ML20216D872

Review of Submittal in Response to USNRC GL 88-20,Suppl 4: 'Ipeees,' Fire Submittal Screening Review Technical Evaluation Rept:Hope Creek Rev 1:980518

ML20216D872
Person / Time
Site:	Hope Creek
Issue date:	07/26/1999
From:	Lachance J, Mitchell D, Pepping R SANDIA NATIONAL LABORATORIES, SCIENCE APPLICATIONS INTERNATIONAL CORP. (FORMERLY
To:
Shared Package
ML20216D819	List: ML20216D833 ML20216D845 ML20216D872 ML20216D890
References
CON-FIN-W-6733 GL-88-20, NUDOCS 9907300171
Download: ML20216D872 (24)

Text

7 , ,,

Review of the Submittalin Response to U.S. NRC Generic Letter 88-20,' Supplement 4:

" Individual Plant Examination-External Events" Fire Submittal Screening Review Technical Evaluation Report: Ilope Creek Revision 1: May 18,1998 Prepared by:

JefTrey L. LaChance Science Applications International Corporation Albuquerque, New Mexico 87106 I

Richard E. Pepping I Donald B. Mitchell Accident and Consequence Analysis Depanment Sandia National Laboratories Albuquerque, New Mexico 87185-0748 Prepared for:

Probabilistic Risk Assessment Branch Division of Systems Technology Office of Nuclear Regulatory Research U. S. Nuclear Regulatory Commission Washington, DC 20555 USNRC JCN W6733 I

l 4

9907300171 990726 4 PDR ADOCK O P

a i l

1.0 INTRODUCTION

This Technical Evaluation Report presents the results of the Step 0 review of the liope Creek Generating Station (IICGS) fire assessment reported in the "Ilope Creek Generating Station Individual Plant Examination for External Events" (IPEEE) [1].

1.1 Plant Description llope Creek is a General Electric (GE) boiling water reactor (BWR) with a design thermal output j rating of 3293 Mwt The nuclear steam supply system (NSSS)is similar to other GE BWR/4 plants.

The reactor core isolation cooling (RCIC) and an emergency core cooling systems (ECCS) are also typical of BWR/4 designs. The ECCS includes a high-pressure coolant injection (liPCI) system (provides flow through both the feedwater system and the core spray spargers), an automatic depressurization system, a multi-loop core spray system, and a low-pressure coolant injection system j (LPCI) which is a mode of operation of the residual heat removal (RHR) system. 11 CGS is equipped with a 12" hard pipe containment vent which can be operated either remotely or iocally A 6" hard 4 pipe vent can also be used for containment venting.

Critical support systems include four trains of emergency power each connected to a diesel generator, four divisions of emergency DC power, a reactor auxiliary cooling system (RACS), a turbine auxiliary cooling system (TACS) and a safety auxiliary cooling system (SACS) which are served by the same pumps and heat exchangers, and a station sersice water system (SSWS) that cools TACS, SACS, and RACS. Emergency loads such as the diesel generators, ECCS pumps and room coolers, and the RHR heat exchangers are cooled by SACS. The TACS is isolated from the SACS upon receipt of a LOCA signal to ensure the safety-related equipment is cooled. The SSWS uses the Delaware River '

as the ultimate heat sink and can be aligned for injection into the vessel through LPCI train B. In addition, both the diesel-driven and motor-driven fire pumps can provide flow to the reactor vessel through LPCI train B.

The major structures at HCGS are the primary containment; the reactor, turbine, auxiliary (also l referred to as the control / diesel building), and service / radioactive waste buildings; the service water l intake stmeture; and various support structures. The liope Creek NSSS is housed within a Mark I i

primary containment. The reactor building contains the reactor vessel, the primary containment, SACS, RACS, and the ECCS. The control / diesel building contains the control room, remote shutdown panel, cable spreading room, critical HVAC systems, DC batteries and switchgear, the diesel generators, and emergency switchgear. The senice water intake structure contains the SSWS pumps. The turbine building houses the turbine generator, auxiliary systems including the air systems, ,

and non-safety switchgear.

i 1.2 Review Objectives l The performance of an IPEEE was requested of all commercial U.S. nuclear power plants by the U.S.

Nuclear Regulatory Commission (USNRC)in Supplement 4 of Generic Letter 88-20 [2] Additional

' 's ,

guidance on the intent and scope of the JPEEE process was provided in NUREG-1407 [3] The objective of this Step 0 screening review is to help the USNRC determine if the llope Creek submittal has met the intent of the generic letter and to also determine the extent to which the fire assessment addresses certain other specific issues and ongoing programs--

1.3 Scope and Limitations The Step 0 review was limited to the material presented in the Ifope Creek IPEEE submittal.

Furthermore, the review was limited to verifying that the critical elements of an acceptable fire analysis have been presented. An in-depth evaluation of the various inputs, assumptions, and calculations was not performed. The review was performed according to the guidance presented in Reference 4. The results of the review against the guidance in this document are presented in Section 2.0. Conclusions and a recommendation as to the adequacy of the 11 ope Creek IPEEE submittal with regard to the fire assessment and its use in supporting the resolution of other issues are presented in Secti.on 3.0.

1 l

l

(

3

}

[

2.0 FIRE ASSESSMENT EVALUATION The following subsections provide the results of the review of the Hope Creek fire assessment. The review compares the fire assessment against the requirements for performing the IPEEE and its use in addressing other issues. Both areas of potential weakness and strength of the fire assessment are highlighted.

2.1 Compliance with NRC IPEEE Guidelines The USNRC guideline., for performance of the IPEEE fire analysis derive from two major documents.

The first is NUREG-1407, and the second is Supplement 4 to USNRC Generic Letter 88-20. In the current screening assessments, the adequacy of the Hope Creek fire assessment in comparison to these guidelines has been made as outlined in " Guidance for the Performance of Screening Review of Submittals in Response to U. S. NRC Generic Letter 88-20, Supplement 4: ' Individual Plant Examinations - External Events,"' [4]. The following sections discuss the utility document in the context of the specific review objectives set forth in this screening review guidance document and assesses the extent to which the utility submittal has achieved the stated objectives.

2.1.1 Methodology Documentation The Hope Creek fire assessment utilizes the Fire-induced Vulnerability Evaluation (FIVE) methodology [5], which was approved for use in NUREG-1407, for screening fire areas and fire probabilistic risk assessment (PRA) techniques for performing detailed analyses of unscreened areas.

The screening FIVE assessment includes a Fire Compartment Interaction Analysis (FCIA) and a quantitative screening which assumes a fire destroys all equipment in a fire area. The quantitative l .

screening step used the HCGS Individual Plant Examination (IPE) models to determine a screening l conditional core damage probability (CCDP) for each fire area. The unscreened fire areas were subjected to a detailed probabilistic assessment of each possible fire initiator / target combination which modeled intermediate fire growth stages. The detailed analysis involved fire damage calculations using a modified version of the FIVE fite propagation methodology and evaluation of the CCDP for each identified scenario using the IPE models and FIVE fire data. Walkdowns were also performed to verify information and obtain information needed for the detailed fire analysis. Some key l assumptions which were utilized in the fire assessment and their validity are addressed below. j Additional assumptions are addressed in the remainder of this report. l A fire in each fire compartment was assumed to result in a reactor trip. Thus, no compartments were qualitatively screened in the FCIA. This is viewed as a positive deviation from the FIVE j y methodology.

. - The Hope Creek licensee credits the systems modeled in the IPE in both the quantitative screening portion of the FIVE assessment and the detailed probabilistic assessment. However, in both assessments; feedwater, condensate, and the control rod drive (CRD) systems were

- assumed to be unavailable. Inclusion of non-safety related mitigating systems in the analysis is  ;

viewed as a positive aspect of the analysis.

i - The submittalindicates that hot shons were included in the analysis. This included evaluation of spurious component operation resulting in LOCAs. The probability of hot shorts was assumed to occur with a probability of 03 given a fire that threatens a cable. Inclusion of spurious component operation due to hot shorts is viewed as a strength of this analysis. liowever, only hot short impacts on the type of accident initiating event were identified. Hot short impacts on mitigating system operation was not addressed.

. The fire assessment does not explicitly credit either automatic or manual fire suppression systems due to the short damage times predicted by the FIVE fire modeling methodology used in the study. Ignoring fire suppression probably results in pessimistic estimates of core damage but does not address the issue of collateral damage from suppression efforts. The potential for collateral damage from the suppression systems and the fire brigade were not qualitatively (or quantitatively) addressed in the submittal.

. The licensee performed a "high hazard area assessment" to determine the potential that an extremely large fire would breach a fire barrier credited in the FIVE methodology for preventing fire growth. The frequency of an uncontrolled fire in areas containing large amounts of combustible materials was estimated.- The frequency estimates include the frequency of fires involving the combustibles and failure of suppression efforts which is required to lead to a large fire. This frequency assessment is viewed as a positive aspect of the analysis but one which already occurs through tne screening and detailed fire assessments.

i l_

e - The FIVE methodology for fire growth and propagation was modified by the licensee to provide more realistic results. The enhancements include conserving the total energy in the fuel, includiny detector delay times, including delays in suppression system actuation, and considering fue:

depletion. All of these modifications are viewed as positive enhancements. However, the modified models have not been verified.

In conclusion, the Hope Creek IPEEE submittal contains a summary description of the methodology used in the fire assessment and a summary presentation of the actual analysis. Important assumptions used in the analysis are identified. The impact of these assumptions are addressed below.

l 2.1.2 Plant Walkdown The Hope Creek fire assessment began with the division of the plant into discrete areas for analysis.

Fire areas are defined as locations completely enclosed by at least two hour rated fire barriers. The licensee states that the fire area boundaries are identical to those identified in the HCGS Updated Final Safety Analysis Report (UFSAR). In most cases, these fire areas were subdivided into compartments with the compartment boundaries meeting the FCIA criteria in FIVE. Although it is not explicitly stated in the submittal that all areas of the plant were reviewed in the fire assessment,  !

a list of the fire areas provided in the submittal appears to include all major buildings in the plant. )

. e ,

Information sources used in the Hope Creek fire assessment include drawings, the UFSAR, procedures, pre-fire plans, and the HCGS IPE In addition, the locations of cables were obtained via a computerized cable database.

The llope Creek submittal provides a comprehensive discussion on the plant walkdowns that were perfonned for the fire assessment. The submittal indicates that the goals of the walkdowns were (1) to confirm documentation used represents the as-built plant, (2) uncover potential intercompartment interactions associated with openings, walls or inadequate fire barriers, (3) aid in addressing the Fire Risk Scoping Study (FRSS) issues, (4) assess the likelihood of critical transient combustible loading, (5) review fire protection features of the plant, and (6) verify assumptions used in the fire damage propagation analyses. In addition, the licensee indicates that the walkdowns were used to develop the fire scenarios quantified in the fire PRA.

The walkdown team was composed of a fire protection engineer and a PRA analyst from HCGS and two fire PRA contractors. The submittal indicates that the fire PRA contractors have participated in several fire PRAs and walkdowns. The licensee used three sets of walkdown checklists to document the fire hazards, equipment locations, fire protection systems, and fire barriers in each room. A review of examples of these checklists provided in the submittal indicates they address critical information needed in a fire assessment that can only be obtained by a walkdown. Written guidance on how to use the checklists was also developed and is provided in the submittal. The submittal indicates that for high radiation and inaccessible areas, a " virtual" walkdown was performed by reviewing digitized photographs of those areas.

The submittal presents some general observations, findings, and conclusions from the walkdowns.

One finding was that openings in fire barriers exist in some rooms. The licensee indicates that several rooms were combined as a result of this finding. Observations for specific rooms were also presented in the submittal. One significant observation is that the offsite power bus bars from both redundant sources run along the ceilings of the diesel generator rooms in conclusion, the walkdown description provided in the submittal covers the purpose, content, and results of the walkdown. The walkdowns cover the main requirements for acceptance, including the collection of data as well as the confirmation of data and assumptions.

2.1.3 Qualitative Fire Area Screening The first level of screening allowed in the FIVE methodology is qualitative in nature and eliminates fire areas in which a fire does not affect any Appendix R equipment and does not result in a reactor trip. In the Hope Creek assessment, a fire in any compartment was assumed to cause a plant trip.

Thus, no fire areas were qualitatively screened.

The llCGS licensee used the FIVE method for defining fire compartment boundaries. That is, an FCIA was perfonned to establish the combination of rooms that have boundaries that meet the FCIA

r . .  !

I screening criteria in the FIVE methodelagy. A total of 209 fire compartments were defmed. The licensee used>/cmplate to document the FCI A. An example of the template is provided in the submittal. As noted earlier, the results of the walkdown were used in some cases to redefine compartments This was done when openings such as unused penetrations, unsealed penetrations, and cable tray nms werejudged to potentially lead to the spreading of hot gases. Note that liVAC return ducts were not included in the list of openings since the licensee concludes, from limited tracing of the ducts, that smoke and hot gas propagation veill not occur through the ducts The FIVE methodology provides a second level of qualitative screening which is performed at the fire compartment level. The FIVE methodology allows a fire compartment to be elimina.ed if all the companment boundaries screen according to the FCIA criteria and if the compartment either contains e no Appendix R equipment or does not result in a plant shutdown due to a fire. The licensee did not qualitatively screen any fire compartments that meet the FCIA criteria as allowed in the FIVE methodology.

In summary, no qualitative screening was performed in the llCGS fire assessment This is viewed as a positive enhancement of the FIVE screening methodology. An FCI A was performed according to the FIVE methodology with the fire compartment boundaries modified to reflect the results of the walkdowns particularly with the respect for hot gas propagation.

2.1.4 Fire Occurrence Frequency In the next phase of the llope Creek fire assessment, quantitative screening was performed based on the 1.0E-6/yr core damage criteria allowed in the FIVE methodology. The first step in the quantitative screening was to develop fire occurrence frequencies for each fire compartment. The submittal indicates that the fire frequencies were generated using the FIVE methodology and data.

This was accomplished by identifying all potential fire sources and by completing Fire Compartment Ignition Source Data Sheets (ISDS) for each companment. The equipment in each compartment that can initiate a fire was identified through the use of the 11 CGS Managed Maintenance Information System, pre-fire plans, UFSAR, and supplemented by plant walkdowns. The resulting fire frequencies were referred to by the licensee as the screening fire frequencies for the fire compartments.

The submittal indicates that the screening fire frequencies used in the study included the contribution from transient ignition sources. The screening transient fire frequency contributions were calculated according to the FIVE methodology. The submittal indicates that the contribution from cigarette smoking and hot pipes were eliminated from these screening transient fire frequencies because the liCGS fire protection program prohibits cigarette smoking and requires all hot work to be accompanied by a fire watch.

The fire frequencies used in the detailed assessment of each source / target combination in the i unscreened fire companments were obtained from the data on the ISDS for each compartment. The l transient fire frequencies were calculated using the FIVE methodology which utilizes the fraction of

~

l the floor area over which a transient fire can damage cables. The calculation of the transient fire 7-n - _ _ _ _ _ _ _ _ _ _ _ _

• . l k ,

1 i i

frequencies was abo based on a daily inspection interval and does not include the probability that the combustible is exposed as allowed in the FIVE methodology since the licensee correctly states that this factor is already included in the generic fire frequency.

1 The licensee claims in the submittal that the llCGS plant-specific fire data is too sparse to be of practical use in modifying the generic FIVE data. Since only one plant-specific fire is mentioned in the submittal (a fire in a 4kV station service transformer located in the yard), this claim appears to be valid.

The submittal provides a list of the fire frequencies but does not provide any information to verify that all the fire initiating event frequencies used in the final quantification were calculated correctly.

However, a comparison of the calculated fire frequencies with NUREG/CR-4840 [6] and the FIVE methodology [5] indicates that the calculated fire frequencies are in general agreement with these references. l 2.1.5 Quantitative Screening Assessment All 209 fire compartments defined in the HCGS fire assessment were subjected to quantitative screening. The quantitative screening consisted of three parts: (1) calculation of screening fire frequencies for each compartment (discussed in the previous cection), (2) identification of an l appropriate plant trip initiator assuming all of the equipment in tFe compartrant is failed by a fire,  !

and (3) calculation of screening CCDP using the Hope Creek IPE models with all equipment in the compartment assumed failed (no credit was given for fire suppression). The submittal indicates that the majority of the fires were assumed to result in a main steam isolation valve (MSiV) closure event.

However,in locations where offsite power was impacted, a loss-of-offsite power (LOSP) event was  !

modeled. For a few fire compartments, a loss of HVAC event was used to model ine plant response.

For the MSIV and LOSP events, the IPE event and fault trees were used with some modifications to calculate the screening CCDPs. These modifications include: (1) assigning loss of HVAC scenarios as resulting directly in core damage, assigning all recovery events a probability of 1.0, CRD l and feedwater pumps were assumed to be unavailable for fires in all fire companments due to '

l insufficient information on cable locations, r.nd the condensate pumps were assumed unavailable for fires in the turbine building. For fires resulting in loss of HVAC initiators, the CCDP of 1.0 defined in the IPE for this event was used. However, a room cooling recovery probability calculated in the IPE (3E-4) was credited for one fire compartment (5604/5611/5618) where this initiator was applicable resulting in it being screened. However, the recovery event was not credited for two other fire compartments where fires result in a loss of HVAC initiator (5703/5704 aLd 5620) and thus were retained for further analysis. The reason for this discrepancy is not known. Other human error probabilities were assigned the values used in the IPE (i.e., no fire-induced impacts were assumed).

The screening CCDPs calculated in this portion of the assessment range from 8.61E-6 to 1.0.

E, ,

1 Six fire companments were assumed to have screening CCDPs of 1.0 and thus were all subjected to detailed assessments. These compartments are the cable spreading room, upper control equipment room, lower control equiprnent room, control room equipment room mezzanine, electrical access area, and the control room.

The resulting screening CCDPs were multiplied by the total fire frequency calculated for the corresponding fire compartment to obtain screening estimates of the core damage frequencies. The results of this screening quantitative analysis resulted in core damage frequencies <l.0E-6/yr and elimination of 170 of the 209 fire compartments. Areas that were screened included the RCIC and HPCI pump rooms, some RHR and core spray pump rooms, the emergency diesel generator tank rooms, battery rooms, switchgear HVAC equipment rooms, one emergency switchgear room, most of the turbine building, and the service water intake structure. It is noted that 38 fire compartments are listed in Table 4.8 in the submittal as being retained for further analysis. Thus, the status of only 208 fire compartments is identified in this table instead of 209 referenced throughout the text.

A unique feature of the HCGS fire assessment we ' e performance of a "high hazard area analysis."

This analysis identified areas containing large amot 9 of combustible materials that, if fully involved in a fire, might have the ability to breach barriers. These large fires were postulated to involve all of the fuel and be uncontrolled by suppression or natural limiting phenomena (e.g., oxygen availability).

The frequency of a large fire scenario in each area was estimated using the FIVE database to estimate the fire initiation frequency and using conservative estimates for failure of automatic suppression systems and mar ual suppression efforts. The CCDPs related to fire propagation through rated barriers were not calculated and thus the results of this effort are fire frequencies only, not CDFs.

However, the worst case consequences were qualitatively addressed. The frequency estimates of large fires ranged from 1E-7/yr to 1E-4/yr. The licensee indicates that all of the high hazard areas except for the diesel generator rooms and the turbine building equipment unloading area were quantitatively screened. The compartments quantitatively screened included areas with high fire  !

frequencies but with limited qualitative consequences. Thus, the licensee concludes that the high hazard areas that were quantitatively screened do not significantly contribute to the risk from fires.

A review of the high hazard areas also indicated that an additional high hazard area not listed by the licensee (the switchyard) also was not quantitatively screened.

In summary, typically important fire compartments were not screened out in the quantitative portion of the Hope Creek fire screening. The remaining unscreened areas included the areas identified above and the switchyard blockhouse, three of the four safety-related AC switchgear rooms, the diesel generator rooms, the RACS pump area, the CRD pump area, the remaining RHR and core spray pump rooms, some electrical equipment rooms, and pans of the turbine building. The status of 208 fire compartments is tracked in the tables provided in the submittal instead of the 209 referenced in the submittal text.

l

)

2.1.6 Fire Propagation and Suppression Analysis q Each of the 38 fire compartments that remained after the quantitative screening were subjected to a detailed scenario-by-scenario probabilistic analysis. Fire growth and propagation studies were I i

r, ,

perfomied for different source / target combinations in some of these unscreened companments. Both steady state and pseudo-transient calculations were performed. The steady state calculations were used to estimate the potential for damage assuming an infinite quantity of fuel and the pseudo-transient calculations considered the potential for extinguishment before damage occurs to the targets.

The FIVE fire screening methodology was used in the fire modeling with several modifications. The modifications listed in the submittal include limiting the ceiling jet and plume temperatures to the adiabatic limiting flame temperature, including detector delay impacts due to detector placement and the time for hot gases to rise to the detector, including suppression actuation delay times and cable soak times in estimating the time to suppression, use of the actual line of sight distance'instead of the horizontal distance to the target in estimating thermal radiation efTects, including the fuel depletion time in estimating times to target damage, and conserving the total energy in the fuel in the evaluations. The licensee implemented the modified FIVE fire screening methodology in a series of EXCEL spreadsheets referred to collectively as FireTran.xit. The described modifications are potentially positive enhancements of the FIVE methodology. Ilowever, their implementation has not been verified.

The submittal indicates that liquid pool fires were generally modeled as unconfined in order to

~

maximize the total heat release rate. A heat release rate of 120 BTU /sec/fF was selected for pool fires based on a sensitivity study performed by the licensee. Quantities of 5 to 20 gallons were modeled in the study depending upon the compartment. The submittal indicates the 5 gallon value is based upon allowable storage limits and the 20 gallon limit was used for the diesel generators.

Cabinet fires were modeled as occurring on top of the cabinet with a heat release rate of 1233 BTU /sec. Trash can fires were assumed to be 32 gallons and used the maximum heat release rate l

found in the FIVE methodology (350 BTU /sec). Trash can fires were assumed to burn for 30 minutes and cabinet fires were assumed to burn for 50 minutes. The burn times for oil fires were not provided but were dependent upon the quantity of fuel. The heat loss factor to compartment boundaries used in the fire modeling were 70% of the heat of the fire.

The submittal indicates that sensitivity calculations were performed to determine fire damage and suppression potential. Typically, the height and horizontal displacement of the target was varied in order to develop an overall picture of how the source affects the targets. In addition, the size of the pool fires (in terms of gallons ofliquid fuel) was also varied in some situations in order to establish threshold sizes for damage.

The submittalindicates there are no active fire doors at HCGS. Ilowever, there are some fusible link active fire dampers. Where these dampers exist, the fire growth and damage calculations were performed by treating these dampers as if they were simple openings. The fire growth calculations were performed to investigate the potential for fire propagation between rooms within a single companment through these simple openings (including unsealed penet. itions and cable runs) This was done by assuming a fire located at the wall of the source room which creates a hot gas layer at the elevation of the opening between the rooms. The submittal indicates the calculations demonstrate that hot gas layers can not damage cables in neighboring rooms due to, among other factors, the reduction of the gas layer temperature when the gas is mixed with the air in the receiver room.

I The submittal indicates that all safety-related cables at HCGS are IEEE 383 qualified. However, some electrical cables are not qualified and have been granted an NRC exemption. All exempted cables are in conduits, installed in non-safety-related areas, are located in areas protected by an automatic fire suppression system, and/or are routed in small quantities. A temperature damage criteria of 698 F was used for cabling (only temperature damage criteria were used in the fire damage modeling). Cable damage was calculated assuming all cables were unprotected even if surrounded by conduit or cable tray enclosures. Furthermore, cable damage was assumed for all cables if cables  ;

at any elevation were determined to be damaged. It is unknown if any exempted cables were modeled I in the fire damage evaluations. The only other target evaluated in the fire modeling was the emergency bus bars located in the diesel generator rooms. The damage criteria for these aluminum bus bars was assumed to be melting temperature of aluminum (1220 F).

Important insights obtained from the fire damage and suppression calculations include: (1) cabinet fires only damage cables directly above them due to plume efTects (the high ceilings and large rooms at HCGS prevent damage from ceiling jets and the hot gas layer below the ceiling jet does not reach damage temperatures), (2) damage was usually, but not always, calculated to occur before extinguishment was calculated to occur, and (3) the steady state calculations for pool fires resulted in target damage not obtained in the pseudo-transient calculations where the fire burned out before the target damage occurs.

The submittal in'dicates that the FIVE fire damage models contain unrealistic assumptions and provide results that exaggerate the amount of damage and minimize the time to cause the damage. As a result, the licensee also states that the damage to overhad cables was calculated to occur in a short period of time. Citing actual fire suppression data, the licensee states the damage times are short in comparison to the time required to suppress a fire. For this reason, the HCGS fire assessment does not credit either automatic or manual fire suppression except for the inherent suppression credit contained in the partitioning factors used in the control and diesel generator room analyses (see the next section) and in the compartment boundary assessment performed in the FCIA. '!

1 In summary, the fire propagation modeling appears to follow the guidelines provided in the FIVE methodology with some modifications. Although the implementation of these modifications has not  :

been reviewed, they appear to be reasonable. Finally, the HCGS fire assessment did not credit fire j suppression except for in the FCI A portion of the assessment and in the severity factors used in the l control room and diesel generator room evaluations.

i 4

2.1.7 Final Quantification and Uncertainty Analysis i

I For each fire scenario within the unscreened fire compartments, a probabilistic assessment of the CDF was performed. This assessment included evaluation of the fire frequey ,= ue Section 2.1.4 in this report), determination of the type of accident resulting from the fire, and determination of CCDP using the IPE model with the equipment damaged by the fire failed.

l i

]

n, ,

All fire scenarios were assumed to result in a plant trip with most scenarios identified as resulting in an MSIV closure event. However, some scenarios were identified as rer,ulting in LOSP, loss of HVAC, inadvertent open relief valve (IORV), or loss of SSWS/ SACS events. In some scenarios, a hot short was identified as required to result in a particular type of accident initiating event (e g.,

a fire-induced IORV or loss of SSWS/ SACS). A hot short probability of 0.3 was assigned as the conditional probability of that type of event occurring given the fire scenario and the complement probability (i.e.,0.7) was assigned to the probability of an MSIV closure event (this preserved the total scenario fire frequency). The submittal indicates that the potential for fire-induced interfacing system LOCAs was reviewed but dismissed due to the presence of check valves or disabled motor-operated valves in the y stems.

The CCDP for each fire scenario was evaluated using the appropriate event tree and required mitigating system fault trees with the equipment calculated to be damaged by the fire given failure probabilities of 1.0. In these evaluations, the CRD, condensate, and feedwater pumps were assumed unavailable. The CCDPs range from 1.3E-5 to 1.0 (calculated for one transient oil fire scenario in the cable spreading room). Only two human recovery actions were modeled. One was alignment of alternate ventilation following a loss of switchgear room HVAC (assigned a value of 3 E-4) and the other was for controlling the plant from the remote shutdown panel following evacuation of the control room (see following discussion). No credit was given for recovery of offsite power or the power conversion system in any scenario. Finally, it is noted that although het shorts appear to have been considered in the evaluation of the type of accident caused by the fire, the impact of hot shons on the subsequent operation of mitigating systems is not addressed in the submittal.

The CDF due to fires in the control room was evaluated in a different fashion. Two types of fires were postulated: small fires limited to the cabinet where the fire occurs but evacuation is not required (the remote shutdown panel may be required to regain control of failed components) and large fires where the operator has to evacuate the control room and use the remote shutdown panel (e g., fires where suppression fails). The frequency of an individual cabinet fire was obtained by panitioning the total control room fire frequency by the ratio of the length of the individual cabinet over the total length of all cabinets in the control room. The corditional probability that a specific cabinet fire is large was obtained by reviewing control room fire data. Bayes Theorem was applied using a uniform prior distribution and evidence that all 12 fires were not large. A conditional probability of 0.028 was obtained. The licensee correctly indicates that this probability implicitly includes the contribution of fire suppression. For the small fires, the type of accident event and corresponding CCDP was calculated in the same fashion as for the other 37 fire compartment scenarios. For the large fire scenarios, the CCDP was assumed equivalent to the human error probability that the operator fails to control the plant from the remote shutdown panel. A probability of 0.06 was used for large fires  ;

i resulting in MSIV closure events and a probability of 0.1 was used for the other types of events which I the licensee judged to require more complex operator actions (i.e., LOSP, loss of SSWS/S ACS, loss {

of HVAC, and IORVs). )

Fires in the diesel generator rooms were also separated into small and large fires with large fires l resulting in damage to the overhead off-site power bus bars and a subsequent LOSP. As with the 1

l j

[. .

1 i

l l

control room, a Bayesian analyds of the fire events was performed by the licensee and resulted in a large fire probability of 0.025. The licensee correctly indicates that this probabilit y implicitly includes the contribution of fire suppression. The licensee also indicates this value compares favorably with the failure probability ofCO2systems which are present in the diesel generater rooms The large fire scenarios are the dominant contributors in the diesel generator rooms contributing 95% of the room-related CDF.

The estimated CDF from fires in the unscreened compartments is 8.lE-5/yr. The control / diesel building, which houses the control area and the diesel generators, contributes 86% of the fire-induced CDF at Hope Creek. The submittal indicates that the control / diesel building is important because there is a confluence of equipment and cables for different electrical divisions in the building, particularly in the cable spreading roorn. The imponant companments and the resulting core damage j

frequencies are identified below. ,

Compartment Compartment Description CDF 5510/5511 Control room 2.5 E-5/yr 5416/5417 Class 1E switchgear room (Channel A) 1.3E-5/yr 5307 Diesel generator room (Channel A) 5.3 E-6/yr 4202 CRD pump area 4.2E-6/yr 5306 Diesel generator room (Channel B) 4.1 E-6/yr 5305 Diesel generator room (Channel C) 3.7E-6/yr 5412/5413 Class 1E switchgear room (Channel B) 3.0E-6/yr 5501 Electrical access room 3.0E-6/yr 5339 Electrical access area 2.7E-6/yr 5605/5631 Upper control equipment room / computer room 2.7E-6/yr 5304 Diesel generator room (Channel D) 2.6E-6/yr 5401/3425 Electrical access area 2.0E-6/yr 4301/4309/4310/ Reactor building 102' elevation-nonh side and 1.8E-6/yr 4311 Division 1 SACS area 5302 Lower (control) electric equipment room 1.7E-6/yr l 1315/1316/1317/ Access and unloading area (turbine building) 1.2E-6/yr 1320/1321/1322 4303 Motor-control center (MCC) area 1.2E-6/yr I

o .

The licencee indicates in the submittal that the most imponant companments are those where multiple electrical channels are found in close proximity (e g., the control room, electrical access rooms, control equipment rooms, MCC areas, and SACS equipment room) or compartments where a LOSP can occur (e g., the diesel genersor rooms, electrical access area, and the access and unloading area in the turbine building).

The submittal also provides breakdowns of the CDF by ignition source and initiating event. Cabinets and diesel generators are the most important ignition sources followed by transient sources. MSIV closure and LOSP are the dominant types of accidents caused by fires at Hope Creek. The importance of MSIV closure stems from the assumption that all fires will at least result in that type of transient. The importance of LOSP arises mostly from fires in the diesel generator rooms and reflects the fact that both sets of offsite power bus bars run through all four diesel generator rooms.

No uncertainty analysis was performed in the fire assessment. However, the licensee does provide a discussion of the sources of uncertainty in the analysis which includes use of generic data, the assumption that all fires result in at least an MSIV closure event, and the failure of all cables within an electrical division when any one cable was found to exceed the cable damage criterion.

2.1.8 Sensitivity and Importance Ranking Studies One core damage frequency sensitivity study was reported in the Hope Creek submittal. As mentioned previously, fire suppression was not credited since the fire modeling predicted it would not occur in time to prevent cable damage. The licensee states the fire modeling is conservative and in recognition of this, performed a sensitivity study which assumed fire suppression could occur in time to prevent damage. Fire suppression was modeled in rooms that have suppression systems (i e.,

some electrical equipment rooms and the turbine building unloading area). The existing CDFs for these rooms were multiplied by the failure probability of a preaction sprinkler system (the value used was not specified in the submittal). This resulted in a total fire-induced CDF of 7.6E-5/yr which is  ;

only 6% lower than th base case value.

l As indicated in the previaus section, the licensee presented the importance rankings for the buildings, ignition sources, compartments, scenarios, and accident initiator types. In addition, the importance of each electrical division was also presented (no one division dominated the results).

l 2.2 Special Issues As a part of the IPEEE fire submittal, the utilities were asked to address a number of fire-related  !

issues identified in the Fire Risk Scoping Study (FRSS) [7] and USNRC Unresolved Safety Issues (USI) and Generic Safety issues (GSI). Specific review guidance on these issues is taken from Reference 4.

+ ,

1 I

2.2.1 Decay lleat Removal (USI A-48)

.. .. a . .

As discussed in Generic Letter 88-20 [2] and NUREG-1407 [3], USI A-45 which is associated with I

-the adequacy of decay heat removal (DifR) at nuclear power plants is subsumed into the IPE f submittals.' . A submittal meeting the intent of Generic Letter 88-20 Supplement 4 is assumed to j satisfy the requirements of USI A-45. Specifically, the fire assessment presented in the IPEEE submittal should address the adequacy oflong-term decay heat removal in the event of fires.

The liope Creek submittal provides a discussion on USI A-45 with regard to the fire assessment.

Specifically, the licensee estimates that the sequences which either directly or indirectly lead to failure

]

, of decay heat removal (identified in the submittal as failure of RHR and containment a nting) i represent an approximate CDF of 6.4E-6/yr or about 8% of the fire-induced CDF. The li<ensee I incorrectly excludes sequences involving LOSP, scenarios requiring control room abandonment, and scenarios resulting in loss of coolant injection (also required for DIiR) from the estimated DlIR-related CDF. It is likely that the real DliR-related CDF is closer to the total CDF of 8. lE-5/yr.

2.2.2 Effects of Fire Protection System Actuation on Safety-related Equipment (FRSS, GSI-57, MSRP)

This issue is associated with the concem that traditional fire PRA methods have generally considered only direct thermal damage effects. Other potential damage mechanisms, such as smoke and fire suppression damage (either from fixed systems or manual actions) have not been considered. In general, this is an area where the data base on equipment vulnerability is rather sparse. It was anticipated that a typical IPEEE analysis would provide for some treatment of both smoke and suppression-induced damage. '

The liope Creek submittal specifically addresses GSI-57. The submittal indicates that 11 CGS uses preaction sprinkler systems with fusible links, CO deluge 2 systems, and backup manually actuated water deluge systems for safety-related areas in which automatic fire suppression is provided. The licensee addresses the potential for seismic actuation of the automatic suppression systems (see Section 2.2.5 in this report). In addition, the licensee used screening criteria to evaluate the effect ofinadvenent suppression in all areas of the plant. The screening criteria included the presence of safe shutdown equipment, the type of suppression system, separation of the safe shuidown equipment, and whether the suppression system is internal to the safety-related equipment.

After applying the criteria, the licensee indicates only one room was identified as having a potential impact from inadvenent fire suppression system actuation. The control equipment room mezzanine (room 5403) contains control circuitry for both electrical divisions and has both an automatic CO, system and a manually actuated deluge system. Inadvertent manual activation of the deluge system could disable both divisions of power by electrically shoning cabies. The submittal indicates that loss

! ofboth divisions in this room was evaluated in the UFSAR and that safe shutdown can be achieved in this situation from the remote shutdown panel.

1

[: _. .

l 2.2.3 Fire-induced Alternate Shutdown / Control Room Panel Interactions (FRSS, GSI-147)

' The issue of control systems interactions is associated primarily with the potential that a fire in the main control room or in the cable spreading room might lead to potential control systems vulnerabilities. Given a fire in the plant, the likely sources of control system interactions are between the control room, remote shutdown panel, and the shutdown systems. Specific areas that should be addressed in the IPEEE fire analysis include: (1) electrical independence of the remote shutdown control systems, (2) loss of control equipment or power before transfer, (3) spurious actuation of components leading to component damage, a LOCA, or an interfacing system LOCA, and (4) total loss of system function. It is anticipated that a licensee's submittal would describe its remote shutdown capability including the nature and location of the shutdown station (s) and the types of control actions which can be taken from the remote panel (s).

The submittalindicates that a remote shutdown panel and redundant shutdown instrumentation and controls are present at HCGS. Once control has been transferred to the remote shutdown panel by operating transfer switches at the panel, the reactor can be brought to a hot shutdown condition (a few pumps and valves must be operated locally to bring the plant to cold shutdown). The submittal indicates that use of the remote shutdown panel may be required for large fires resulting in damage to normal safe shutdown circuits in the cable spreading room, lower control equipment room, control equipment room mezzanine, main control room, lE panel room HVAC corridor, upper control equipment room, and the diesel area HVAC room. However, the use of the remote shutdown panel appears to have only been modeled in the fire assessment for control room fires.

The submittal indicates that if offsite power is lost, the diesel generators will start automatically.

Indication of their successful start is provided at the remote shutdown panel. If they do not automatically start,' the diesel generators can be manually started at the diesel generator control panels. There is no indication in the submittal that electrical isolation exists for the diesel generators j between the control room and the local control panels. j in summary, the discussion of this issue in the Hope Creek submittal is brief and does not specifically address the potential for loss of control equipment or power before transferring to the remote

{

shutdown panel or spurious actuations due to hot shorta The submittal indicates there is a procedure  ;

for transferring control to the remote shutdown panel.

-2.2.4 Smoke Control and Manual Fire Fighting Effectiveness (FRSS, GSI-148)

Smoke control and manual fire fighting effectiveness is associated with the concern that nuclear power plant ventilation systems are known to be poorly configured for smoke removal in the event of a fire, and hence, significant potential exists for the buildup of smoke to hamper the efforts of the fire brigade to promptly and etTectively suppress fires. Sensitivity studies have shown that prolonged fire fighting times can lead to a noticeable increase in fire risk. Smoke, identified as one of the major contributors to prolonged response times, can also cause misdirected suppression efrorts and hamper the operator's ability to safely shutdown the plant.

s -

In the Hope Creek IPEEE fire analysis, manual fire fighting is not credited in the fire scenarios considered in the assessment. Nevertheless, the submittal addresses the effectiveness of manual fire fighting activities at Hope Creek. Based on this review, the licensee concludes the plant's fire brigade and manual fire fighting capability are effective.

Specifically, the submittal describes the availability of programs and procedures for identifying fires and calling the fire brigade. The makeup of the fire brigade, their equipment, the fire brigade training program, and the frequency of drills are all briefly described in the submittal. The Hope Creek submittal indicates that all fire brigade personnel are trained in the use of self-contained breathing apparatus and ventilation techniques through the use oflive smoke training. The licensee also indicates that pre-fire plans have been developed and are used in the fire brigade training. Records are maintained for all fire drills and personnel training and all fire brigade equipment is inspected on a periodic basis.

The submittal did not discuss the potential that manual fire fighting activities might lead to damage to equipment from misdirected fire suppressants, but did qualitatively address the potential efTects of smoke on equipment. After discussing the potential impacts on various types of equipment, the licensee concludes that the effect of smoke on equipment is insignificant with respect to the impact of the fire itself. The submittal also indicates that the impact of smoke on personnel performance was included in the control room evaluation since evacuation is assumed required for large unsuppressed fires. A separate smoke removal system for the control area of the plant is also discussed in the submittal and involves opening a normally closed shutoff damper and fit e damper and then manually l

starting a smoke exhaust fan. j 2.2.5 Seismic / Fire Interactions (FRSS, MSRP) l The issue of seismic / fire interactions involves primarily three concerns. First is the potential that seismic events might also cause fires internal to the plant. Such threats might be realized from inadequately secured liquid fuel or oil tanks, through breakage of fuel lines, or through the rocking of unanchored electrical panels (either safety or non-safety grade). The second concern is the potential that seismic events might render fixed fire suppression systems inoperable. This could include detection systems, fixed suppression systems, and fixed manual fire fighting support elements such as the plant water distribution system. The third concern is that a seismL event might spuriously actuate fixed fire detection and suppression systems. Actuation of a suppression system may lead to flooding problems, habitability concerns (in the case of CO 2systems), diversion of suppressants to non-fire areas rendering them unavailable in the event of a fire elsewhere, the potential over-dumping ofgaseous suppressants resulting in an over-pressure of a compartment, and spraying ofimportant plant components. It had been anticipated that a typical fire IPEEE submittal would provide for some treatment of these issues through a focused seismic / fire interaction walkdown. The Hope Creek submittal addresses all three concerns. Plant walkdowns were performed to address these issues.

l The submittal identifies a list of equipment investigated as potential seismic-induced fire sources.

l This list included hydrogen piping and storage tanks, emergency diesel generator fuel oil piping and l

1 L

associated storage tanks, turbine-related sources (lubricating oil tanks and piping, hydrogen , and the hydrogen seal oil unit and associated tanks and piping), other sources of flammable liqus ; and gases such as the reactor recirculation pumps, and unanchored non-lE cabinets. The specific location of thu equipment is not provided in the submittal but is identified as being delineated in the UFS AR.

The submittal indicates that the seismic ruggedness of all these sources was reviewed during the seismic walkdown and all were found to be sufliciently rugged. The anchorages of both IE and non-IE cabinets were included in the walkdowns. The anchorages for all non-lE cabinets were either screened out or found to have median capacities in excess of 1.5g. Therefore, the licensee concludes that seismic interactions of non lE cabinets and IE equipment is not a significant fire concern.

As previously indicated, HCGS uses preaction water sprinklers, CO 2deluge systems, and backup manually actuated water deluge systems in safety-related areas in which automatic fire suppression is provided. The submittal indicates the seismic ruggedness of the fire suppression systems was reviewed in the seismic walkdown. The submittal indicates that all fire protection systems are designed such that failure or inadvertent operation does not result in loss of safety-related systems.

Specifically, piping located over safety-related equipment is supported such that it will not fait during a safe shutdown earthquake, water system pressure is low enough so that pipe whip protection is not required, and the pressurized portion of the CO 2 systems is located outdoors. The licensee concludes that the piping, sprinkler heads, and other components of the suppression systems are seismically robust. However, the fire water pumps are identified as being located in a block wall structure that is not seismically qualified. In addition, the fire water tanks are also identified as not being seismically quahfied.

The Hope Creek submittal also discusses the potential for seismically-induced actuation of fire suppression systems. The licensee indicates that the preaction sprinklers used in many areas of the

. , plant are not susceptible to seismic actuation because they require two independent, dissimilar failure L modes to allow water discharge. Since the systems in safety-related areas are heat activated instead l ' of smoke activated, the licensee indicates that inadvertent actuation by smoke or dust is not an issue

! in these areas. The submittal also indicates that the CO 2systems serving safety-related areas have seismically qualified components to avoid inadvertent discharge. A low ruggedness relay evaluation performed by the licensee is also referenced in the submittal. The licensee states that none of the l identified low ruggedness reiays are in the fire protection or detection systems 2.2.6 Adequacy of Fire Barriers (FRSS)

The common reliance on fire barriers to separate redundant components needed to achieve safe shutdown has elevated the risk sensitivity of fire barrier performance. Degraded fire barrier penetration seals and unsealed penetrations in some barriers can contribute to this source of fire risk since fires in one area might impact other adjacent or connected areas through the spread of heat and smoke. In general, it is expected that a utility analysis would provide for some treatment of such potential by considering that (1) manual fire fighting activities might allow for the spread of smoke and heat through the opening of access doors, and (2) that the failure of active fire barrier elements such as normally open doors, water curtains, and ventilation dampers might compromise barrier

-18

c- .

1 1

integrity. Resolution of the fire barrier issue involves verification that fire barriers are properly installed and maintained under a surveillance program.

The submittal indicates that 10% of the penetration seals are inspected every I8 months to verify their integrity. Fire doors are included in the HCGS fire protection surveillance requirements and are subjected to inspection every six months. Fire dampers are also inspected at least once every 18 months to verify their operability. Two NRC Inspection and Enforcement Notices (ins) pertaining to fire dampers are also addressed in the submittal: IN 89-52 which addresses potential closing problems with curtain type fire dampers under system air flow conditions and IN 83-69 which identifies three specific fire damper installation deficiencies. The licensee indicates the concerns in these ins were addressed in a major fire damper improvement conducted at HCGS in 1985 and is documented in the UFSAR. Similarly, three ins related to penetration seals were also identified and discussed in the submittal: IN 88-56 dealing with potential problems with silicone foam barrier penetration seals and IN 88-04 and IN 88-04 Supplement I both dealing with inadequate qualification of seals. The licensee indicates the concerns raised by these ins have been addressed in internal memos and are either not applicable to HCGS or adequately controlled.

2.2.7 Effects ofIlydrogen Line Ruptures (MSRP)

The use of flammable gases in the plant, including hydrogen, introduces the potential that a rupture of the gas flow lines might lead to the introduction of a serious fire hazard into plant safety areas.

It had been anticipated that a typical fire IPEEE analysis would include the consideration of such sources in the analysis.

The Hope Creek submittal does not specifically address this issue beyond the seismic evaluation of hydrogen piping discussed in Section 7.2.5 of this report.

2.2.8 Common Cause Failures Related to lluman Errors (MSRP)

Common cause failures resulting from human errors include operator acts of omission or commission that could be initiating events or could affect redundant safety-related trains needed to mitigate other initiating events. It had been anticipated that a typical fire IPEEE analysis would include the consideration of such failures in the submittal.

The Hope Creek IPEEE submittal indicates that only two human recovery actions were included in the fire assessment. No other discussion of human errors is provided in the submittal.

2.2.9 Non-safety Related Control System / Safety Related Protection System Dependencies (MSRP)

Multiple failures in non-safety-related control systems may have an adverse impact on safety-related protection systems as a result of potential unrecognized dependencies between control and protection systems. The licensee's IPE process should provide a framework for systematic evaluation of interdependence between safety-related and non-safety related systems and identify ta

potential sources of vulnerabilities. It had been anticipated that the fire IPEEE analysis would include the consideration of such dependencies in the submittal.

The submittals discussion on the ability to shutdown the plant from the alternate shutdown panel is summarized in Section 2.2.3 of this report.

2.2.10 Effects of Flooding and/or Moisture intrusion on Non-Safety- and Safety-Related Equipment (MSRP)

Flooding and water in';usion events can affect safety-related equipment either directly or indirectly through flooding or moisture intrusion of multiple trains of non-safety-related equipment. This type of event can result from external flooding events, tank and pipe ruptures, actuations of the fire suppression system, or backflow through part of the plant drainage system. It had been anticipated that the fire IPEEE analysis would include the consideration of such events in the submittal.

As indicated previously, llCGS uses preaction sprinkler systems with fusible links, CO2 deluge systems, and backup manually actuated water deluge systems for safety-related areas in which automatic fire suppression is provided. The licensee addresses the potential for seismic actuation of the automatic suppression systems (see Section 2.2.5 in this report). In addition the licensee used screening criteria to evaluate the effect ofinadvertent suppression in all areas of the plant. The screening criteria included the presence of safe shutdown equipment, the type of suppression system, separation of the safe shutdown equipment, and whether the suppression system is internal to the safety-related equipment.

2.2.11 Shutdown Systems and Electrical Instrumentation and Control Features (MSRP)

The issue of shutdown systems addresses the capacity of plants to ensure reliable shutdown using safety-grade equipment. The issue of electrical instrumentation and control addresses the functional capabilities of electrical instrumentation and control features of systems required for safe shutdown, including support systems. These systems should be designed, fabricated, installed, and tested to quality standards and remain functional following external events. It had been anticipated that the fire IPEEE analysis would include the consideration of this issue in the submittal.

A summary of the submittals discussion on decay heat removal is presented in Section 2.2.1 of this report.

2.3 Containment Performance Issues Unique to Fire Scenarios The licensee for Hope Creek examined the containment performance for fires in the 38 fire compartments that were not screened in the FIVE assessment. This was accomplished by examining the potential for bypass, isolation failure, direct containment failure, and containment system I

t

degradation or failure. The submittal indicates the evaluation was performed using the list of equipment, cabinets, and cables compiled for the fire PRA and included the effects of the failure of those components including the effects of hot shorts. It is specifically noted in the submittal that the potential for interfacing system LOCAs is small due to the presence of check valves or disabled motor-operated valves in all high-low pressure interfacing systems. The licensee concludes that there are no fire-induced containment failure modes that are significantly difTerent than those treated in the IICGS IPE.

2.4 Plant Vulnerabilities and Improvements The llope Creek licensee defines a vulnerability as a scenario which contributes inordinately to the HCGS CDF, as compared to other plants of similar type and vintage, thus representing a substantial design weakness of the plant. No fire-related vulnerabilities were identified in the submittal. In addition, no fire-related plant improvements were discussed.

1 l

3.0 CONCLUSION

S AND RECOMMENDATIONS The Hope Creek fire assessment was performed using the FIVE methodology to perform quantitative screening of fire compartments. An FCIA was also performed per the FIVE methodology but no qualitative screening was performed. Unscreened fire compartments were subjected to a detailed probabilistic assessment ofeach identified source / target combination in the compartments The FIVE fire modeling methodology was used, with some modifications, to evaluate fire propagation and damage potential. No credit was given for fire suppression in either the screening or detailed assessment (except for the crediting of fire suppression systems in the FCIA). Both levels of assessment used generic fire frequency data from FIVE and used the IPE models to quantify the i conditional core damage probability for all fire scenarios.

In general, the Hope Creek IPEEE submittal adequately documents the methods and results of the fire assessment. The assessment also addresses the FRSS issues, USl A-45, and GSI-57. Some areas of strengths and minor weaknesses were identified in the analysis. They include:

i Strenuths i

t No qualitative screening was performed The licensee assumed that a plant trip would occur for  ;

all fires. )

• The licensee performed a separate "high hazard area analysis" to address whether areas containing large amounts of combustibles represent a significant fire risk.

l

• The analysis includes credit for non-safety related equipment in the analysis.

The FIVE fire damage models were modified by the licensee to provide more realistic results.

However, these modifications have not been verified.

• The licensee performed fire propagation sensitivity studies to establish critical amounts of combustibles and to develop an understanding of how source fires affects the targets.

• Available PRA models were used to evaluate the conditional core damage probability in both the screening and detailed assessments.

• Only two human recovery actions were credited in the assessment. The impact of the fire on the probability assigned for these actions is addressed in the submittal.

• The fire assessment includes hot shorts in the evaluation of the type of accident initiating events caused by the fires.

• Although fire suppression was not explicitly credited in the analysis, the licensee did perform a sensitivity study on the impact of automatic fire suppression on the CDF.

Minor Weaknesses

• The~ analysis of hot shorts appears to be limited to determining the type of accident initiated by the fire. The potential for hot shorts negatively impacting mitigating systems is not specifically addressed.

. The licensee used fire severity factsrs in the evaluation of the control room and the diesel generator.redms. An evaluation that modeled fire growth and suppression considering plant specific factors would be more representative.

. The estimate of the DHR-related CDF in the USI A-45 discussion improperly eliminated important contributors The reviewer reconunends that a sufficient level of documen'tation and appropriate bases for anaiysis have been established to conclude that the subject licensee submittal has substantially met the intent of the IPEEE process. No further review is recommended.

4.0 REFEllENCES

1. "flope Creek Nuclear Station Individual Plant Examination tbr External Events," Public Senice Electric and Gas Company, July 1997..

2. USNRC," Individual Plant Examination of Fxternal Events for Severe Accident Vulnerabilities -

10 CFR s50.54(f)," Generic Letter 88-20, Supplement 4, April 1991.

3. USNRC," Procedural and Submittal Guidance for the Individual Plant Examination of External Events (IPEEE) for Severe Accident Vulnerabilities," NUREG-1407, May 1991.

4. S. Nowlen, M. Bohn, J. Chen, " Guidance for the Perfbrmance of Screening Reviews of Submittals in Response to U.S. NRC Generic Letter 88-20, Supplement 4: ' Individual Plant Examination - External Events,'" Rev. 3, March 21,1997.

5. EPRI, " Fire-Induced Vulnerability Evaluation (FIVE)," EPRI TR-100370 April 1992. {

l

6. M. Bohn and J. Lambright, " Procedures for the Extemal Event Core Damage Frequency Analyses for NUREG-1150," Sandia National Laboratories, NUREG/CR-4840, November 1990.

7. J. Lambright, " Fire Risk Scoping Study: Investigation of Nuclear Power Plant Fire Risk, including Previously Unaddressed issues," Sandia National Laboratories, NUREG/CR-5508, January 1989..

1 I

I 1

l l

1 J