Minnesota 55089 5636 Sturgeon Lake Road Welch, Minnesota 55089
- Adonis A. Nebiett Assistant Attomey Genera'
. Mr. Roger O. Anderson, Director Office of the Attomey General Nuclear Energy Engineering
~
455 Minnesota Street Northem States Power Company Suite 900 414 Nicollet Mall St. Paul Minnesota 55101-2127 Minnaapolis, MN 55401 U.S. Nuclear Regulatory Commission Resident inspector's Office 1719 Wakonade Drive East Welch, Minnesota 55089-9642 i
Regional Administrator, Region 111 l
U.S. Nuclear Regulatory Commission 801 Warrenville Road Lisle, Illinois 60532-4351 Mr. Jeff Cole, Auditor / Treasurer Goodhue County Courthouse Box 408 Red Wing, Minnesota 55066-0408 Kris Sanda, Commissioner Department of Public Service 121 Seventh Place East Suite 200 St. Paul, Minnesota 55101-2145 i
NovemDer 1996
.-. m
i LER No. 282/96-017 LER No. 282/96 012 Event
Description:
Las of offsite phwer to safeguards buses on both units Date of Event: June 29,1996 Plant: Prairie Island I and 2 L ent Summary Both units were operating at 100% power on June 29,1996, when strong isolated thunderstorms caused the failure of three 345 kV ofTW transmission lines to the plant substation. Both tmit scactors tripped in response to a loss ofloed. All four emergency diesel generators (EDOs) started as expected. Both Unit 2 safeguards buses and one Unit I sefcguards bus were immediately powered by their respective EDGs. A single 345 kV transmission line continued to supply power to the plant substation, the Unit I normal buses, and one Unit I safeguards but (bus 15) However, the voltage supplied by the single offsite power supply line was so unstable that approximately 7 min after the trip, bus 15 automatically transfened to its associated EDO. A stable ofTsite source was not reesteblished for approximately 5 h, and both units were cooled by natural circulatica cooling until offsite power was restored (Ref.1) The estimated conditional core damage probability (F SP) for this grid based lots of offsite power (LOOP)is 5.3 = 10*. This CCDp estimate is applicable to
.1, units.
e Event Description On June 29,1996, at approximately 1418, the Blue Lake 345 kV transmission line tripped off line and remained offline following a single phase to ground fault resulting from a passing thuriderstonn Both uruts n ained at 100% power following the loss of this offsite connection to 345 kV bus I in the plant substation, shown in Fig.1. Approximately iI min later, at 1429, severe weather with " straight line" winds destroyed several support structures, causing three phase faults in both Red Rock lines. Both Red Rock 345 kV lines tripped. At this point, only the Bpon 345 kV line reinained in service and powered 345 kV bus 2 in the Prairie Island switchyard. Both tmit geacrators were aligned to 345 kV bus I and subsequently tnpped from 100% powr because of the loss ofload. All reactor coolant pumps (RCPs) on both units tripped because of low frequency, resulting in decay heat being removed by natural circulation cooling only.
All four EDGs started as designed upon the LOOP. 'lafeguards buses 16 (Unit 1) and 25 and 26 (Unit 2) were immediately sequenced onto their respectise EDGs. Safeguards bus 15 (Unit 1) continued to be supplied by offsite power via the Byron line. Voltage on the Byron line was low and unstable, and safeguards bus 15 automatically loaded onto EDO Di approximately 7 min afler the reactor trip. Unit I nonnal 4 kV buses.
(non safety related) transferred to the IR transfonner supplied from the 345 kV bus 2 via the Byron line gnd continued to be powered from this unstable source throughout the event. The 2 normal 4 kV buses at Unit 2 transferred to the 2R transformer which is supplied power from the 345 kV bus I; this bus uas deenergired at the time. At this time, the licensee decla::d an Unusual Esent.
I ENCLOSURE 1
LF.R No. 282/96 012 l
At approxirnately 1800, the Unit 2 normal 4.kV buses were powered from 345 kV bus 2, which was still in a degraded voltage condition (below 330 kV). At 1925, voltage on the Byron line was restored within the normal range after the utility purchased additional power and reset transfonner taps in the switchyard.
Because all EWh cre operating satisfactorily, emphasis was placed on restoring forced circulation cooling in both units rather than transferring the safeguards bus power supply to the appinpriate oEsite source. This priority was established because, in case of an EDO failure with an offsite powet source available, the load sequencers would have automatically transferred safeguards bus loads to the offsite power source. At 1953, forced circulation cooling was eshblished at Unit 1. At 2028, forced circul6n cooling was established at i
Unit 2.
On June 33,1995, at 0135, the Blue Lake 345 kV line was restor,d. The process of transferring the safeguards '>uses from the EDGs to an offsite power source Legan at this point and was campleted at 1035.
The licensee exited the Urmsual Event at this point.
Additiotial Event Related information The diesel driven cooliag water pumps that provide plant service water started as designed during the event to supply a source of cooling water throughout the event. Additionaby, the Unit I motor.d.iven cooling water pump operated throughout the esent while powered from the degraded offsite power source. Therefore, a heat sink was available throughout the esent for the component cooling water system, the EDGs, and the auxiliary feedwater (AFW) pumps (Ref. 2) All AFW pu.rps operated as designed.
The EDGs on each unit can be cross tied to the opposite unit using two 4160 V breakers connected in series.
Also the motor driven AFW pump on each unit can be cross tied to the opposite unit through two manually operated valves (Ref. 2).
Modeling Assumptions Five hours passe 3 before a stable offsite power supply at normal voltage was reestablished. This did not occur until the utility purchased additional power and reset switchyard trsnsformer taps, if the EDGs had not started as expected, the utility may hase expedited the retum of the oHsite Byron 345 kV line, but it is not known if this would have been possible. Therefore, the Byron line uas not considered a viable offsite power source for the safeguarde buses, and the event was modeled as a grid centered LOOP on each unit. The probability of not recovering offsite power in the short term is included in the initiating event probability (IE.
LOOP). This term was set to the probability for a grid based LOOP assummg operators fail to recover offsite power in the short term (4.8 = 104).
The grid-based LOOP probability of short term and long term offsite power recovery for a grid entered LOOP and the probability of a RCP seal loss of coolant accident (LOCA) following a postulated station blackout (SBO) were developed based on data distributions contained in NUREG.1032, Evaluation ofStation Blackout Accidents atNuc/ car Powerplants (Ref 3) The RCP sea! LOCA models were developed as part of the NUREG ll50 probabilistic risk assessment (PRA) efforts. Both are described in Revised LOOP Recovcry and rWR Seal LOCA Models (Ref. 4). The probabilities for the following basic events (defined 2
.,~.
-4~.,
-m
_.. ~..
,-,---e-.
--e-
,,,,--g
-v
-- - - -- - - -,,e-a.
LER No. 282/96 012 in Table 1) are based on these models: IE. LOOP, OEP.XilE NOREC.6H, OPE.XilE.NOREC.BD, OPE.XilE.NOREC.SL, and RCS.MDP.LK SEALS.
Each of the four EDGs at Prairie Islandian supply the power requirements for the hot shutdown loads for l
its associated unit and one train of essential loads of the opposite unit in case of an SB0 on the second unit (Ref. 5). This is accomplished through two manual cross tie breders between buses 15 and 25 or buses 16 and 26. A basic event was added (EPS.XHE.XE.XTIE) to the model to account for the failure of the operator to initiate the cross tie between buses according to the established procedure (Ref. 6). This cross tic basic event assumet. that a safety injection signal does not exist on both units at the same time.
EPS.XHE.XE XTIE was set at 3.2 = 10'8 based on the individual plant examination (IPE, Ref. 5, Table 3.3 3). A basic esent was also added to account for the mechanical failure of the two cross tic breakers in series (Ref. 3, Table 3.31); however, this event did not influence any of the significant core damage sequences. The base case common cause failure probability of the EDGs (EPS. DON.CF.ALL) was adjusted from 1.6 = 10'8 to 7.0 = 10 to account for all four EDGs (Ref. 7. Tc.ble 5 9, a., = 0.0164; and Table 512, d
a., = 0.0174).
The IPE (Ref. $, page 2 7) indicates that the first station battery will fail aner 2 h Because a stable offsite power source was not restored until about 5 h into the LOOP, basic esent OEP.XHE NOREC.2H was set to TRUE (i.e., the probability of this event is 1.0 ghen that power was not restored in the short term). This hed little impact on the CCDP calculated for this event.
Each motor driven AFW pump can be cross tied to supply feedwater to the opposite unit. A basic event uas added (AF\\WXHE.XEMflE) to the model to account for the failure of the operator to initiate the cross tic between units. AF\\WXHE.XE.XTIE was set at 3.2 = 10-8 (Ref. 5, Table 3.3 3). A basic esent was also added to account for the potential failure of the cross tie vahes.
Analysis Results 8
The CCDP for this event is estimated to be 5.3 = 10. The dominant core damage sequence for this ment (sequence 28 an Fig. 2) involves
- aLOOP, a successful reactor trip, failure of the emergency power supplies (5B0),
success of the AFW s) stem, no challenge to the power-operated relici valves (PORVs),
failure of the RCP seals during the LOOP, and
=
failure to recover ofTsite power aner the RCP seals fail.
This sequence accounts for about 38% of the total contnbution to the CCDP. Sequence 37 is similar to LOOP sequence 28, except LOOP sequence 37 involves a PORY lift and successful reclosure. Combined, these two sequences account for 60% of the total contnbution to the CCDP.
3
l 1.ER No. 282/96 012 Sequences involving battery depletion (sequences 21 and 30) account for 16% of the total contribution to the CCDP. An SBO is invohed in 98% of the dominant core damage sequences. Failure of the AFW system only occurs in 8% of the dominant core dr.nage sequences.
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 logie 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.
In a sensitivity study, the probability of failing to recover offsite power in the short term (IE. LOOP) was set to TRUE [ LOOP with no short. term (30 min) recovery possible] SBO values were adjusted to account for the revised period based on the code sssociated with the Rcvncd LOOP Rccovery and fil7f Scal LOCA Afodcls (Ref. 4). The calculated CCDP in this case is 6.2 = 10. The dominant sequence remained the same 4
for this sensitivity study.
Finally, the IPE indicated that an additional 2 h was available to recover offsite power before core damage following battery depletion. An estimate of the CCDP assuming 2 h before battery depletion and an additional 2 h before uncovenng the core is calculated to be 4.1 = 10 Again, adjusted station blackout 4
values were based on the code associated uith the Revssed LOOP Recovery and Pillt Seal LOCA Afodels (Ref. 4) The dominar.t sequences remained the same as for the initial analysis.
Acronyms AFW auxiliary feedwater system CCDP
- onditionni core damage probabihty EDO emergency diesel generator IE initiating event IPE individual plant examination kV kilova LER licensee event report LOCA loss of coolant accident LOOP loss of offsite power PORV power operated relief valve RCP reactor coolant pump SBO station blackout References
- 1. LER 282/96 012, Rev. O, " Loss of Offsite Power to Unit 2 and degraded Offsite Power to Unit i Followed by Reactor Trips of Both Units," July 29,1996.
- 2. Northern States Power Company, Prairie Island Nuclear Generating Plant, Updated Safety Analysis Report.
4
LER No. 282/96-012 4-
- 3. Evaluation ofStation Blackout Accidents at Nuclear Power Plants, NUREG 1032.
- 4. Revised 1.00PReconryandPMR&,al1.0CA Models, ORNLMRCILTR 89111, August I989.
$. Prairie Island Nuclear Generating Plant,/ndwidualPlant Examination.
6, Prairic Island Procedure IECA.O.0, Rev.11." Loss of All Safeguards AC Power."
- 7. Common-Cause Failure Ikta Collection andAnalysis System,INEL 9410064, December, l995.
5
LER No. 282/96 012
(
e-lww:
w w.
wn :
im
- $ "Y
- 8
/
W,....
.l
-l l
/
/
/
I)
I)
/l)
I I
/l
/
/
I I
i
/
/
l)
I)
I) l l
/l
/.
/.
i-o i
i
- t. w is
/,
/
/
l) l)
l
/l) '
I I
/
/
l l
l>.sxvui s
la Caesseem te.dems Urst 3 thna l
- i. ii. i.elarmee Ares tru
- t. w ri o
o Fig.1. Switch >wd at Prairie Island.
6-
....;,.. :~_.-
LER No. 282/96 012 ll 5555858888855885588885585888885588888888 i
...*..esesserts:AnnaAnshanassaanskase
.l lll [
L lll l I
11 l ll l lil !
l l
7-Ill; :
11ll !
Il il :
11ll nii !
l' c
11 i m
f
.k I
I ;l i El W
g b
lt[ !
II !
I Fig. 2. Dominant core damage sequence for LER No 282/96 012, 7
l.
LER No. 282/96 012 Table 1. Definitions and Probabilities for Selected Basic Events for LER No. 282/96-012 s
Modified i
Event Base Current for this I
naine Description probability probability Type event l
IELOOP truusting Event-LOOP 3.9 E 006 4.8 E 001 ORID Yes LOOP i
l IE 50TR Irutsatmg Event-Steam 1.0 E 006 0 0 E+000 lONORE Yes Oenerator Tube Rupture IE$LOCA initiatmg Event-$ mall-Break 1.0 E 006 0.0 E@0 IGNORE Yes LOCA l
IE TRANS Initiatmg Eient-Transient 3.3 E 004 0.0 E@0 10NORE Yes Alv TDP FC 1DP Turtaine Dnvon AIV Pump 3.$ E 002 3.5 E 002 No Fails AIV X11E-NOREC EP Operator Fails to Recover Alv 3.4 E401 3 4 E 001 No System Durmg an $80 Alv XilE XE XTIE Operator fails to Cross Tie the 3.2 E 002 3.2 E 002 NE'V No Motor Drisen Alv Pump EPS-DON CF ALL Common Cause Failure of EDO:
1.0 E404 7.0 E 004 No LPS DON FC 1 EDO 1 Fails 4.2 E 002 4.2 E 002 No EPS DON FC 2 EDO 2 Fails 4.2 E 002 4.2 E 002 No EPS XilE NOREC Operator Fails to Recover 1.0 E+000 1.0 E+000 TRUE No Emergency Powcr EPS XilE XE XTIE Operator Fails to Cross Tee the 3.2 E 003 3.2 E 003 NEW No Safiguards ac Buses OEP XilE NOREC 2H Operator rails to Recover OITsite 2.1 E 001 1.0 E+000 TRUE Yes Power Within 2 h OEP XHE NOREC 6H Operator rails to Recover Offsite 9.9 E 002 3.6 E 004 ORID Yes Power Withm 6 h LOOP OEP-XHE NOREC BD Operator rails to Recover OtTsite 6.1 E 002 3.30002 GRID Yes Power Before Battery Depicuan LOOP OEP Xil".NOREC SL Operator Fails to Recover Ofhite 5.9 E 001 4.$ E 001 ORID Yes Power After a Seal LOCA LOOP PPR SRV CO SBO PORVs Lin Dunng an $80 3.7 E401 3.7 E 001 No PPR SRV.OO 1 PORY I fails to Reclose After 3 0 E 002 3.0 E 002 No Openmg 8
th
,e 1.Fk No. 282/96-012 Table 1. Definitions and Probabilities for Selected Basic Events for LER No. 282/96 012 Modified Event Base Current for this name Description probability probability Type event PPR SRV4G2 ICRV 2 Fails to R* low AAer 3.0 E@2 3 0 E@2 No D *nias i
RCS-MDP-LK SEALS RCP $ eels Fail Wiht Coolms 2.3 E o0i 2.1E001 GRID Yes and injection LOOP 9
t i
LER No. 282/96 012 Table 2. Sequence Conditional Probabilities for LER No. 281/96 012
\\
Conditional Event tree Sequence core damage Percent name number probability contribution (CCDP)
'.,00P 28 2.0 E.005 37.9 LOOP 37 1.1E005 22.2 LOOP 38 7.5 E 006 14.1 LOOP 21 5.$ E 006 10.4 LOOP 39 4.0 E 006 7.$
LOOP 30 3.2 E.006 6.)
Total (all sequences) 5.3 E 005 f
I d
10 ew
-,... w n,. -_-
- - ~. -- - - - ~ ~ ~ ~ - ~ ~ - - -. - ~ ~ -
ee m'w weyw-g.,--.->
-w-.m v
---ree.
y-----y wr
+.
-r----.y.m9 yvrw m-
--g=*
F S
y-y w
y-v
__.m.____
P LER No. 282/%012 Table 3. Sequence Logic for Dominant Sequences for LER No 282/96 012 Event tree Sequmce name number Logic LOOP 28
/RT.L EP,/AFW.L.EP,/PORV.SBO, SEALLOCA, OP.SL LOOP 37
/RT L. EP, /AFW.L.EP, PORV.SBO,
/PORV.EP, SEALLOCA, OP.SL LOOP 38
/RT.L. EP, /AFW.L.EP, PORV.SBO, PORV.EP LOOP 21
/RT.L. EP, /AFW.L.EP, /PORV.SBO,
/SEALLOCA, OP.BD LOOP 39
/RT.L. EP, /AFW.L.EP, PORV.SBO,
/PORV.EP,/SEALLOCA, OP.BD Table 4, System Names for LER No 282/96 012 System name Logic AFW.L.EP No or Insumcient AFW Flow During an SBO EP Failure of Both Trains of Emergency Power OP.BD Operator Fails to Recoser Offsite Power Before Battery Depletion OP.SL Operator Fails to Recover Offsite Power After a Seal LOCA PORV.EP PORVs Fail to Reclose (No Electric Power)
PORV.SB0 PORVs Open During an SB0 RT.L Reactor Fails to Trip During a LOOP SEALLOCA RCP Seals Fail During a LOOP 11
i !
LER No. 282/96-012 l
l Table 5. Conditional Cut Sets for fligher Probability Sequences for LER No. 282/96 012 I
j Cut set Percent 1
number contribution CCDP' Cut sets' LOOP Sequence 28 2.0 E 005 1
98.8 2.0 E 005 EPS-DONCF ALL, EPS XilE NOREC, RCS MDP LK SEALS,
/PPR SRV40-SDO,OEP XilE NOREC SL 2
0.8 1.6 E 007 EPS don rC l. EPS DON TC 2. EPS XilE XE MIE, EPS X11E NOREC, RCS MDP LK SEALS,
/PPR SRV CO SBO,OEP X1tE NOREC SL LOOP Sequence 37 1.2 E 005 I
1 98.8 1.2 E 005 EPS DON CF ALL, EPS XilE NOREC, RCS MDP-LK SEALS, PPR SRV CO SHO,OEP XilE NOREC SL l
2 0.8
- 1.0 E 007 EPS DON FC l, EPS DON 4C 2, EPS XHE XE nlE, EPS X1lE NOREC, PPR SRV40 SBO, RCS MDP LK SEALS, OLP X11E NOREC SL LOOP Sequence 38 7.5 E 006 1
49.4 3.7 E 006 EPS-DGN CF ALL, epi XilE NOREC, PPR SRV CO SBO, PPR SRV OO l 2
49.4 3,7 E 006 EPS DONCF ALL,EPS XilE NOREC,PPR SRV-CO SDO, PPR SRV 00 2 LOOP Sequence 21 5.6 E 006 I
98.8 5.5 E 006 EPS-DONCF ALL, EPS XilE NOREC,
/RLS MDP LK SEALS,/PPR SRV CO'SBO, OEP X1tE NOREC BD 2
0.8 4.4 E 008 EPS DGN IC l,EPS DON lC 2.EPS XilE XE XTIE, EPS XilE.NOREC,/RCS MDP LK SEALS,
/PPR SRV CO SBO,OEP XilE NOREC BD j
LOOP Sequence 39 4.0 E 006
^
l 99.4 4.0 E 006 EPS DGN CF ALL, EPS X1tE NOREC, Af%TDP FC TDP, AFW Xilt-NOREC EP
+
12
-y-,
m v
e--gm, w
.,y,
,.,,_--g9 m
--e-aem.
t o
LER No. 282/96-012 Table 5. Conditional Cut Sets for Higher Probability Sequences for LER No. 282/96-012
~
Cut set Percent number contribution CCDP' Cut sets
- LOOP Sequence 30 3,3 E 006 1
98.8 3.2 E-006 EPS don.CF.ALL EPS X1tE NOREC. PPR SRV CO SBO,
/RCS-MDP LK SEALS,OEP XIIE NOREC BD 2
0.8 2.6 E 008 EPS DON rC l. EPS-DOH-rC 2. EPS XI!E XE XIIE.
EPS XI!E NOREC PPR SRV CO SBO,
/RCS MDP LK SEALS,OEP XilE NOREC-BD Total (all sequences) 5.3 E-005
'The conditional probability for each cut ut is determined by multiplying the ped,abikty of the irutisting event by the protabihtses of the bas.c tvents in that minimal cut ut, The probabil.' ties for the initiatmg events and the basic events are given in Table 1.
6Basic event EPS XIIE lCREC is a type TRUE event and these type of events are normally not included in the output of fault tree Nuction programs, but has been added to aid in understandmg the sequences to potential core damage answinted with the event.
i l
13
---m,
1 i
LER No. 282/96-012 1
I I
LER No. 282/96-012
(
[
Event
Description:
. Loss of offsite power to safeguards buses on both units i
Date of Event: June 29,1996 1
4 Plant: Prairie Island I and 2 I
e Licensee Comments 1
Reference:
Letter from Joel P Sorensen, Plant Manager - Prairie Island Nuclear Generating Plant, Northern States Power Company, to U. S. Nuclear Regulatory Commission, " Response to Request for Review of Preliminary Accident Sequence Preewsor Analysis of Operational l
Event at Prairie Island," June 25,1997.-
i Co====t 1:
All areas appear to be acewate except for the PRA modeling credit given for the cross tie 4
- capability of the emergency ac power buses between units. Each diesel generator can supply t
the train related safe shutdown loads of both units. The operator is directed to perform this 4
i i,ction (called manual voltage restoration) in steps 8 and 9 of the emergency procedures for loss of all ac power. The Prairie Island IPE used a human error probability (HEP) of 0.0032
[
(IPE Table 3.3 3) for this action. It is r=== -vi that the ASP analysis be re perfonned, i
1 with credit for the ac cross tic capability being modeled as the operator HEP for perfonning j
the cross tic per the emergency procedwes.
- Response 1:-
Because each of the EDGs can supply the train related safe shutdown loads on both units, the Prairie Island model was reconfigured to allow credit for cross tying the emergency ac F
power buses toia. units. The operator HEP from the Prairie Island IPE (0.0032-Table 3.3 3) was used for the probability of the operator failing to cross tic the safeguards buses.
(EPS XHE XE XTIE). Because both units experienced a LOOP, the reliability of the opposite unit EDG was s o considered by modifying the base-case common cause failure.
probability to account for 4 EDGs (EPS DON-CF ALL). Additionally, the two 4160 V breakers in series were modelod; however, this event did not influence any of the significant core damage sequences.
Because the capability to cross tie the output of the motor driven AFW pumps between units exists, the Prairie Island model was further refined. The operator HEP from the Prairie Island IPE (0.032-Table 3.3 3) was used for the probability of the operator failing to cross-
?
tie the motor-driven AFW pump (AFW.XHE XE XTIE). The nominal motor driven AFW I
1 ENCLOSURE 2
-r
+--f
- .deM*e+rme-MrN'4--'>t
-'t
- f-
= we w ee--t--=e~'
+
Y-*v--m-
=++mtw--v-mw.ywe-
.Nr*-k'g---mm-tw'1-t<Eraes-r--GerTry--'Wv-tr---1+-+ft+r mw"Ws' tN'N-MTN**tF-Y M
-grhnN
t s
i LER No. 282/96 012 i
i pump failure probability was considered. Also, the failure probability of the two manual cross tie valves in series was considered. Neither of these last two basic events appear in any l
of the dominant core damage sequence cut sets, 1
Accounting for the ability to cross tic the EDGs and the motor driven AFW pumps reduced the CCDP by a factor of five to $.3 = 10'8. The dominant core damage sequence remained j
the same as before the model revisions.
1 5
a i
Comment 2:
It is requested that the ASP report acknowledge the limitations of the NRC models in the areas of syrtems that can be cross tied across units, since accurate modeling of these capabilities would lower the conditional core damage probability calculated for this event.
3 l
Response 2:
Similar to Section 2.5 of the previous ASP report, frecursors to forential Severe Core Damage Acc/ dents: 1995 A Status Report, NUREO/CR4674, Volume 23, the 19% status report will address the limitations of the ASP models. With respect to the Prairie Island i
j analysis, the Prairie Island model was reconfigured to properly credit cross tying emergency I
ac power buses and motor driven AFW pumps between units. As noted above, specifically
}
4 accounting for the ability to cross tic between units at P. ie Island lowered the CCDP by a factor of 5.
i a
i
}
h I
I I
2 n..
,,-.-.~.-.,,---n~-,----,---,---~,.,,.,-w,,_..,,,,--,.--,,-
c
.n,.
..,,,n,.,
-