ML20203B721
| ML20203B721 | |
| Person / Time | |
|---|---|
| Site: | Wolf Creek  |
| Issue date: | 12/08/1997 |
| From: | Maynard O WOLF CREEK NUCLEAR OPERATING CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| WM-97-0142, WM-97-142, NUDOCS 9712150138 | |
| Download: ML20203B721 (15) | |
Text
-
. - _ -. -... - -...... -... -.. - -. _... ~
... ~. ~. _.
I s
r i
j WQLF CREEK NUCLEAR OPERATING CORPORATION l
Otto L Maynard Presdent and Chef Execuwe Officer i
December 8, 1997 i
t l
WM 97-0142 U. S. Nuclear Regulatory Commission 1
ATTH:
Document Control Desk 5
l Mail Stat.icn Pl-137 j
Washington, D. C.
205L5 i
i References Letter dated October 8, 1997, from K. M. Thomas, NRC, to O.
L. Maynard, WONOC
Subject:
Docket No. 50-402:
Response to Request. for Additional Information hegarding the Wolf Creek Individual Plant Exemination of Externr 1 Events Subn.ittal Gentlement Attachert in Wolt Creek Nuclear Ope rating Cor1.orati<>n*.s (WCNOC) rcsponse tc the Requent for Additional Informaticn (RAI) containcd in the Reference for the i
Wolf Creek Generating Station (WCGS) Individual PJant Exarnination of External Event s (IPECE). The RAI concerns questiens in the areas of seismic and fire analysis.
The att aichment prov. ides the speci fic rerponses to the TsAI.
If yo.: should have any questions zegarding this submittal, please contact me at (316) 364-0831, extension 4000, or Mr. Michael J. Angus at extension 4077 l
Very truly yours,
/
[
Otto L. Maynard l
i OLM/jad Attachment cci W.
D. Johnson (NRC), w/a E. W.
Merschof f (NRC), w/a j
i J. F, Ringwald (NRC), w/a K. M. Thomas (NRC), w/a l
9712150138 971200 PDR ADOCK 05000482 P
pog 3
ll i
l i
i.i.
. l.
.% S O m.
7QyVV PO. Don 411 i Durkngton, KS 66839 i Phone: (316) 364-8831 l
An tQual OmortuMy En.,*w M F IC VET f
3--w..
v
...m.
,,-.w.....m.,,,,
w my.
~,,..._.m....
-m-
....,c.w,,.-..,e
_. =
At t,a chment to WM 97-0142 Page 1 of 14 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION Seismic Analysis Questions Question 1:
NUREG-1407 (section 3.2.$.8, page 14) suggests that non-seismic failures and human actions are to be clearly identified and the assessment assure that they are low enough in probability to not compromise the seismic margins.
Typically, some of the more risk significant human actions for PWRs ares connect service water as an auxiliary feedwater source, feed and bleed, control steam generator relief valves for steam generator cooling, RCS cooldown and depressurization to use the RHR system, reduce containment spray pump flow, room cooling recovery, and establish cold leg recirculation.
The submittal does not provide sufficient assurance that human errors can. be accomplished af ter an SME.
Per NUREG-1407, show the analysis and logic used to verify that human actions do not control the probability of losing a success path af ter an earthquake.
For each human action, provide the location and the timing of performance.
In addition, explain what affect an earthquake would have on performance of these actions.
Response
Table 1 lists the human actions needed for the two success paths used in the IPEEE Seismic Margins analysis'.
The table was developed by reviewing the current WCGS Human Reliability Analysis calculation and determining which were needed to respond to a Small Break LOCA, which is the basis of the Seismic Margins success paths.
These actions are performed using seismically qualified, safety grade equipment.
OPA-CCWSERLP is the only action that may be af fected by a SME. OPA-CCWSERLP is a manual switchover of the service loop to the opposite train of CCW.
This action is necessary to provide cooling to the RCP seals, and as such only five minutes is allowed to perform the task. As noted above, the two success paths modeled for the Seismic Margins analysis assume a Small Break LOCA.
Failing to perform this task will only result in an eventual increase in the LOCA size due to a RCP seal failure. Also, because this action is ANDed with the failure of a CCW train, it has little impact on the overall core damage.
OPA-CCWHX is the only human _ action performed locally. It requires the operators to manually close a CCW Heat Exchanger Bypass valve (or a bypass isolation valve). The CCW Heat Exchanger Bypass valves are located in the Auxiliary Building.
With 30 minutes available' to reposition the valve, it is - judged that a SME will not affeet the success rate of this action.
variety of manual actions that are OPA-AFWACT models the failure of a required only when the corresponding automatic actions fail.
Since these actions are all ANDed with the automatic actions, their failures have little impact on overall core damage.
In addition, these actions are procedure driven, whereby operators are required to verify that an action has performed properly, and if not, then manually perform the action.
The majority of the remaining actions have time limits of at least 30 minutes. With this amount of time available, it is not likely a seismic 4
=-
r-mi-um
- pig, s
7 e-.-
e q
-w-
<mec
--vr w
rr tc
.
- Att;achment to WM 97-0142 Page 2 of 14 event will be a factor in their success / failure.
OPA-CCWSERLP, OPA-HPR and OPA-LPR have less than 30 minutes time available to complete.
The
. injection-recirculation switchover actions do not take place until after the Refueling Water Storage Tank (RWST) is emptied, which (for a concurrent Small Break LOCA) is well after a seismic event would have terminated.
Becaus-of this, their failure probabilities would be similar to that used in the IPE analysis.
TABLE 1 Operator Description Location Failure Time Action Probability Limit OPA-AFWACT ACTUATE AFW COMPONENTS AFTER FAILURE OF Control 3.60E-03 Varies AUTOMATIC ACTUATION SIGNAL Room OPA-CCWHX CLOSE CCW HEAT EXCHANGER BYPASS VALVE Local 2.70E-03 30 min OPA-CCWLPISO ISOLATE CCW SERVICE LOOP Control 2.00E-02 60 min Room 8
OPA-CCWSERLP SWITCH CCW SERVICE LOOP TO OPPOSITE Control 1.60E-02 5 min TRAIN Room OPA-ESl RCS COOLDOWN AND DEPRESSURIZE - SLO control 2.10I-03 10 hrs Room OPA-HPR PERFORM INJECTION-RECIRC SWITCttOVER Control 3.20E-03 20 min Room OPA-LPI ESTABLISH LOW PRESSURE INJECTION Control 1.00E-04 60 miri FUNCTION Room OPA-LPR PERFORM INJECTION-RECIRC SWITCHOVER Control 3.90E-03 10 min Room OPA-OF2 ESTABLISH BLEED AND FEED COOLING Control 2.60E-03 30 min Room OPA-OPl PERFORM RCS COOLDOWN AND DEPRESSURIZE -
Control 1.20E-02 30 min ES-ll Room OPA-RHR STOP RHR PUMPS DURING A HIGH PRESSURE Control S.00E-04 135 min EVENT Room OPA-SGOVERFL MAINTAIN STEAM GENERATOR LEVEL Control 4.31E-03 30 min Room Question 2.
The submittal discussion regarding seism'c degradation of fire protection equipment is limited to the interaction of piping in safety related areas.
The evaluation should also include an examination of potential loss of fire protection capability itself due to seismic events. Examples of items found in past s t uc* i e s include (but are not limited to):
- Fire protection pumps and tanks CO tanks and bottles
- Sprinkler standoffs (either penetrating suspended ceilings or simply hanging from the ceiling)
- Use of cast iron mains to provide fire water to fire pumps Provios the location of each of these items (if any at WCGS), how they are anchored and/or supported, and whether or not they are seismically qualified.
-,w
Attachment to WM 97-0142 l
- - Page 3 of-14 l
Response
I No f.i re protection system components were included in the ' Wolf Creek seismic. St.f e Shutdown, Equipment hist (SSEL).
No fire protection
[
components are required for operation of any of the front line systems
- l or support systems for the primary or alternate safe shutdown success paths. In accordance with NUREG-1407 and - EPRI NP-6041, systems or components which were not required for operation of a front line system in a success path, or which do not support a front line system in a i
success path, were not included in the IPEEE seismic evaluation.
Accordingly,. an evaluation of the seismic -capability of the-fire protection system components was not included in the reduced scope seismic examination.
L Section 4 of NUREG-1407 states in part, "Some fire issues identified in the Fire Risk Scoping Study (FRSS), -(1) seismic / fire interaction, (2) effects of fire suppressants on safety equipment, and (3) control system
- interaction, should be addressed in the. IPEEE."
The Fire Risk Scoping
- l Study, NUREG /CR-508 8', - identifies (a) seismically induced fires, (b) t f
seismic actuation of fire suppression
- systems, and (c) seismic j
degradation of fire suppression systems as specific items that should be.
addressed with respect ' to seismic / fire interaction. As discussed = in i
3 Section 4.8.1 of the Wolf Creek IPEEE results summary, seismic
- degradation of fire suppression systems" - (as described in the - FRSS) involves seismically induced mechanical failure of suppresolon system components and the potential effects of such failures on the safe i
shutdown capability - (i.e.,
incapacitation of safe shutdown components caused-by seismic-failure and failing of fire suppression components).
This issue does not focus on the " functional survivability" of the suppression system. Accordingly an evaluation of the seismic capability of the fire protection system components was not included in the reduced scope examination.
Section 7.2 of the FIVE' methodology indicates that fire suppression
- systems in close proximity to seismic safe shutdown equipment should be
- evaluated during the seismic walkdown for their " survivability".
The term " survivability" means - the suppression system in question.does not disable safety systems required to shutdown and cool the reactor plant.
It does not correspond to ensuring fire suppression system operability.
3 Section 4.8.2 of the Wolf Creek IPEEE results summary includes a r
description addressing fire suppression system " survivability".
FIVE does not require the issue of fire suppression system operability following a seismic event to be addressed.
Accordingly, an evaluation t
of the functional operability of fire protection system components following a seismic event was not included in the fire risk evaluation.
4 An evaluation of the functional operability of the fire protection
- system components-is considered outside the scope of either.the IPEEE seismic or-fire risk evaluations.
The design requirements for the fire protection system at WCGS do not include the requirement of functional
- operability of the-system following a seismic event.
p Question-3:
EPRI NP-6041 E(Step. 3 of section.2) recommends development of generic anchorage. capacities as part of preparatory work. It states that "it is impossible to make judgments - on the adequacy of seismic ruggedness i
without.an -understanding of the seismic demand corresponding to the HCLPF level and some - measure - of equipment anchorage capacity."
The-submittal. mentione the performance of a " bounding evaluation" for anchorage in only a few places (e.g.,
section 3.5.6).
I
~
f=w7-t'-RFrrt**==="PVwrMf 7
5'r"*W*Ot-*':7 t'
"v
$ 38 PT*t-*'*-W WsreWP,--'g--h'Fg'**gy-6e+--v
,yy n i-erM-W-y*y' yin g, gw y-qww-ycw--
- wg-%-'twiwerewe w--
ww--
- Trismvv y, 4u 7we gw TW-+r
97-0142
- Attachment to WM i
Page 4 of 14
[
- a. Provide an example of a bounding anchorage evaluation, and
~
b..For each equipment category of section 3.5 of the submittal, provide t
the guidance used by-SRT and the procedure it used during the WCGS IPEEE to screen out anchorages.
l Responses l
- a. As. stated in section 3.5.6.1 of the IPEEE results summary, two bounding calculations were performed for two categories of Motor Control-Centers - (MCCs).
The highest elevation design basis in-structure response spectra was used for calculating forces.
Studs were checked for combined shear and tension, and embedded,:hannels were checked for weak axis bending.
Both calculations demonstrated that the MCC anchorage has sufficient margin to accommodate an SME of 0.3g pga.
The bot.nding calculations are either attached to - the screening and evaluation cheets (SEWS) form'or issued as a controlled calculation and noted on the SEWS form.
l
- b. Each category of equipment was screened for anchorage for reduced scope based on one or more of the following:
Detailed-design review
- Anchorage details Specific calculations for CCW Heat exchangers, RHR Pumps, Containment Spray Pumps and D/G Panels Bounding calculations for MCCs, Room Coolers, 125 VDC Panels, 7.5 KVA Inverters-
=
Seismic Review Team's (SRT's) judgment EPRI NP-6041 checklist (Appendix F)
The specific. approach used for each category is identified in section 3.5.6 of the IPEEE results summary *.
8
~
As a part of the detailed design review (per NP-6041, section 2, step 3),_ the following design documents were gathered and reviewed for each I
category of equipment.
- a. Specifications for equipment.
- b. RRU (required response spectra) for design of equipment.
- c. TRS (test response spectra) for design of equipment.
- d. Equipment qualification reports.
- e. Seismic qualification review team (SQRT) files.
- f. Specific anchorage calculations (existing).
- g. Generic anchorage calculations (existing).
h.
0.25g design spectra and 0.2g license spectra.
- 1. EPRI NP-6041 (Table 4-3) recommends 5% damping value while Wolf Creek
[
design -spectra for equipment and piping have used 3% damping (conservative).
i Most components.were screened to 0.3g level even though required-to be screened only to 0.29 level per reduced scope.
The screening was based onathe available margin demonstrated in the above mentioned documents and the experience of'the:SRT.
Components not screened to 0.3g level are specifically mentioned in section 3.5.6 of the submittal.
i Question 4 The evaluation-of USI A-45 states that "all components and structures relating to decay heat ~ removal are adequate for the plant design basis
~
?
'.. +,__-- :_-;-:-.,~.m.--,
5-,.,,
.v
... _ -.. ~. -. -. _.,, _
_x
~
j Attachment to WM 97-0142
?
Page 5 of 14
[
1
'SSE".
Provide an evaluation of USI A-45 for.the 0.3g RLE as requested
. [
in NUREG-1407 including, but not limited to, the basis for screening out
- i
'RWST, pond and service water-system.
[
Response 1-t j-Wolf Creek Nuclear Operating Corporation (WCNOC) notified _. the Nuclear
?
5 Regulatory Commission (NRC) in a letter dated May 20, 1994, that in
- light of the Lawrence Livermore National Laboratory's (LLNL) revised J
- seismic - hazard curves', a reduced scope seismic evaluation would be j
performed for the-Wolf Creek Generating Station (WCGS).
On August 15,
- 1994, the NRC issued a response' to WCNOC's May 20, 19946 letter.
The i
~
NRC's August.15 letter' stated in part, "Until an NRC position is formulated, we cannot approve your revised approach to perform a reduced j
- seismic examination."
WCNOC completed the IPEEE evaluation as per the i
8 Hay 20, 1994* 1etter and submitted the IPEEE results Summary to the NRC f
on June 27, 1995.
Section 3.1, Methodology Selection, of the WCGS 1
IPEEE results Summary
- provided a detailed discussion _ and rationale for the method of analysis.
Subsequent to the completion of the IPEEE study and submittal of the results, the NRC. issued Supplement-50 to GL 80-20 on September 8, 1995.
Reading under the heading
' Requested Information',
" Licensees who
~
previously submitted their requests to modify their seismic IPEEEs may choose not to submit any _ response to this generic letter supplement; j-should that be the case, NRC will respond separately to their previous-i request."
Supplement 5 also states "... licensees may use the revised l
LLNL seismic hazard estimates-instead of_ the 1989 LLNL seismic hazard i
estimates".
WCNOC did not respond to Supplement 5 and no further correspondence on the subject was received from the NRC. WCNOC believes the reduced scope evaluation of WCGS meets the, intent of increased r
understanding of seismic severe accident behavior and identification of 1
seismic severe accident vulnerabilities.
Wolf Creek was evaluated as a reduced scope plant.
Therefore, the l
review-level earthquake (RLE) for the Wolf Creek IPEEE is 0.2g.
However most of the components have been screened to a 0.3g level based on'the
~
available margin in design documents.
d The USI-A45 issue is discussed in the Wolf Creek IPE submittal (section
- 3.4.3)'.
Systems / functions reg' tired based on the IPE discussion are i
included in the IPEEE seismic shutdown paths.
All the cor m onents in these shutdown paths are screened to 0.3g RLE except as noted in section
'3.5.6 of the IPEEE results summary *.
- r 2
results summary ).
2.:The Ultimate Heat Sink (pond) is a seismic category 1 structure, hence not required to be eaaluated per EPRI NP-6041, Table 2-3.
- 3. The service water system (EA) is not part of the preferred or alternate - success path,_ but the Essential Service Water system- (EF) is.
Accordingly,' applicable components-of the Er system are included in the safe shutdown equipment list (SSEL) and are screened to a-0.3g 2
level except as noted in'section 3.5.6 of the IPEEE results Summary.
The buried piping is screened to a 0.29. level (section 3.5. 9. 6 of t
IPEEE results summary),
Question 5:
.I The submitta) notes that. horizontal SME in-structure ' response spectra l
demand exceeds the ~ design. basis spectra demand by as much as 50%.
]
+
4, y
+,.e n.
v.
, - - -, ~ + -,
,.f---
ry,
. ~. -.
m,,,,w_um,.,w,_wn_m-.,_,.,,.,.w..mm,,,-nn-
-,.-.-mm
,...y-
.,,-,.g,-,.,
Attachment to WM 97-0142 Page 6 of 16 Determine and provide seismic capacities for those eg"ipment items that have demand on the order of 40% to 50% higher than the design basis, in order to verify the judgment made during the walk down screening.
- Responses i
l At the end of Section 3.1 of the IPEEE results Summary it is noted that "Where appropriate, this IPEEE summary report discusses. activities undertaken as a part of the
' focus scope' examination process implemented prior to the change in commitment and direction to a
' reduced scope' examination".
Accordingly, this information (Table 3.11) was provided for information only.
Also, see section 3.4 for more i
details about preliminary focus scope examination activities.
For a reduced scope plant, development of a SME in-structure response is not required.
The platat design basis SSE ground spectra and the design 4
basis SSE in-structure response spectra define the SME.
Accordingly, determination of the seismic capacity of these equipment items is not required and was not performed as part of the reduced scope 3
examination activities. All of these components were screened to a 0.3g level based on available margin in design documents, a specific calculation or a bounding calculation except as noted in section 3.5.6 3
of the IPEEE results summary.
Question 6:
The submittal notes that three items could not be assigned capacities more than 0.3g PGA by the SBT.
The submittal also states that if it was not for these indeterminate items, the plant HCLPF would be greater than 0.3g PGA.
Among these items are i
- four 60 cell batteries and racks because of spacing between the batteries and rails twelve LSELS/ESFAS cabinets because they are not bolted together The submittal also asserts that these items could be shown to have a 2
larger seismic capacity than the SSE if detailed seismic margin assessment would be performed.
Consistent with the guidance of NUREG-1407 for focused scope plant and EPRI NP-6041 for performance of an EPRI seismic margin study, determine and provide the seismic capacity of these components.
Response
Section 1.4 of the IPEEE results Summary
- states that " twelve components (four battcry racks and eight cabinets) which are acceptable in terms of the WCGS seismic design basis (0.2g) were not screened against an SME of 0.3g".
Section 3.8 states, "Therefore, the primary and alternate shutdown paths for WCGS (and the overall plant) are assigned a HCLPF capacity equivalent to original plant design basis SSE of 0.2g.
The SRT believes that the HCLPF capacity could be shown to be 0.3g pga with a rigorous evaluation of the two component groups identified above and, perhaps, well in excess of 0.39 pga".
However,- the final-HCPLF assigned to these components was 0.2g pga per reduced scope requirements.
Rigorous seismic capacity evaluations were
- not performed for these components and are not required for reduced
-scope-examination activities.
r
Attachment to t1H 97-0142 Page 7 of 14
(
[
Fire Analysis Questions QuestioL 1:
Section 7.1.2 states that hot shorts are significant for a nunter of fire areas.
Provide the fire areas for which hot shorts are significant.
Describe the method used to account for hot shorts in the PRA of unscreened compartments.
This description should include how F
cabinets and components were reviewed to determine the followings (1) if a hot short could occur, (2) the probability of hot shorts, and.
(3) the way in which the core damage frequency associated with hot shorts was developed.
r Responses For the unscreened fire areas, hot shorts provided significant I
contribution to the Core Damage Frequency (CDF) for fire areas A-17, A-18, C-9 and C-10.
For fire areas A-17 and A-18, hot shorts are significant with regards to a postulated spurious opening of a
pressurizer Power Operated Relief Valve (PORV). The postulated spurious 4
opening of a PORV in these fire areas was aleo assumed to occur concurrent with loss of the power supply to the associated PORV block valve resulting in failure of the block /alve in the open position.
For fire areas C-9 and C-10, spurious operation, due to hot shorts, of a number of Motor-Operated Valves (MOVs) being powered from Motor Control Centers (MCCs) in these areas provided significant contribution to the CDP for postulated fires in several of the cabinets in these areas.
The probabilistic safety assessment (PSA) of unscreened compartments accounted for hot. shorts in the following manner.
In general, hot shorts that might be postulated due to fire damage of cables or cabinets in a fire compartment were assumed to occur.
Affected cable schemes were not subjected to a detailed analysis to determine whether hot shorts could or could not occur.
The only deviation from this general approach concerning hot shorts was an evaluation of the impact of hot shorts on the power and control cables for MOVs running from the MCC cubicle providing power for the MOV to the MOV itself.
The control cable running from the MCC cubicle to the MOV actuator is normally energized.
A hot short condition providing power to this normally energired cable would not result in spurious operation of the associated valve.
The power supply to the MOV actuator is three phase 480 Volt AC.
A single conductor hot short condition associated with the MOV power supply cable would not result in spurious valve operation.
In order to achieve spurious valve operation due to fire damage of the MOV power supply cable, simultaneous three phase hot shorts in the proper phase
. rotation would be required.
Spurious valve operation due to occurrence of this type of hot short condition was not considered credible.
Accordingly, fire damage to the control and power cables routed from an MOV actuator to the associated MCC cubicle was not postulated-to result in hot short conditions which would spuriously reposition the MOV.
If the only cables in a fire area which are associated with a given MOV are these control and power cables, then-spurious operation of that MOV was not included in the determination of CDF for the fire area.
Where hot shorts were postulated to occur due to a cabinet or compartment fire, they were assumed to occur with a probability-of 1.0.
The only exception to this was'the consideration of extended hot short conditions for specific fire scenarios where a pressurizer PORV was postulated to fail to the - open positicn due to a hot short condition
Attachment to WM 97-0142 Page 8 of 14 with the control and/or power cables for the associated PORV block
.ve also impacted by the same fire such that the block valve would be failed in the as is (open) position.
For these specific fire scenarios, the probability of an extended hot short condition was considered.
As 3
described in Section 4.3.5 of the Wolf Creek IPEEE rasults Summary,
NUREG/CR-2258" provides a recommendation that an extended hot short probability of 0.068 be applied.
While the hot short may initially result in spurious component operation, it was considered that an open circuit condition was more likely with time.
NUREG/CR-2258" suggests using an upper bound value of 0.2 for hot shorts that need only last five minutes to realize system or component failure, and a lower bound
-value of 0.01 for hot shorts that need to last for 30 minutes.
In the determination of the CDF for these fire scenarios, a probab311ty of 0.07 was used for failure of the open PORV to close.
This extended hot short probability value was applied for scenarios in fire areas A-17, A-18 and for selected fire scenarios in fire areas A-16 and A-80.
The _ CDF due to a fire in an area or in a specific scenario was determined by multiplying the fire initiating event frequency for that area or scenario by the Conditional Core Damage Probability (CCDP) resulting from the postulated fire.
The CCDP included consideration for hot shorts by the assignment of a logical failure to the failure modes for the component in question, including those failure modes which would be considered to result from a hot short condition.
The only exception to assignment of a logical failure for component failure modes was the assignment, for specific fire scenarios, of a value of 0.07 for the probability of an open FDRV to reclose as described above.
Question 2:
The treatment of propagation of fire from cabinets was inconsistent.
Propagation of fire from cabinets was assumed to not occur in the control room, whereas it could occur in all other cabinets of the plant.
It is noted that the study distinguished among open cabinets, sealed cabinets in high traffic areas and sealed cabinets in low traffic areas.
The justification for the propagation o' fire from sealed cabinets of 0.69 is clearly. stated.
However, the derivation of 0.15, as the propagation probability from sealed cabinets in high traffic areas, is not clear.
Describe the testing, inspection and/or surveillance program in place at WCGS that would ensure that cabinet seals are (a) always in place in the cabinets for which credit was taken in the IPEEE, and (b) effective in preventing fire propagation.
Provide a derivation with explanation of the probability of 0.15 for sealed cabinets in _ high. traf fic areas.
Provide a derivation with explanation for the assumptions that fires will not propagate from cabinets in the control room.
Response
l LAs indicated in Section 4.0.2 of the Wolf Creek IPEEE results Summary,
the top penetrations for a number of electrical cabinets at WCGS have an installed vapor and dust barrier seal. Wolf Creek Drawings M-lYOO6A, M-
-lYOO68 and M-lY006C" list the equipment at WCGS which requires the installation: of vapor and ' dust seals.
While WCGS does not have a separate and specific program for the inspection and maintenance of these seuls, preventive maintenance inspections are performed on the cabinets using approved Preventive Maintenance (PM). procedures.
The purpose of these procedures is to clean, inspect and test the cabinets.
If; repairs are needed on - the cabinets (including the vapor and dust seals), an Action Request is-generated per procedure AP 16c-001, " Action
Attachment to WM 97-0142 Page 9 of 14 Requesti"" to accomplish the necessary work.
Repair of the existing vapor and dust seal, or installation of a new seal,-is accomplished in accordance with prvcedure CNT-908, " Installation and Repair of Vapor and Dust Seals"".
'the inspection accomplished during performance of the PM procedures, along with the procedural guidance to affect any necessary real repairs, provides reasonable assurance that the installed vapor and dusi seals are in place, intact and in good condition.
The existence and satisfactory condition of these seals was verified during the IPEEE fire risk evaluation walkdowns for the appropriate electrical cabinets located in non-screened areas.
These penetration seals are composed of the same type of fire resistant materials as utilized for qualified fire
- rated penetration seals (Dow Corning 3-6548 silicone RTV foam or Dow Corning Sylgard 170 silicone elastomer).
The seal supplier has indicated that tests have been performed on this seal material and fire ratings-have been established-for installed seal thicknesses of.
approximately 3 inL. or greater.
The vapor and dust penetration seals r
at WCGS are typically in the 1 to 2 inch thickness range.
The installed vapor and-dust penetration seal configurations have not been subjected to a qualification testing program and are not considered a fire rated seal.
While it is acknowledged that the subject vapor and dust penetration seals are not-qualified fire rated seals, it was judged that the seals would provide an erfective barrier to fire propagation from-electrical cabinets.
Based on a review of the vapor and dust penetration seal installation drawings, inspections of the seals performed during the fire risk evsluation walkdowns and the fire resistant nature of the materials utilized in the construction of the seals, it was judged that assignment of'a probability of 1.0 for fire propagation from these cabinets would provide unrealistic and overly conservative resul?s.
Accordingly, it was decided that a probabilistic approach would be utilized for the determination of the likelihood for propagation of fire from the electrical cabinets with these seals to the cables which might be above i
the cabinet.
The probabilities of propagation for cabinet fires was based on electrical cabinet fire data from NSAC/178L".
NSAC/178L provided data on a total of 84 electrical _ cabinet fires.
Of these 84 electrical cabinet
- fires, 26 were determined to self extinguish.
Accordingly, a propagation probability of 0.69 was determined for a sealed cabinet-in an area not well traveled and without suppression as 1.0 minus the self extinguish probability of 0.31 (26/84).
For sealed electrical cabinets in f requently travaled arean, it was assumed that the propagation probability would be reduced as the more frequent presence of plant personnel in the area would increase the likelihood of local detection and manual suppression of a cabinet fire.
The NSAC/178D' data indicated that 35 of the 84 electrical cabinet fires were manually suppressed with - a known method, and 20 out of 84 were manually suppressed with an unknown method.
For sealed electrical e
cabinet fires in frequently traveled areas, it was assumed that the data on manual suppression of the cabinet fire would apply.
In this case,
- all of the known manual suppression methodology data and one half of the unknown- -suppression methodology data were assumed applicable.-
Accordingly, a propagation probability of 0.15 was determined for a sealed cabinet in a well traveled area and without automatic suppression as 1.0 minus the sum of the_self extinguish probability plus the manual supp:ession with known method probability plus one half of the manual suppression with unknown method probability (1 --((26 + 35 + 10)/84)).
Based on the historical : evidence and -the method used to seal cabinet cable..- penet rations at Wolf Creek, it is 1 considered that the approach-outlined above_for the determination of electrical cabinet. fire propagation probabilities is reasonable and generally consistent with that usod by others in the industry.
Indeed, the approach utilized at
' At,tachment to WM 97-0142
' ' Paga 10 of 14 l
i Wolf Creek is conservative when compared with that recommended in the EPR1 Fire PRA ' Implementation Guide Appendix E of the EPRI Guide" 3
Indir:ates that non-vented - electrical cabinets- (with the exception of 3
- those containing high. energy equipment). do not propagate fire in the (probability of propagation is-zero). The position is taken *
+
r absence of other ventilation, even: small penetrations will allow sufficient air exchange to replace oxygen being consumed by tne fire, and an incipient fire will self extinguish when there is no longer t-enough oxygen to support combustion.
For high energy cabinets (2 480 i
VAC),
fire propagation via non-rated fire penetration seals is considered possible.
However, the conditional probability of a severe -
i fire originating in such cabinets (sufficient to result in propagation) was determined to be 0.12 based on the historical evidence of fires in switchgear room electrical cabinets (Appendix D of the EPRI Guide").
i rinally,-it is noted that electrical cabinets at Wolf Creek contain IEEE contained 4
383 rated cables.. The fire f requency and prop,a,gation data within the EPR1 Fire Events Database (NSAC/178L) includes = events from older plants which une non-rated cables.
The direct application of this data to Wolf; Creek introduces additional conservatism to the approach used.
For control room cabinets it was assumed that'a cabinet fire would not spread from the confines of the cabinet in which it originated because the cabinets have solid metal or fire resistant boundaries.
This is supported by the results of the Sandia cabinet
~ fire: tests (NUREG/CR-4 $27") in which all test fires self extinguished and also by i
the report on cabinet fires in the NSAC/178L" data base.
d Question 3:
The general turbine area and the two radiation access areas were screened out by-including the automatic suppression system i
unavailability as a multiplication factor on the fire ignition frequency and conditional core damage probability.
FIVE allows this method-of I
screening if it can be demonstrated that the suopression system is code compliant, and the time of extinguishment i less than the time of i
damage.
For those areas, provide justification that the suppression systems are installed in a manner that complies with all applicable NFPA standards, 1
and that the suppression systems would be effective to prevent damage to
. cables.
Include in the response the damage criteria used, the time to 3
damage ~ cables or equipment, and the time to extinguish the fire.
+
Response:-
The._ Wolf Creek _ Updated Safety Analysis Report (USAR) Appendix
- 9. 5 A",
which_-delineates WCGS compliance to Regulatory Guide 1.120,
" Fire Protection Guidelines for~ Nuclear' Power Plants," states, "the automatic water extinguishing systems are: designed, constructed, and tested based on NFPA 13-1975 and 15-1973, as_ applicable".. This is in agreement _with paragraph
- 3. 4.2 of the system description for the power block fire protection system (KC) which states,'" Tests and inspections should be in accordance1with NFPA Standard 13A".
Alsoi the purchase specifications for-the automatic water extinguishing systems (10466-M-650, -M-650A and
-M-654) require -conformance to applicable portions of NFPA 13-1975 and 2 NFPA 1973. in the design, materials, manufacture, installation,
-testing, inspection, stamping, certification and documentation.
l
[
i m
a a
e a=
~'s:ww
'p wyy1,=i-n-Qe g-=-,e+-
W yyaim.-g-
m+N9-ym-g wegg g y p-p y-prn--
gn,y wm erWvwg wrW r'm v-w'mps rw==*rwg*
n'9-ginumW-r-r-T prw fyget weg up' ti g'
p="
sing P- *r
-'s fr7 F TT*W 1r W
w' r
At,t a c hme nt to WM 97-0142 Page 11 of 14 For the Access Control Fire Areas, Areas C-5 and C-6, all ignition sources and combustibles are located below a nonc ombtwtible suspended ceil,ing protected by a wet pipe fire suppression system.
All Safe Shutdown Equipment (SSE) cabling is located above the noncombustible suspend 3d ceiling in either cable trays or conduit.
Fire detection and suppression are provided both below the suspended ceiling and above the SSE cabling located above the suspended ceiling.
Since fire protection is located between all ignition sources and the SSE cabling, operation of the automatic fire suppression system was credited.
All of the SSE cabling is protected by the automatic suppression system.
While manual fire suppression was not credited, manual suppression of most fires is considered highly likely since these Access Control fire areas are occupied 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> a day.
The 'iurbine Building general area, Area T-2A, does not include any SSE conponents.
This area does include i number of non-SSE PSA affecting components and cabling, the failure of which is considered to result in a Loss of Main Feedwater transient (TRO), except for power feed cable from the switchyard to one of the ESF transformers.
The ESF transformer power cable enters the Turbine Building in the far northwest corner, is routed through the PA02 electrical cabinet and then runs through a cable tray along the west Turbine Building wall before exiting to the ESF transformer.
Walkdowns of the Turbine Building general area were performed to determine the types of ignition sources present and their separation distances relative to the ESF transformer power cable.
Based on cable damage distances generated from the generic COMPBRN runs, it was determined that damage to the subject cable would not be sustained in the absence of propagation from primary fixed ignition sources to adjacent combustibles (i.e.,
the heat release from the primary ignition sources, such as electrical cabinets, pumps, etc. are not capable of causing damage to the subject cable directly).
The only fixed ignition source which would directly damage the ESF transformer power cable is cabinet PA02, through which the subject cable is routed.
The large open construction of the Turbine Building general area, with floor to ceiling openings and considerable floor gratings, precludes hot gas layer buildup.
Given that the entire area around the target cable and the fire sources in the area are covered by an automatic suppression system, which is designed to the appropriate NFPA standards, the response time of the suppression system was considered adequace to protect the ESF transformer power cable and no damage was postulated unless the suppression system failed to operate.
Thus, the only sources which could cause direct damage to tha subject ESP cable are transient sources and electrical panel PA02, through which the ESF cable is routed.
Automatic fire suppression was not credited for protecting the cabb against fires associated with these sources.
The Conditional Core Nmage Probability (CCDP) values utilized to screen the Access Control Areas and Turbine Building general area followed the guidance provided in Step 4.2,
" Preliminary Quantitative Screening" in the EPRI Fire PRA Implementation Guide", which indicates, for a fire area with a full-zone automatic suppression system, the fire-induced core damage frequency (CDF) is quantified with the following two terms:
Conditional Core Damage Probability (CCDP), with all equipment in the fire area damaged, multiplied by the probability of failure of the full-zone automatic suppression system, plus
- CCDP, with only the equipment not protected by the full-zone suppression system damaged, multiplied by the successful (as designed) operation of the automatic suppression system.
. Attachment to WM 97-0142 Page 12 of 14 2
The CDF values for the Access Control Areas (C-5 and C-6), as presented l
in Tables 4.14 and 4.15 of the Wolf Creek IPEEE results Summary, were quantified in accordance with the first CCDP term above.
The second-a CCDP term was not included in the CDF' determination for these areas as all equipment was protected by the full-zone suppression system.
The CDF for the Turbine Building general area due to transient ignation sources is 1.17E-08 as presented in Table 4.14 of the Wolf Creek 1F?.EE 3
results Summary.
Since transient ignition. sources could be locatd
. such - that direct damage to the ESP transformer power cable would - be
_ postulated to occur, the CDP due to transient ignition sources did not credit automatic fire suppression system operation.
'The CDF for the Turbine Building general area due to fixed ignition sources, including credit f or automatic fire suppression operation, is 2.91E-07 as presented in Table 4.15 of the Wolf Creek -IPEEE results Summary.
This CDF-value does not, however, account for the direct 2
damage to the ESF cransformer power cable postulated from a PA02 cabinet fire in accordance with the second term from Step 4.2 of the EPRI Fire PRA Implementation Guide".
The CDF contribution due to a fire in PA02 was calculated as follows.
The ignition source frequency for fixed ignition source PA02 is determined based on the generic ignition frequency for electrical 3
cabinets in the Turbine Building, from FIVE,- Reference Table 1.2, divided by the number of electrical cabinets in the area"
[1.3E-02/103).
The Turbine Building general area CCDP of 3.06E-04 (IPEEE 8
results Summary, - Table 4.15) is used for the tailure of cabinet PA02 and the ESF transformer power cable, since the cverall CCDP is dominated by failure of the transformer power cable.
The CDF contribution due to a PA02 cabinet fire would bei
( (1. 30E-02 /103) ( 3. 06E-04 ) ]
Freq,m
=
3.86E-08 The total CDP due to a Turbine Building.- get'eral area fire would z
therefore be 3.41E-07 (1.17E-08 + 2.91E-07 + 3.86E-00] which is less than the 1.0E-06 screening velue.
With regards to the fire damage criteria used to ascertain the potential for direct damage to the. ESF transformer power cable from surrounding sources, these are provided in Section 4.3.2.2 of the Wolf Creek IPEEE 8
results Summary,
A cable damage temperature of 523 K or a damage flux of 5700 W/H2 were used, which are quite conservative given that cables at Wolf Creek are IEEE 383 rated.
These damage criteria were used.in the performance of the generic COMPBRN runs.
The results of the generic COMPBRN runs wore used during the Turbine Building general area walkdown as.a basis for determination whether a fire associate - with a fixed ignition source would directly damage the subject ESP transformer power cable.
Given the established separation of the fire sources from the targets in both the Turbine Building general area and the two Access control areas, immediate exposure of the target cables ' to eny fire source (except transients and cabinet PA02 in the Turbine Builcing) would be prevented.
The fire suppress. ion system actuation devices would, however, be exposed almost.immediately to the subject fire and a demand for actuation of the suppression ' systems prior. to damage of the cables was considered a certainty, No actual modeling cf the suppression system response time
- -. - _ ~.. - - -
?
At.tachment to WM 97-0142 Page 13 of 14
{
with respect to the time to damage was required or performed in these' l
cases.
j i
Question 4 l
l A review of the submittal with respect to the PRA of unscreened J
compartments reveals that it was generally assumed that fires that propagate out of cabinets would damage cables and equipment within 10 to 20 feet of the cabinet.
It was also assumed that halon suppression l
would be effective unless a system failure occurrod in preventing damage to equipment outside of this radius.
This implicitly assumes that the i
systems are designed, installed and maintained in a code compliant l
manner.
Provide justification that suppression systems, taken credit for in the PRA-of unscreened compartments, are designed, installed and maintained in accordance with appropriate = industry standards, such as-those published by the NFPA.
kesponses i
USAR Appendix
- 9. 5A",
which delineates WCGS compliance to Regulatory Guide 1.120,
" Fi re Protection Guidelines - for Nuclear Power Plants,"
l states "Halon extinguishing systems are based on NFPA 12A-1973".
It states further *...the Halon systems are maintained and tested based on NFPA 12A-1973".
These statements are in agreement with paragraph 3.4.4 l
I of the system description for the power block fire protection system (KC) which says " Test and inspection procedures'should be in accordance i
1 h NFPA Standard 12A".
NUREG 0881, Supplement 5",
discusses a modification of the Halon system in the control room whereby an 4
t automatic time initiation of the backup Halon storage bank is used to increase the Halon coak time in the control room cable trenches to about i
7 minutes (versus the recommended 10 minutes).
The staf f has reviewed 1
the concentration, soak times and the design criteria for the Halon fire suppression systems and has concluded that even with this deviation, the gaseous fire suppression systems meet the guidelines.
Finally, the purchase specification for Halon extinguishing systems at WCGS (16577-M-658) requires conformance to NFPA 12A in the design, materials, manufacture, installation, testing, inspection, stamping, certification 3
i and documentation.
1 Letter ET 95-0055, dated June 27, 1995 from R. C. Hagan to NRC,
" Final Response to Generic Letter 88-20, Supplement 4" NUREG/CR-5088, " Fire Risk Scoping Study," NRC, January 1989 8
TR-100370, Fire-Induced Vulnerability Evaluation (FIVE), Electric t
Power Research Institute, Final Report, April 1992 NP-6041-SL, "A_ Methodology for Assessment of Nuclear Power Plant Seismic M
'in," Revision 1, EPRI, August 1991 l
Letter WM 94-0081, dated May 20, 1994, from N. S. Carns to NRC,
" Revision-to NRC Commitment Made in WCNOC's Response to Generic
-Letter 80-20, Supplement 4" NUREG-1488, " Revised Livermore Seismic Hazard Estimates for 69 Nuclear Power Plant Sites. East of the Rocky Mountains," NRC, April 1994 t
Letter dated August 15, 1994, from USNRC to N. S. Carns, Wolf Creek Generating Station - Individual Plant Examination of
- External Events -(IPEEE)
(TAC No, M83696) l
.NRC Generic Letter 88-20, Supplement 5, Individual Plant Examination'of External Events for Severe Accident 4
Vulnerabilities, September 8, 1995 i
?
z-7
. s,
_.___..._,-_.-----.J,_,-...,,_..A._,
.,.,. ~
-.----. - - -f At,tachment to WM 97-0142
_c Page 14 of 14 f
Letter WM 92-0152, dated Septenber 28, 1992, from B. D. Withers to j
.. NRC, " Wolf Creek Generating Station Individual Plant Examination" NUREG/CR-2258,
" Fire Risk Analysis for Nuclear Power Plants,"
- t
' September 1981 i
WCGS Drawings H-1Y006A, H-1YOO6B and M-1YOO6C, Electrical Equipment Requiring Vapor & Dust Seals"
}'
" Action Request" WCGS Procedure CNT-908, Installation and Repair of Vapor and Dust Seals" NSAC/178L,
" Fire Events Database for U.S. Nuclear Power Plants."
EPRI-105928, "EPRI Fire PRA-Implementation Guide." December 1995 NUREG/CR-4527 "An Experimental Investigation of !!.ternally Ignited Fires in Nuclear Power Plant Control Cabinets, Volumes 1 and 2,"
April 1987
_ Wolf Creek Updated Safety Analysis Report (USAR), Appendix 9.5A.
WCGS' Calculation Number AN-94-041, Rev. O, WCGS IPEEE Project,
!PEEE Ignition Source Frequencies NUREG 0881, Supplement 5,
" Safety Evaluation Report Related to the Operation of Wolf Creek Generating Station Unit #1" t
3 r
I i
. k 4
i P
^
g, w,
i4,y s.-g
,9 q.+g-eg gg--.
y ypve
- --ge,y..,.,p,-y,,-pygw+gr--p.--p9.-e-w.emmm-a-rp--+,2-e g.g gpg--g+-p-.g--p+,gq qwe,-$g%.'4.m,-g