ML20217G048
| ML20217G048 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 01/02/1997 |
| From: | GENERAL PUBLIC UTILITIES CORP. |
| To: | |
| Shared Package | |
| ML20217G044 | List: |
| References | |
| REF-GTECI-A-46, REF-GTECI-SC, TASK-A-46, TASK-OR GL-87-02, GL-87-2, GL-88-20, GL-96-06, GL-96-6, NUDOCS 9710090164 | |
| Download: ML20217G048 (48) | |
Text
Mfrty and Risk An:1pn l' age i of 48 Rnision 0 RISK ASSESSMENT OF DEFEltitED OYSTEll CitEEK PRO lECTS l
9710090164 971001 PDR ADOCK 05000219 P
pop I
asena<a w Ia w n
9 klet) and Rid An J) sis P:ge 2 of 48 Resism o TAllLE OF CONTENTS i
i 1.01NTHODUCT1ON.......................................................................................................................3 2.0 MET110D............................................................................................
- 3 3.0 E VA L U A T l D N O F P R O J E CTS................................................................................................ 5 3.1 Step 1 - dvaluation of the Status of the Projects Proposed for Deferral..................
......5 3.2 Step 2. Evaluate the Safety or Risk impact.................................
.....6 3.2.1 Generic Letter 96 06 Modincations.............
.....6 3.2.2 Seismic Quall0 cation Mod 10 cations - Phase il.................................................... 6 3.2.3 Control Room 11uman Factors Design Review...
...............7 3.2.4 Anticipatory Scram flypass Logic improvement........
...................8 3.2.5 Thermo-Lag Fire 11arrier Modincatlons..................................
..............................8 3.2.6 Severe Accident Management Guideline Development..........................................
.9 3.2.7 Reactor Water Cleanup Leakage Monitoring. LOCA Detection and Isolation............... 9 3.3 Step 3 Categorire Safety / Risk impacts...........................................................,..... 10 3.4 Step 4 Evaluate the Integrated Safety / Risk Impact............................................ 12 4.0 CO N C L U S I O N S.....................................................
..............................................I3 5.0 RE FER ENCES...........
..........................I4 APPEND 1X A: DEFFERRED PROJECT LIST.......
....................................................I6 APPENDIX H: RISK IMPACT ANALYSIS..............
.....................I8 11.1 Generic Letter 96-06 Modi 0 cations...........
,...........................19 11.2 Seismic Qualineation Modi 0 cations - Phase 11.......
.........,..........26 11.3 Anticipatory Scram Dypass Logic improvement........................... --............... 29 B.4 Thermo-Lag Fire Barrier Modi 0 cations......................................................... 32 B.5 Reactor Water Cleanup Leakage Monitoring, LOCA Detection and Isolation..,................... 34 APPENDIX C: FIRE INDIVIDUAL PLANT EXAMINATION METiloiv0 LOGY OVERVIEW. 40 derened 01 doc 10/02/97
_ ~ _ _.
hid) c.ndlM Anxiple hre 3 ol ds Rei kkm 0 1.0 IN1RODUCilON i
Proside Risk Analpls support of the planning activities that consider the following potential Opter Creek I
scenarios:
l r
1.
Continued Operation to the end of licensed life (2009) 2.
Sale of Oyster Creek to a third pany 3,
i:arly Shutdown in September,2000 Specincally, prmide the risk impact of the deferral of projects until the ISR refueling outage in support of the potential early shutdown in the y ear 2000.
2.0 MElllOD The Opter Creek commitments have been reviewed and grouped into three categories by the Project and Regulatory Review 1 cams. The three categories are:
Defer now, before final plant decision is made Car;cel after final plant decision is made implement as originally committed e
for the projects which are to be deferred, provide a risk analysis of the impact of the deferral, in addition, provide an integrated assessment of the risk impact associate with the deferral of the proposed projects. The process for the evaluation is divided into four steps.
First, Esaluate the Status of the Projects within the framework of the various risk analysis studies performed for Oyster Creek. In this step it is determined whether the risk impact of project deferral can be reflected or inferred using the previously descloped risk analysis studies.
Review the available risk analysis studies (ic., Probabilistic Risk Assessments (PRAs) and Esternal Event (IPLEE) analyses.)
Review the deferred projects.
Define whether impact of the deferred projects can be directly or indirectly inferred from available risk evaluations.
Sceond, Es aluate the Safety or Risk Impact of the proposed deferred individual projects.
If the risk impact of the deferral of the project can be directly produced using the available risk studies, perform the evaluation and provide the risk impact.
If the risk impact cannot be directly inferred, however, minor modifications to existing evaluations can be performed, perform modifications and provide the risk impact.
If the risk impact cannot be either directly or indirectly inferred from esisting risk evaluations, either:
Perfonn additional risk evaluations and provide the impact, or dercrred 03 dos 10/02M7
~ _ _ _ _ _ _ _ _
hilcty endlM Andpls hre4 M48 kes ision o Qualitatively assesses the risk impact in a framework that lends itself to incorporation with quantitathcly assessed risk impacts produced in the steps above.
As part of this step, individual projects with signl0 cant risk impacts may be addressed in part.
That is, risk signincant portions of an individual project may be recommended for completion on the current schedule with the remainder of the project being deferred if the risk impact is large and cannot be reduced by performing portions of the project or compensatory measures, the project will be recommended for completion on the original schedule. This ensures that the proposed risk increase remains small.
Third, Categorlie All Safety / Risk Impacts using categories of high, medium and low, for the quantitatively produced risk impacts this consists of assigning numerical increases in core damage frequency or large early release frequency to pre-denned ranges. In the case of qualitatively evaluated imprats this consists of an assessment based on judgement. Assignment of a risk category allows for the integration of the risk impacts in cases where ditTerent Ogures of merit may be used to evaluate projects or activities.
Fourth. Esaluate the Integrated Safety /Hisk Impact. Using the categories established in step three, provide a Onal integrated risk assessment. In the case of the qualitative evaluations, weighting factors based on judgement may be required.1his step allows for the risk impacts to be considered in an integrated manner and as part of an overall risk management approach.
The Ogures of merit used in the evaluation of the quantitative risk impact are core damage frequency and large early release frequency (LERF). These Ogures of merit are chosen since most previously performed risk studies evaluate the frequency of core damage or large early release frequency. Other qualitative factors such as, consideration of alternative endstates, (e g., signincant transients) are documented in the individual evaluations. These qualitative factors car affect the allocation of a project to a given risk category.
In overview, the above methodology agrees closely with the methods for the use of PRA methods in risk informed decision making outlined in the NRC drall Standard Reuew Plan, *Use of Probabilistic Risk Assessment in Plant Specine, Risk Informed Decisionmaking: acneral Guidance" (reference 1). For comparison purpo:,es:
Steps 1 and 2 are equivalent to the Orst element in the drall SRP, Denne the Proposed Change.
Steps 2 through 4, correspond to Element 2 of the drall SRP, Conduct Engineering Evaluations.
The third element of the drail SRP, Develop Implementation and Monitoring Strategies is also addressed in steps 2 through 4 on an individual project basis. Each project is evaluated for the prtential for risk reduction, including compensatory measures, for example, fire watches have been posted in Orc zones which contain thermolag Ore barriers. The evaluation of implementation and Monitoring strategies is performed on an activity or project basis depending on risk impact of the project deferral and the risk reduction achievable with potential compensatory measures. Also, i
performing parts or portions of projects are considered potential compensatory measures. for l
example, the most risk signincant portions of a project may proceed as planned while less risk signincant portions are deferred for a single cycle.
The fourth element in the drafi SRP is represented in the submittal of the integrated schedule to the NRC. The submittal and supporting documents contain suf0cient information to support the conclusions of the acceptability of the deferrals and are available for staff review.
derened 03 doc 10/02M
salm endr u Anainh i
t'Mf t $ Or 48 kn hhm 0 3.0 EVAL.UAllON Of PROJECTS As stated previously, the Opter Creek commitments and projects base been resiewed and pouped into three categories by the Project and Regulatory Rctiew Teams. The three categories are: Defer, Cancel or implement as originally committed.1he following projects have been proposed to be DEI 1:RRED:
D.I. Generic Letter 96-06 hiodi0 cations 112. SQUG - Seismic Quall0 cation hiodincations D.3. Control Room iluman Factors Design Review (llack Panels)
D.4. Anticipatory Scram Logic hiodification D.$. 1hermolag i ire llarrier hiodincations D.6. Sesere Accident hinnagement Guidelines D.7.
Reactor Water Cleanup Automatic Isolation hiodincation Complete descriptions of the projects proposed for deferralis available in Appendit A of this report.
3.1 Step i - 1:saluation of the Status of the Projects Proposed for Deferral lhe goal of this step is to detennine whether the rkk impact of the deferral of the above projects can be estimated using the available risk analpes done for Oyster Creek.1he risk analyses perfonned in support of Oyster Creek include the Oyster Creek Probabilistic Risk Assessment (OCPRA) and the Opter Creck IPE for External 1:sents (IPEEE).1he Opter Creek IPEEE includes a Sekmic PRA La well as a hiodined t ire PRA.
Project or Activity Applicabk Rhk livaluation D,1. Generic Ectter 96 06 hiodifications Level 2 nCPRA D.2. SQUG - Seismic Quali0 cation hiodifications Seismic PRA D.3 Control Rooro lluman Factor Design Review (llack Panels)
Qualitative D.4. Anticipatory Scram Logic hiodification OCPRA D.5. 1hennolag i i.v Ilarrier hiodincations Iire IPEEE D.6. Severe Accident hianagement Guidelines Qualitathe D.7. Reactor Water Cleanup Automatic Isolation hiodincation Level 2 OCPRA in the " Applicable Risk Esaluation" column the following are used: OCPRA, Level 2 OCPRA. Seismic PRA, Fire indh idual Plant Examination for External Es ents (IPEEE), or Qualitative.
The OCPR A (reference 2) refers to the plant specific Level 1 PRA performed in response to the IPE generic letter.
The Level 2 OCPRA (reference 3) refers to the full scope Level 2 PRA perfonned in response to the IPE generic letter.
The Seismic PRA and the Fire IPEEE refer to the quantitatise evaluation performed in response to the IPE for [hternal Events analpis (reference 4). The Oyster Creek Fire IPEEE is a modined probabilistic risk assessment due to the use of a screening approach.
Detah.re available in reference 4 and Appendix C.
In the case where no existing risk analysis can be used in the detennination of the quantitative risk impact of the deferral of the project, then the term " Qualitative" is used in the " Applicable Risk Evaluation" column.
defened.0) doc IWO2 W
Lre sad Rhk Arciph Pere 6 or48 ncinhm o All es aluations pm md to determine the rhk impact of the defertal of projects are discussed in summary in the following rep i section and in detail in Appendix 11.
3.2 Step 2. I'.5aluate the Safety or Risk impact 1his report section prosides a summary of the methods and results of the determination of the safety /rkk impact of the deferral of projects. Details on the specinc evaluations are available in Appendix D of this l
reput.
3.2.1 Generic 14tter 96-06 Modifications Perfonn the proposed modi 0 cations in response to Generic Letter 96-06 during the 18R refueling outage.
The generic letter questions the operability of systems with regard to their capability to withstand ambient heating following a loss of coolant accident (LOCA). Preliminary analpis indicates that two of the three issues contained in the generic letter do not apply to Oyster Creek (reference 6). The third issue, containment penetration overpressurization due to ambient heating following isolation during a LOCA applies to Opter Creck. Without overpressure protection, the concern is that entrapped water between the intmard and outboard isolation valves is heated, expands, and increases in pressure challenging the strength of the particular penetration.
Operability determinations hase been performed indicating that all sptems considered susceptible to oserpressure are considered operable for the interim duration until either procedural chanym and or hardware modi 0 cations can be made (reference 6,7,8). OpU has committed to perform correcthe actions which involve phpical modi 0 cations to the plant be documented in the integrated Schedule for Opter Creek, pursuant to license condition 2.C.(6) of the full Term Operating License.
'the analpis of the risk impact of the deferral of the 96-06 modi 0 cations until the 18R outage is performed using insights desetoped in the Level I and Level 2 OCpRAs. LOCAs which discharge to the drywell and result in core damage are adjusted to renect endstates which bypass the primary containment. Three sensitivity cases are evaluated to detennine the risk impact of project deferral. Case 1, evaluutes the risk impact if all LOCAs nhict discharge to the dryw ell airspace are assumed to fail the containment integrity.
Case 2, evaluates the risk impact if large I.OCAs which discharge to the dryweh airspace are assumed to fail the containment integrity. Case 3, evaluates the risk impact if $n of large LOCAs which dkcharge to the drywell airspace are assumed to fail the containment integrity.
Based on this analysis, the large early release frequency increase for case 3, is 5.5x10* per year, This equates to a 7.4's in the large early release frequency. Ds ails of the analysis and results are presented in Appenda D.
3.2.2 Scismie Qualification Modifications-Phase il The scope of the project is to implement modi 0 cations w hich address outliers resuking from Oyster Creck's unresobed safety issue (USI) A 46 Program which was perfonned in response to the NRC's Generic Letter 87 02. Seismic seri0 cation walkdowns perfonned utilizing SQUG methodology were conducted during 1994 (reference 24). Phase I modifications hase been completed.
1he evaluation for the risk impact of this modi 0 cation is performed using insights from the Seismic PRA performed in support of Generic Letter 88 20, Supplement 4 (reference 4). Modi 0catims on the core spray and containment spray anchorage and the platform in the southwest corner room are expected to be completed on schedule. Other SQUG modi 0 cations, with the exception of the diesel generator building roo' slabs, were evaluated in the Seismic PRA frag:1?y analysis "as-built" and therefore, do not signincantly affect risk.
derened-03 doc 10'Ov97
Stftty and Rid Anilpin hp 7 of 48 Rn ition o l
?
I 1hree sensitivity cases are evaluated with respect to the capacity of the diesel generator building roof.1he 1
first, models the capacity of the roof at 0.18g which provides a 50?k chance of building failure gigen the safe shutdown canhquake (SSE). The sscond and third cases, model a capacity of the diesel generator building roof at 0.36g and 0.54g, which correspoud to 2 and 3 times the SSE acceleration.1hese cases are l
based on the fact that seismically designed equipment typically has a capacity of 2 to 3 times design (i c.,
l SSE).
1 1he results for case 1,2 and 3 are core damage frequency increases of 2.2sl0,1.0x10+ and 3.6x10 per 4
4 year, respectively. This corresponds to a 61.196,28.696 and 9.996 increase in the Seisn ic PRA core damage frequency, respectively. Details of the analysis and results are presented in Appendix it 3.2.3 Controlltoom iluman Factors Design Resiew in summary, the work included within the scope of this project includes the upgrade of the human engineering of the control room back panels IR through $R and 6R through IIR (including 9XR) as well as llXR,12R,12XR,14R,14XR,llF and 16R (reference 23). The scope of work for each of the panels includes:
I 1.
Review of the panels by GPU lluman I actors 2.
Walldowns with P!mt Operations 3.
Development of three sets of dr9,ings of these panels (relabeling, label specifications, and final"as builts".
4.
Repainting. relabeling, and annuciator matched demarcation of these panels, including necessary cosmetic panel repairs (e g., sanding, hole filling, etc.)
To date, many plant changes have resulted in improved back panels. Since the initial control room human factors review, signl0 cant changes to the back panels have occurred. Con' al room panels hase been upgraded or replaced as a result of many recent plant modifications. Panel equipment is rep! aced or upgraded in accordance with current company standards which meet or exceed those of NUREG 0700, hiajor modifications include:
1.
hiain Generator Protection l'pgrade Project. This project afTected panels llR, llXR,12R and 12XR.
2.
Digital Feedwater and Digital Recirculation Control hiodification. This project affected panels 8R and 9R.
3.
Recirculation Flow Scram Ekctronics hiodification. This modification affected panels 3R and $R.
4.
Panel 2R was relabeled in accordance with the Back Panel L,,beling Project.
In addition to the above mentioned projects, other plant programs, initiatives and corrective actions base identified poorly or confusingly labeled equipment on the control room panels which have since been relabeled in accordance with company standards.
No quantitative figure of merit is available for the assessment of the risk impact of the deferral of this modification. The current plant specific PRAs are based on the current control room design. hiodifications dererred 03 doc 10/o2 H
hofct) ondithL Ancipls PopeRof4R Ites hion o to the back panels in the control room may increase the human' machine intstface, howes er it is likely that this will not significantly affect the probabilities of operator errors.
3.2.4 Anticipatory Scram flypass Logic Improsement l
This project was initiated to improse upon actions taken in response to LER 95-005 (reference 12). The actions taken to date include the reseting of PSil switches to conservative $ctroints. This conservatism is required, to assure that under certain plant configurations, where steam is redirected, that thennal power remains below 40% when these anticipatory SCRAM signals are bypassed. Ilecause the PSil switches tap oft the third stage extraction stearn lines, the operation of the switches is not a true indicator of reactor thermal power. They are a better indicator of turbine load.1hus the parameters monitored by tl e PSil switches are not indicative of total plant thermal power except during normal " full power" plant steam alignments.
The modification would replace the current PSil switches with more precise switches. In addition, local control switches and indicating lamps would be installed to provide indication when the PSil switches are closed. The new PSil switches will proside a permissive signal that will allow bypassing of the affected anticipatory SCRAM signals. Group annunciation of when the anticipatory scram bypass is pennitted will be provided to the control room.1his will allow return of the setpoints from the current 25% to the 40%
pow er level.
Currently, operators are not aware when the turbine stop valve closure and turbine control valve fast closurc scrams are bypassed (l.c., no control room or local indication). Lack of indication of when the scrams are bypassed results in lost generation due to unnecessarily low power reductions when turbine scrams must be bypassed (e g., grid work), in addition, without indicailon of the engaged scram signal bypass, operators could assume that the scram is engaged when in fact it is not, resulting in an inadvertent scram.
The risk of deferring this project from the 17R to the 18R refueling outage is estimated using the insights and results of the Level i OCPRA. Smee, not performing the modification in the 17R refueling outage could result in the potential for an inadvertent scram (reference 11) and the safety significance of the non-consenative setpoint is considered minimal (reference 12), the turbine trip frequency is increased by one turbine trip over the operating cycle. Details to the risk evahtation are contained in Appendix 11, 3.2.5 Thermo Lag Fire llarrier Modifications The scope of this project is to install modifications to bring the Thenno-Lag 330-1 fire barrier systems installed at Oyster Creek into compliance with 10CFR$0 Appendix R.
The NRC has raised several concerns as to adequacy of these systems and has issued information notice 92 46, ilulletins 92-01, and 92-01, Supplement I and Generic Letter 92 08 on this subject. The NRC now considers the fire rating of these systems to be indetenninate and is requiring compensatory measures (l.c., fire watches) until the issue is resolved.
'lhe risk impact of altering the current completion date of the Thermo-Lag project is estimated using the Oyster Creek Individur ' Plant Examination for External Events (reference 4). Specifically, the risk impact is estimated using the fire Individual Plant Examination (IPEEE).
The core damage frequency estimates produced in the Oyster Creek fire IPEEE do not model the effect of the Thermo-Lag fire barriers therefore. upgrade of the Thermo-Lag barriers would sene to reduce the current estimates of the core damage frequency. Iloweser, due to the high combustible loading as well as the importance of the 480 VAC system in the mitigation of fire events at Oyster Creek the modifications of the 480 VAC Switchgear Rooms are scheduled to proceed as planned. Details of the evaluation are presented in Appendix 11.
derened-03 doc 10/02M
Scich cod Rhk Anal >un Paec v cus Ret hion o 3.2.6 Sesere Accident Management Guideline Deselopment Desclopment and implementation of the Sescre Accident Management Guidelines (SAMos)is currently scheduled to be completed by December 1998. This submittal presents the risk impact of the deferral of des clopment and implementation of the SA MGs until December of 2000.1he risk impact of this deferral is estimated qualitatisely since actions directed by the SAMGs are not modeled in any existing risk evC ations and have not been completely developed.1he guidance for coping with severe accidents is expected to provide limited risk benc0t over the two > car period remaining between the completion of the guidance and the potential closure date of Oyster Creek (l'all 2000). On this basis, the deferral of the implementation of the Severe Accident Management Guideline is assigned to the low risk category.
3.2,7 Reactor Water Cleanup Leakage Monitoring. LOCA Detection and Isolatior.
The purpose of these modincations is two fold. l'irst, the installation of thennocouples on the discharge d relief valves in the reactor water cleanup (RWCU) system to allow for the easy determination ofleakage by operations or maintenance (reference 17). Second, the installation of temperaturc sensors at the entrance to the reactor water cleanup recirculating pump room to detect leaks which constitute a LOCA on the high pressure portions of she RWCU system (reference 18).
The installation of thennocouples on the discharge of the relief valves in the RWCU system to allow easy detennination of leakage by operatwo or maintenance is being perfonned for "improsed radiological conditions" and dose reduction.1he risk impact of the deferral of this modincation is considered low based on judgement and the fact that this modihcation was initially proposed for the purposes of does reduction and convenience.
The installation of temperature sensors to detect leaks in the RWCU which constitute a LOCA on the high pressure portions of the RWCU system is being performed in response to the long term correctise actions for GE Nuclear SIL 604 (reference 19) and Oyster Creek Deviation Report 961097 (reference 20).1he concern of the GE SIL is that at certain power lesels below 10096, automatic isolation of the cleanup system on low reactor water level may not occur due to the capacity of the feedwater system to maintain level despite ins entory losses out the break.
The risk impact is determined using insights from the Level I and Level 2 OCpRAs. The risk impact is detennined by denning an new initiating esent representing a break in the RWCU line.1he initiating event frequency is based on the probability of the RWCU pipe break including the failure to isolate probability (i e., operator response) to the event.1he initiating event impact conservatively assumes the failure of the all equipment in the reactor building due to the harsh environment following the failure to isolate the break, in this fashion the increase in core damage frequency is estimated. This initiating event results in the bypass of the primary containment Therefore, the large early release frequency is also estimated using insights from the Lesel 2 OCpRA.
4 The increase in core damage frequency due to a RWCU line break in the high pressure sections is 1.2x10 4
per year. This equates to a 3.296 increase. The large early release frequency is calculated to be 1.2x10 per year, the same as the core damage frequency increase. The percent increase in the large early release frequency is 16'6. Details of the analysis and results are presented in Appendix B.
3.3 Step 3. Categorlie Safety / Risk Impacts This report section provides the categorization of the safety / risk impact for each of the proposed project deferments. The risk impacts are categorized into either high, medium or low categories according to the increase in total core damage or large early release frequency as denned on Tables I and 2, below. In the case where the risk impact categoritation differs between the core damage frequency and the large early release frequencies, the higher of the two categorizations is assigned.
ddened-03 dos 10'02/97
)
hrr0 and Rhk Antipin l' age 10cf 48 Rn hkm 0 In the case where non quantitathe results have been used to assess the risk impact, the assignment of the j
risk category is based on judgement. for the quantitative assessments, additmnal details on the assignment of the risk rategory are provided in the detailed analysis Appendix 11. For the non quantitative assessment, details are provided in this report section.
I Table ! -Categoritation of Risk impacts j
Affecting Core Damage Frequency l
Risk Core Damage Frequency j
Category Percent increase Range 4
liigh - liigh 1000?b i
j liigh 10096 100096 j
Medium -
109. 10046 i -
Low
<1096 Table 2 -Categoritation of Large 3
Early Release Frequency increa6es l
4 Risk Range a
Category (Percent increase) liigh
>10046 i
_ Medium 1096 - 10096 i
f _^ Low
<10?b j
Table 1 Categoritation of Risk Impacts Affecting Core Damage Frequency, is derived from the Oyster l
Creek On Line Maintenance Risk Management Procedure (reference 16) and,in part, from the LPHI PRA J
Applications Guide (reference 15). Table 2 - Categorization of Large Early Release Frequency increases, is j
derived, in part, from the EPRI PRA Applications Guide which indicates that increases in large early release frequency for Oyster Creek of greater than 36.4% are signincant and require additional analysis.
Using the criteria in Tables i and 2 above, the risk impact is given in Table 3 and discussed in the i
following paragraphs.
Generie Letter 96-06 Mod 10 cations. The Generic Letter 96-06 Modi 0 cations do not affect the core i
damage frequency since the concern is the merpressure of piping penetrations following loss of coolant 1
accidents. This overpressure is conservatively assumed to result in a failure of the piping penetration such j
that the primary containment integrity is compromised, if this assumption is applied to all losses of coolant the increase in the large early release frequency is 34.3% Small LOCAs which discharge to the drywell 1
may not result in the same environment (l.c., slower containment and piping heatup) versus large LOCAs.
If small LOCAs which discharge to the drywell are excluded, the increase in the large early release frequency increase becomes 14.8?&. Ilowever, if the probability of GL 96 06 related pipe break given a
{
large LOCA is not unity (1.0) then the risk impact would be less. The same is true if the probability of containment integrity failure is not unity (1.0) given a LOCA (e.g., specinc break location (s)). Assuming a j
GL 96 06 pipe break occurs in only 50% of the LOCA cases, the risk increase is less than 1096 in large i
carly release frequency, Using Table 2, the increase is then cctegorized as Low.
i SQUG - Seismic Qualineation Modineations. The SQUG phase I modi 0 cations have been completed.
j Signincant phase il modi 0 cations 'are recommended for completion, including the core spray and containment spray pump anchorages and the platform in the southwest corner room. The remaining
]
modi 0 cations, with the exception of the diesel generator building roof slabs, were included in the Seismic
]
PRA as "as built", Therefore, the current Seismic PRA includes the capacity of these components. The deferred 03 doc l&o2m
'. - - - -. -.. - - - - -. -. ~. - - -. -... -- -. -
~
. - -. - -,.. -. ~,
Acfcty cridlusk Anolpin Pcgeilordt untsk n o I
diesel guerator building roof slab anchorage was not evaluated "as built". Reducing the capacity of the building results in an increase in the seismic core damage frequency.
To evaluati the risk impet of the diesel generator building roof, three cases are evaluated.1he first case, reduces the capacity of the diesel generator building to 0.18 g re0ccting a 50?6 chance of building failure giver, the tsfe shutdown carthquake (SSE). The second and third cases renect higher capacities of 0.36g l_
and 0.54g which it equal to the 2 and 3 times the SSE, respectively, lhe cases are based on the i
compensatory measure of verifying the diesel generator building roof anchorage for the SSE. Typically, seismically designed equipment has a mean fragility of 2 to 3 times the SSE acceleration.1he cose damage frequency estimates in case 3 indicate a Low category. Case 2 has an increase in core damage frequency equal to lx10* w hich is traditionally considered low.
Control Room iluman Factors I)esign Reslew (Hack Panels). The contro) toom human factors design review for the back panels does not significantly alTect either the core damage or large early release frequencies. Many changes have been completed since the proposal of the project. The remaining modi 0 cations are not expected to signincantly improse the human / machine interface. As such, the deferral of this project is assigned to the Low category based on judgement.
Anticipatory Scram legie Modineation. The risk imps ct of the anticipatory scram logic modincation is assessed using the Level 1 OCPRA. Deferring the mod'.tcation is assumed to result in a turbine trip over the operating cycle for which it is deferred. This rotts in an increase in the core damage frequency of 6.896. Tbc impact on the large early release frequency is not calculated since turbine trip events typically result in.ntact containment endstates. Using Table 1, the risk impact of the deferral of this mod 10 cation is assigned to the low category.
1hermo Lag Fire Harrier Modifications. The risk impact of the deferral of the lhermo-lag modi 0 cations is estimated using the Fire IPELE. The l' ire IPELE did not model the effect of the lhermo-Lag Orc barriers. As such, upgrade of the barriers would serve to lower the existing core damage frequency estimates. All Dre zones which contain 1hermo Lag were screened in the Ore analysis (i.e., core damage 4
frequency less than lx10 per year) with the single exception of the "A" 480 VAC Switchgear Room.
Given the importance of the 480 VAC Switchgear to the mitigation of transients at Oyster Creek, it is recommended that the Thermo-Lag replacement for both "A" and "B" 480 VAC Switchgear Rooms proceed as originally planned. The remaining Ore zones containing 1hermo-Lag were screened in the Ore analysis (i.e., core damage frequency less than lx10* per year). Therefore, these Ore tones are considered leu risk signincant. On this basis, the risk impact of the deferral of the Thermo-Lag Upgrade, excluding the 4R0 VAC Sw itchgear Rooms, is categorited as Low.
Severe Accident Management Guidelines. The deferral of the Sesere Accident Management Guidelines does not signincantly affect the core damage frequency or large early release frequency.1he PRAs rencet the plant design, maintenance and operations practices at the time of their development and do not include the affect of the guidelines. The Severe Accident Management Guidelines may provide limited benent over the two year period between the completion of the guidance and the potential closure date of Oy ster Creek (Fall 2000). On this basis, the deferral of the implementation of the Severe Accident Management Guideline is assigned to the Low risk category.
Reactor Water Cleanup Automatic isolation Modification. The risk impact of the deferral of the Reactor Water Cleanup Automatic tr.olation Modi 0 cation is estimated using the Level I and Level 2 OCPRAs. Estimation of the frequency and impacts of a break in the high pressure portion of the RWCU piping indicate that the large early release frequency increases considerably, Given the signincant increase in the large early release frequency as well as the potential for a bypass of the primary containment, it is recommended that this modincation be implemented as originally planned. The risk impact of this deterred 03 doc 10/01W
hefety chd Risk Anclysis Page 12 of 4s Resiston 0 modification is assigned to the high category based on the increase in the large early release frequency and degradation to a signincant fission product boundary.
Table 3 - Risk Impact Categories of Ptajut Deferrals Project Title CIW LEHl' Percentage Hisk increase increase Increase Category Generic Letter 96-06 Modifications "
None 5.5x10 7.4? &
Low i
4 4
SQUO - Seismic Quatincation Mod 10 cations 3.6x10 *-
N/A 9.996 Low
- Control Room fluman Factors Design Small N/A N/A Low Review (llack Panels)'4 4
4 Anticipatory Scram Logic Mod 10 cation 2.6x 10 N/A 6.896 Low i
Thermo Lag Fire liarrier Modi 0 cations **
Small"'
N/A-
.N/A Low
- Severe Accide.it Management Guidelines
- Small Small N/A Low 4
4 Reactor ",ater Cleanup Automatic Isolation 1.2x10 1.2x10
-1696
- Medium 4
4 Modification - Case I (Case 2 in brackets) *
(8.0x10 )
(8.0x10 )
(106?6)
(lligh *)
- 1. Unce the ecsults from scnuuut) caw 3 which are deemed to bcst reprocnt the riA impact
- 2. Case 3 and a risk category ofI ow is displa)cd. A low categor) is assigned due to compematory measurcs.
- 3. Qualitathcly euenned 4 ITct sdes the 480 VAC Switchycar Rooms which are to be performcd as scheduled
$. Percentage increase is in terms of the large early release frequency, f>
Recommended to be complete on schedule due to degradatlon of signincant ftssion pniduct barrier and relatis cl> high 1.I Rli 3.4 Step 4. Evaluate the Integrated Safety / Risk Impact The evaluation of the integrated risk impact is perfonned in two steps. The Orst step involves the detennination of whether the six projects for deferral are independent. That is, does the risk impact of the deferral of the projects have dependencies which innuence the overall risk impact in a non linear fashion.
For example, if two or more projects were to affect a fission product barrier, a single project may have a low risk impact while the combined affect of the two or more projects could have a signincant or high risk impact. Dependencies.re uncovered through the review of the projects to determine if the deferral of the project:
Impacts the same system, structure or component (SSC)
AfTects the same safety function Afrects the same fission product boundary or Reduces the margin of safety for multiple accidents (e.g., external, internal or shutdow n esents)
A review of the detailed risk evaluations indicate that there are no dependency issues. In the absence of any risk impact dependencies, the risk impact can be calculated using simple addition of quantitative risk impacts Since the evaluation contains qualitative assessments as well as conservative quantitative results-and, different Ogures of merit (i.e., core damage frequency and large early release frequency), judgement is used in the combination of the integrated risk impacts.
The total core damage frequency increase is 6.2x10# per year. The total large early release frequency is
$,$x10' per year. On this basis the integrated risk assessment would be considered low and therefore acceptable, The projects which were evaluated non quantitatively have small risk impacts either in the core damage or large early release frequency. In addition, the projects which are performed on the current schedule, either in total or in part, can present reductions in the total core damage frequency or large early release when compared with existing risk studies. Quantitative values have not been developed for these deferred-03 doc 1(r01H e-,..
..iw.
.w,,2vwm.-..,-..,.-cw.1,
.-,,~,-vwm,
- w. m w...~,.,
-..,,,,m,-.my.wm
0 hortt) and Rid Analph
~
Pcscl}of4R o
Res hion 0 risk reductions, howeser they can signincantly offset any risk increase. Therefore, the integrated assessment of the total risk impact of the defened projects is comidered low.
4.0 CONCI.USIONS This analysis evaluated the risk impact of the deferral of projects for a single cycle.1he risk impact was evaluated using, to the extent pmsible, the existing plant specine PRAs and risk evaluations. Where the risk impact could not be evaluated using the existing risk studies, minor changes were made to allow their use, or the risk impact was performed qualitatively. To assess the total risk impact of the deferral of the projects, individual as w ell as integrated risk impact evaluations w cre undertaken.
in addition to the risk impact evaluation, ar.cntion was paid to defense in depth and maintaining adequete safety margins as well as ensuring that the incremental change in risk was small. To this end, projects with signincant risk impact were recommended for implementation on the original schedule. Portions or elements of projects which contributed signl0cantly to the risk impact or which represented signincant decreases in the safety margin or defense in depth were also recommended for comple: ion as originally scheduled. Perfonning several projects on schedule, either in total or in part, can reduce risk when compared with existing risk evaluations and serve to offset, in part, the small risk increases.
Of the seven (7) projects originally scheduled for deferral only four (4) are deferred in their entirety. These are:
Generic 1.etter 96-06 hiodincations Control Room lluman Factors Design Review (llack Panels)
Anticipatory Scram Logic hiodi0 cation Severe Accident hianagement Guidelines.
a Of the remaining three (3) projects, one (1) is recommended for implementation as originally scheduled (Reactor Water Cleanup Automatic isolation hiodification) and two (2) are only deferred in part (SQUG -
Seismic Quali0 cation hiodincations and Thermo Lag Fire llarrier hiodi0 cations), With the irnplementation as planned of the projects and portions of projects which hase significant risk impact, the individual risk impact as well as the integrated risk impact remains acceptable.
derenede doc WO2M7
Sciety and Rid Analpin Page 14 of 48 Ret bion 0 5.0 kr.FERI:NCES 1.
Nuclear Regulatory Comrnission, Standard Review Plan, "Use of Probabilistic Risk Assessment in Plant Specine, Risk informed Decisionmaking: General Guidance" Draft for Comment, DraR SRP Chapter 19, Revision L, March 27,1997, 2.
GPU Nuclear Corporatiot, " Oyster Creek Probabilistic Risk Assessment (Level 1)",
Volumes I through 6, November 1991.
3.
GPU Nuclear Corporation, "Oyner Creek Probabilistic Risk Assessment (Level 2)",
Volume 1. June 1992.
4.
GPU Nuclear Corporation, " Oyster Creek Individual Plant Examination fo Lxternal Events", December 1995.
5.
Nuclear Regulatory Commission, " Generic Letter 96 006: Assurance of Eqt ipment Operability & Containment integrity During Design liasis Accident Cont.itions".
September 30,1996.
6.
GPU Nuclear Corporation, Letter 6730-97 2012. " Generic Letter No. 96 06, Dated September 30,1996,120 day Response", January 28,1997.
d 7.
GPU Nuclear Corporation, Letter 6730-97 2059, " Generic Letter No. 96-06, Dated September 30,1996,120-day Response, Updated" February 26,1997.
8.
GPU Nuclear Corporation, '.etter 6730-97 2145, " Generic Letter No. 96-06, Dated September 30,1996, Corrective Actions follow up to 120-day Response Letter", June 3, 1997.
9 GPU Nuclear Corporation, Technical Data Report, " Oyster Creek Applicability to Generic 1.etter 96-06", TDR 1210, Revision 1 March I8,1997.
10.
GPU Nuclear Corporation, Request for Project Approval, "Scismic Quali0 cation Modi 0 cations - Phase !!", P w get Activity Number 403092, Resision 0, August 7,1997.
11.
GPU Nuclear Corpration, Request for Project Approval, " Anticipatory Scram 13ypass Logic Improvement",ilA Number 400018, Revision 0, August 7,1997.
12 GPU Nuclear Corporatiot, Licensee Event Report, Oyster Creek Unit I, " Anticipatory Scram !!ypass Switches' Non Conservatis e Setpoint Due to Original Plant Design", LER 95 005, Revision 0, August 25,1995.
derened-03 doc t 0/02N7
Stftt) cnd Rid Antlpis l' esc 15 of 4ll Res hion 0 13.
GPU Nuclear Corporation, Request for Project Approsal, "Thermo L4, Fire llarrier ModlBeation", il A Number 403042, Resision 1. August ?,1997, 14.
GPU Nuclear Corporation, Fire Protection Program.
- Fire I; e Analysis Report".
Document Number 9901746, Revision 9 Volumes I and 2.
15.
Electric Power Research Institute (EPRI), "Probabilistic Safety Assessment Applications Guide", EPRI TR 105396, Final Report, August 1995.
16.
GPU Nuclear Corporation Oyster Creek Policy and Procedure Manual, *0n Line Maintenance Risk Management",2000-ADM 3022.01, Revision 4, September 5,1997.
17.
GPU Nuclear Corporation Request for Project Approval, " Cleanup System Leakage Monitoring", Activity Number 400294, Revision 0, August 7,1997.
18.
GPU Nuclear Corporation, Request for Project Approval, "RWCU LOCA Detect &
Isolate (SIL604)", iludget Activity Number 400017 Revision 0, August 7,1997.
19.
GE Nuclear Energy, Service Information Letter (SIL), " Reactor Water Clean up System litcak Detection" SIL No. 634, November 6,1996.
20.
GPU Nuclear Corporation, Deviation Report, " Deviation Report of RWCU Break Detection", DVR Number 961097, December 4,1996.
21.
PLG, incorporated," Database for Probabilistic Risk Assessment of Light Water Nuclear Power Plants (Failure Data)", PLG 0500, Volume 2, Revision 0, July 1989, 22.
PLG, incorporated, " Database for Probabilistic Risk Assessment of Light Water Nuclear Power Plants (Common Cause Failure)", PLG-0500, Volume 4, Revision 1. July 1989 23.
GPU Nuclear Corporation, Request for Project Approval," Control Room iluman factors Design Review", Budget Activity Number 328030, Revision 2, August 7,1997, 24 GPU Nuclear Corporation, Request for Project Approval, " Seismic Quali0 cation Modi 0 cations - Phase 11", Budget Activity Number 403092, Revision 0, August 7,1997.
25.
GPU Nuclear Corporation, GPU Nuclear Testing Procedure, " Control Room and Cable Spreading Room lleatup Test", Procedure 254/13 MTX 26.12.2.6.
26.
GPU Nuclear Corporation, GPU Nuclear Calculation, "PRA Control Room Loss of Ventilation Profile". Calculation C 1302 862 5360 010, Resision 0.
defetted43 doc 10/02N7 l
=
e 5:let) and Risk Andph Pcge Ifeof 4R Rct hiun u l
l i
l API'ENDIX A DEFFERRED l'ROJECT LIST defcned-03 doc 1WO197
O
.NRC Regulatory Commitments i
Defer Before Final Plant Decision l
i Commitment Regulatory c
Approach Generic Letter 96-06 Modifications - Pressure concerns for piping penetrations Integrated Schedule (BA #31G690, BA #320011)
Update SQUG - Seismic Qualification Modifications. NOTE: Significant number of Integrated Schedule modifications have been completed. (BA #403092)
Update Control Room Human Factors Design Review - Repaint, refurbish, and relabel Integrated Schedule control room panel 1R through 10R, 6XR,11XR,12R,12XR,14R,14XR,16R, Update I 11F,9XR and 11P. NUREG 0737, Supplement 1 (BA #328030)
Anticipatory Scrarn W c Modification LER 95-05 (BA #400018)
Integrated Schedule i
Update Severe Accident Management Program Generic Letter 88-20
" Individual Plant Integrated Schedule _
Examination for Severe Acciderit Vulnerabilities" Update Thermotag Fire Barrier Modifications 16 and 17R. NOTE: Modifications to Integrated Schedule 460V rooms will not be deferred. (BA #403042)
Update r
Reactor Water Clean Up - Provide an automatic RWCU system isolation on a Integrated Schedule line break - SIL 604 - LER %-015. (BA #40G294, BA #400017)
Update J
?5?
ts m
hafety and Rkk Ardpk I' cst 18 of 45 o
Reg bion 0 APPENillX 11 RISK IMPACT ANALYSIS
?
defened-03 doc 10/02/97
heten enJ Rid An:1pis Pere thi48 Heelon 0 This appendix prosides deinils on the risk impact analysis done in support of the project deferral. Only the projects evaluated for quantitative analysis are described here.
11.1 Generle Letter 96-06 Modifications 1he genetic letter questions the operability of sptems with regard to their capability to withstand ambient heating following a loss of coolant accident (LOCA). Preliminary analysis indicates that tuo of the three issues contained in the generic letter do not apply to Oyster Creck (reference 6).1he third issue has been detennined to apply to Oyster Creck.
Project Descript;on and Proposed Change lhe third issue is the oserpressuritation of containment penetrations due to ambient heating following isolation during a LOCA. Without overpressure protection, the concern is that entrapped water between the inboard and outboard isolation valves is heated, expands, and increases in pressure challenging the strength of the particular penetration,l'ive (5) penetrations require modincation to reliese oserpressure:
1.
Reactor fluilding Clossd Cooling Water (RilCCW) Return from the Drywell 2.
Shutdow n Cooling Supply to the Reactor 1
Isolation Condenser "A" Condensate Return 4.
Isolation Condenser "B" Condensate Return 5.
Rectreulation Sampling Operability determinations have been performed indicating that all systems considered susceptible to overpressure are operable for the interim deration until either procedural changes and'or hardware modi 0 cations can be made (reference 6,7,8). GPU has committed to perfonn corrective actions whkh involve physical modi 0 cations to the plant be documented in the integrated Schedule for Oyster Creek, pursuant to license condition 2.C.(6) of the l'ull Term Operating License.
Risk impact baluation 1he analysis of the risk impact of the deferral of the 96 06 modi 0 cations until the ISR outage is perfonned using insights developed in the Lesel 1 and Level 2 OCPRAs. The Generic Letter 96-06 is primerily concerned with the integrity of the containment following a LOCA. That is, the overpressure of containment penetrations resulting in failurs of the penetratior and containment integrity. The figure of merit or risk measure used in the determination of the risk impact of the 96 06 modi 0 cations is, therefore.
Large 12arly Release Frequency. Three cases are used in the estimate of the risk impact. The cases are ordered from most conservative to least consersative.
Sensitivity Case 1 in Case 1, the risk impact is estimated by assuming that all LOCAs which discharge to the drywell and result in core damage, also result in overpressurization and failure of a containment penetration. This includes the effect of small LOCAs een though small LOCAs would not result in the severe environmental conditions that occur during large LOCAs. The contribution of all LOCAs to the total core damage frequency is taken from the Level i OCPRA. Table B 1 provides the contribution of all LOCAs to the total core damage frequency.
Since it is assumed that piping overpressure results in the failure of a containment penetration, a large containment bypass is created. This is a conservatis e assumption since a large containment bypass requires either of the following to occur: (1) a single large pipe rupture at the containment penetration or (2) two pipe breaks w ith one inside the containment and another outside.
defencd 03 doc t O'02 W
Axiety and Rhk Anxlph l'tge 20of 4R i
l Res hkm 0 1he contribution of the LOCAs is normally an " containment intact" plant damage endstate. To model the assumed containment integrity failure, the nonnat plant damage endstate of" containment intact" is adjusted l
to large early release endstate.1his leads to increase in the total large early release frequency i
approximately equal to the core damage frequency of the LOCA contributions. A "Large Early Release frequency Worksheet"(Ll:Rf) is provided as lable 112 and displays the estimation of the increase in LLRf.
From Table 111, the frequency of all LOCAs with discharge to the primary containment airspace is equal to 2.59x10' per year. In Table 112, the Lesel l Key Plant Damage State. PIFW, whlch is a " containment intact" endstate,is reduced by the above LOCA frequency of 2.$9x10 per > car.
4 Key Plant Damage State PlFW (!!ase Case)- LOCA frequency Contribution - New PirW KPDS 4
4 d
Ll6x10 - 2.$9x10 - 8.98x10 j
4 The percent variance on Tabk 112, is then 8.98x 10 disided by the base case of 1.16x 10+ or.22%
Also on Table 112, the i.evel i Key Plant Damage State, MKCU, which is a large early release containment endstate,is increased by the above LOCA frequency of 2.59x10 per year.
Key Plant Damage State MRCU (base case) + LOCA Frequency Contribution ~ New MKCU KPDS 1.72x10 4 2.59x10' = 4.3:x10#
4 4
The percent variance on Table Il 2, is then 4.31x10 dhideJ by the base case of 1.72x10 or + l$l?6.
The changes to these key plant damage states results in an increase of the large early release frequency 4
4 4
from the base case of 7.56x10 per year to 1.02x10 per year or 2.59x10 per year. This corresponds to an increase in the large early release frequency of 34.396.
The Case i analysis of the risk impact of the deferral of the Generic Letter 96-06 modilications remains bounding due to the conservative assumptions regarding pipe rupture, pipe rupture locations as well as the assumption that small LOCAs result in the overpressurization of the susceptible containment penetrations.
Semitivity Caw 2 Case 2, a less conservative sensitivity case, is evaluated to estimate a less conservathe risk impact. This sensitivity case evaluates the risk impact assuming that the issue of piping oserpressure is restricted to the large LOCAs into the drywell airspace which result in core damage. That is, small 1.OCAs result in a less severe environment due to the slower heatup of the drywell. The slower heatup allows for the initiation of containment spray and/or the automatic depressurization system. The effect of the cooling of the containment spray system and use of the automatic depressurization s) stem to remove heat to the torus and results in less heat being discharged to the drywell. With less ambient heatup of the drywell and, therefore less ambient heatup of piping penetrations, it is less likely that piping failures due to overpressuritation will occur. Using the "Large LOCA with discharge to the drywell airspace" row from Table 111, the evaluation performed in case I abos e is repeated. 'the results are display ed on Table 114.
I in the sensitisity case, the increase in large early release frequency is 1.12x10" per year which corresponds to a 14.896 increase. This evaluation remains conservative due to assumptions with regard to assumed pipe breaks following exceeding code allowable stresses and the assumed break location (or multiple breaks) which fail containment integrity.
Semitivity Caw 3 in case 3 the risk impact is evaluated by assuming that piping overpressure is restricted to the large LOCAs into the drywell airspace which result in core damage and only 1096 of piping oserpressuritations result in defened Oldoc 10/02M
~._
hiftty cnd Rtd Analpin P:ge 21 of All Rni6 ion 0 a pipe break which fails containment isolation. This is reasonable assuming that ultimate failure pressures of pipes are typically signincantly higher than the design or code allowable pressures. I'or the total of fhe penetrations this is equal to 5 times 10?6, or a 50?k chance of containment integrity failure due to pipe overpressuritation.
'Ihc efTect of small LOCAs is also not considered in this case. As stated in the evaluation of case 2, small LOCAs result in a less t.evere environment due to the slower heatup of the dr>well. The slower heatup allows for the initiation of containment spray and'or the automatic depressurization system. The coonng effect of the containment spray system and use of the automatic depressurintion system to remo"e heat to the torus, results in less heat being discharged to the drywell. With less ambient heatup of the drywell and piping penetrations it is less likely that piping failures due to overpressuritation will occur. The frequency used is 50?b of the frequency in case 2.
In this sensitivity case, the increase in large early release frequency is 5.5x10 per Scar w hlch corresponds to a 7.4? 6 increase.
Results and Conclusions The results of three sensitivity cases used to evaluate the afTect of the deferral of Generic Letter 96-06 Modi 0 cations is displayed on Table 11.5, below. As the results indicate, the large early release frequency ranges from a percent increase of 7.4?s to 34.3?6.
Increases in the Large Early Release I requency are categorlied according to the criteria on 'lable 2 (found in the main report). The risk impact of the sensitivity cases range over risk categories of Low and Medium.
Ilased on judgement, case 3 is deemed to best represent the deferral of Generic Letter 96 06 Modi 0 cations and the risk impact is categorlied as low.
Table 115-Summary of Generic Letter 96-06 Risk Evaluations Case Description Large Early Release l'requency Risk increase percent Category Value increase Case I: All LOCA core damage 4
frequency contributions (which discharge 2.59x10 34.3?6 Medium to drywell airspace) result in containment integrity failure.
Case 2: All Large LOCA core damage frequency contributions (which discharge 1.12x10 I4.8?6 Medium to the dr> w ell airspace) result in containment integrity tailure.
Case 3: 50?. of 1.arge I.OCA core damage frequency contributions (which 5.5 x 10
7.4? b Low discharge to the drywell airspace) result in containment integrity failure, deterred-03 doc 10/02N7
,. _ _ ~ _..
Sofety r.*.iPbb Anoluis Page!!9.'4R e
Resision 0 TABLE B 1 OCPRA INITIATING EVENT IMPORTANCE MODEL Name: OCPRA 13 Initiator Contributions to End StMe Group : ALL Total Frequency for the Group = 3 7982E 06 Inibator Frequency Unaccounted Percent LOSP 1.24E 06 1.00E-09 32.73 %
TTRIP 4.64E 07 3.64E 09 12 23%
RT 2.84E 07 1.80E 09 7.48%
LOFW 2.60E 07 1.40E-09 6.85%
CMSIV 2.57E 07 3.18E-09 6.76 %
LOTB 1.48E-07 9.61E 10 3.90 %
LOCV 1.48E 07 2.69E 09 3 89 %
LOIS 1.22E 07 7.26E 10 3 21 %
EPRL 1.19E 07 2.54E 09 3 13%
LBIE _ M..
l_ _'
.1,00E 07 T
~; 8.30E 11!"
. J 2.87%.
LOFC 1.02E-07 2 69E 09 2.69%
SBE l~"
7
.f 9.46E-08 E
? 2.01E 10' J2;49%F IEMRV 9.04E 08 1.64E-09 2.38%
PLOFW 7.83E 08 1.89E 09 2.06 %
LBIO 7.65E-08 2.86E 11 2 01 %
sal'J
~
T; 5.24E-08 T
~ 2.13E-10T SBO 4.67E-08 2.25E 11 11.38%J 1.20 %
LOlA 3.15E 08 1.96E-09 0.63 %
EPRH 2.93E-08 1.52E-09 0.77 %
LOCW 2.15E-08 1.35E 09 0.57 %
IADS 1.69508 9.01E 11 0.42 %
SAOTB 2.28E 09 4.80E 11 0.06 %
LAICS
~
' 1.48E-09 5.36E 11-
' O.04 % 7 LAIMS 6, Li.37E-09 :
. 5.30E 11
_ 0.04 % 0_
m LAOMS 4.62E 10 3 36E 12 0.01 %
LACIC 3.99E 11 3.36E 12 0.00 %
SAORB 2.66E 11 1.03E 11 0.00 %
SAOIC 8.52E 12 1.24E 11 0.00 %
TOTALS 3 80E 06 2 98E-08 100 00%
L5fgG LWA5 Whit WI50MfgG L3 h W I' m
.1.12E 07 :-
1.9.E 10 :
. 2.94% =
~
gg
,--* =
- 2. E r:
8.
E.1.
.. 1 %
derened 03 dos tn'02M fr
Safety and RM Analpis
/ age 23 of 48 i
{
Resision o i
l Table B 2: 96 06 LERF Estimation Spreadsheet Reference Cue: Bane Case (Risk Model: OCPRA 131 l
Case under study: 96 06 LERF (Case 11
)
i
, Level 1 Initiating Events Level 1 Key Plant Damage States j
l.E. '
Value Reference Variance g input input Reference Percent j
CMSIV 4.17 E-01 4.17 E-01 0%
3 KPDS Value -
Base Case Variance l
EPRH 5.61E 02 5.61 E-02 0%
I PlF W 8.98E-07 1.16E 06 22 %
EPRL 1.76E-01
- 1. 76E-01 0%
l l ADS 1.33E-03 1.33E-03 0%
OlAU 5.75E-07 5.75E 07 0%
l IEMRV 3.31E 02 3.31E 02 0%
OJAU 1.83E 07 1.83E 07 0%
LAICS 8.21E 05 8.21 E-05 0%
MKCU 4,31 E-07 1.72E-07 151 %
l LAIMS 1.15E 04 1.15E 04 0%
MJAU 5.88E-08 5.88E 08 0%
l LAOIC 6.962 08 6.96E-08 0%
NJHW 1.56E-08 1.56E-08 0%
1 LAOMS 6.44E 08 6.44E-08 0%
Total CDF 3.80E-06 3.80E-06 0%
{
LBI 5.67E-04 5.67E-04 0%
LBIO 8.37E-06 8,37E-06 0%
LERF Estimation LOCV 2.24E-01 2.24E 01 0%
Percent of CDF Analyzed =
84.36 % '
LOCW 2,71 E 02 2.71 E-02 0%
Total analyzed frequency = 3.20E 06 f
[
LOFC 1,71 E 01 1,71 E-01 0%
- Category FA - Large Early j
LOFW 1,51 E-01 1.51 E-01 0%
MKCU 100 %
4.31 E-07 4.31E 07 LOlA 4.33E-02 4.33E-02 0%
NIFW 30.85 %
1,04E-06 3.22E-07 i
LOIS 7,51 E-03 7,51E 03 0%
F OIAU 0.95 %
5.75E 07 5.47E-09 l
Total 7.58E 07 i
LOTB 1.03E 02 1,03E 02 0%
[
Percent of Total Analyzed =
23.65% '
i PLOFW 1.78E-01 1,78E 01 0%
Category 18 - Containment Bypass j
RT 7,21 E-01 7,21 E-01 0%
OJAU 100 %
1.832-07 1.83E 07 3
0 MJAU 100 %
5,88E 08 5.88E-08 9 l
SAOIC 1,59E-06 1.59E-06 0%
NJHW 100 %
1.56E-08 1.56E 08 5
)
SAORB 7.70E-07 7,70E-07 0%
i Total 2.57E 07 j
SAOTB 3,64E 04 3.64E-04 0%
Percent of Total Analyzed =
8.03 %
SBl 7.81 E-03 7.81E 03 0%
Total LERF (sum of above) =
1.02E-06
Reference LERF =
7.56E-07 TTRIP 8.97E-01 8.97E-01 0%-
Percent Change in LERF =
34.26% g Percent of Total Analyzed =
31.69 %
EPRI PSA Applications Guide Delta for this Case CDF 5.40 %
0.00%
LERF 36.37 %
34.26 %
u
- ~, MBi, 'tG h5L'O4. --
.m%
Comments:
4 deferred 03 doc 10/02/47
. ~.. -
-. - ~
hafet) and Rhk Analpis l'aye 24 of 4R Rnision 0 4
d 1
I E I I
l l
Table B 3: 96 06 LERF Estimation Spreadsheet 4
Reference Case: Base Case (Risk Model: OCPRA 13)
Case under study: 96-06 LERF (Case 2) a D... > l l l
Level 1 Initiating Events f
Level 1 Key Plant Damage States J
1.E.
Value Reference Varian:o input input Reference Percent CMSIV 4.17E 01 4.17 E-01 0%
l !
KPDS Value Base Case Varianc,e EPRH 5.61E 02 5.61 E-02 0%
M PIFW 1.04E 06 1.16E-06
-10 %
' EPRL 1,76E 01 1.76E 01 0%
NIFW 1.04 E-06 1.04E-06 0%
jl LADS 1.33E-03 1.33E-03 0%
OIAU 5.75E-07 5.75E-07 0%
! IEMRV 3.31 E-02 3.31 E-02 OY OJAU 1.83E-07 1.83E-07 0%
- LAICS 8.21 E-05 8.21 E-05 0%
[
MKCU 2.84E-07 1.72E 07 65 %
- 1. AIMS 1.15E 04 1.15E-04 0%
h MJAU 5.88E-08 5.88E-08 0%
i LAOIC 6.96E 08 6.96E-08 0%
l NJHW 1.56E-08 1.56E-08 0%
- LAOMS 6.44E-08 6.44E-08 0%
Total CDF 3.80E-06 3.80E-06 0%
'- LBI 5.67E-04 5.67E-04 0%
- LBIO 8.37E-06 8.37E-06 0%
LERF Estimation i
,LOCV 2.24 E-01 2.24E 01 0%
{ [
Percent of CDF Analyzed =
84.35 % )
LOCW 2.71 E-02 2.71 E-02 0%
L-Total analyzed frequency = 3.20E-06 J
[LOFC 1.71 E-01 1.71 E 01 0%
$ Category 1A Large Early p
1 LOFW 1.51 E-01 1.51 E-01 0%
l 0 LOlA 4.33E-02 4.33E-02 0%
'l MKCU 100 %
2.84E-07 2.84E-07 (
NIFW 30.85 %
1.04E 06 3.22E-07 I
h LOIS 7.51 E-03 7.51 E-03 0%
OIAU 0.95 %
5.75E 07 5.47E 09 L mLOSP 3.26E-02 3.26E 02 0%
L Total 6.11 E-07 ;'
I LOTB 1.03E-02 1.03E-02 0%
II Percent of Total Analyzed =
19.07 % s b PLOFW 1.78E-01 1.78E-01 0%
l Category 1B - Containtnent Bypass a
[RT 7.21E 01 7.21 E-01 0%
l OJAU 100 %
1.83E 07 1.83E-07 il l
' S AI 9.27E-03 9.27E-03 0%
MJAU 100 %
5.88E-08 5.88E-08 y E SAOIC 1.59E-06 1.59E-06 0%
NJHW 100 %
1.56E-08 1.56E-08 (
JSAORB 7.70E-07 7.70E-07 0%
Total 2.5 7 E-07 y
. SAOTB 3.64E-04 3.64E-04 0%
Percent of Total Analyzed =
8.03%
l SBl 7.81 E-03 7.81 E-03 0%
Total LERF (sum of above) =
8.68E-07 i SBO 2.86E 06 2.86E-06 0%
Reference LERF =
7.56E-07 E TTRIP 8.97E-01 8.97E 01 0%
I' Percent Change in LERF = l 14.82% [N
[
Wggg Q4@ gMh Percent of Total Analyzed =
27.10% lN r e
- -- % i m 4 -w,__ _ f m & # 93 M S@ W E A
i
=u--> m eaemanwam
- ]
I Risk significant cutoffs:
Risk Significant Cutoff Delta for this Case f
CDF 51.40 %
0.00 %
Cin S I
LERF 36.37 %
14.82 %
[
QhPK,&W%WWe%l%WMeemLt%Ma UWu ' 7MMMM&W n % '
%WWVW:W &lW^.
h Comments:
[
h NewsmywusMFMwwhinqERWs "4N i
' w C.M m
W AA i
+ <
+*
+ "
v defened4)3 doe WO2!97
Safety and Rhk Analpis Page 23 o[48 t
Resision0 1
Table B-4: 96 06 LERF Estimation Spreedsheet 4
Reference Case: Base Case (Risk Model: OCPRA.13)
Case under study: 96 06 LERF (Case 3)
Level 1 Initiating Events Level 1 Key Plant Damage Stades
- 1. E.
Value Reference Variance E Input input Reference Percent CMSIV 4.17E 01 4.17E 01 0%
B KPDS Value Base Case Variance -
EPRH 5.61 E-02 5.61E 02 0%
PIFW 1.10E-06 1.16E 06 5%
l EPRL 1.76E 01 1.76E-01 0%
NIFW 1.04 E-06 1.04E-06 0%.
4 l
l ADS 1.33E 03 1.33E 03 0%
OIAU 5.75E 07 5.75E-07 0%
IEMRV 3.31 E-02 3.31E 02 0%
OJAU 1.83E 07 1.83E 07 0%
LAICS 8.21 E-05 8.21 E-05 0%
MKCU 2,28E 07 1.72 E-07 33 %
I LAIMS 1.105-04 1.15 E-04 0%
MJAU 5.88E 08 5.88E 08 0%
}
LAOIC 6.96E 08 6.96E 08 0%
NJHW 1.56E-08 1.56E 08 0%
l LAOMS 6.44E-08 6.44E 08 0%
Total CDP 3.80E-06 3.80E-06 0%
j LBI 5.67E 04 5.67 E-04 0%
l LBIO 8.37E 06 8.37E-06 0%
LERF Estimation l
LOCV 2.24E 01 2.24 E-01 0%
Percent of CDF Analyzed =
84.35 %
l LOCW 2.71E 02 2,71E 02 0%
Total analyzed frequency = 3.20E 06 LOFC 1.71 E-01 1.71 E-01 0%
Category FA Large Early 2
LOFW 1,51E 01 1.51E 01 0%
MKCU 100 %
2.28E-07 2.28E-07 j
LOlA 4.33E 02 4.33E 02 0%
NIFW 30.85 %
1.04E-06 3.22E-07 ~
j LOIS 7.51E 03 7.51 E-03 0%
i OIAU 0.95 %
5.75E 07 5.47E-09 i
LOSP 3.26E 02 3.26E 02 0%
Total 5.55 E-07 l
LOTB 1.03E 02 1.03E 02 0%
Percent of Total Analyzed =
17.32 %
l PLOFW 1.78E 01 1.78E 01 0%
- Category 18 Containment Bypass RT 7.21 E-01 7,21E 01 0%
OJAU 100 %
1.83E 07 1.83E-07 l
sal 9.27E-03 9.27E 03 0%
f MJAU 100 %
5.88E-08 5.88E 08 i
SAOIC 1.59E-06 1.59E 06 0%
NJHW 100 %
1,56 E-08 1.56E-08 l
SAORB 7.70E-07 7.70E 07 0%
Total 2.57E-07 SAOTB
. 3.64E 04 3.64E 04 0%
Percent of Total Analyzed =
8.03 %
l SBl 7.81 E-03 7.81E 03 0%
Total LERF (sum of above) =
8.12 E-07 l
SBO 2.86E 06 2.86E 06 0%
Reference LERF =
7.56E-07 i
TTRIP 8.97E 01 8.97E 01 0%
Percent Change in LERF =
7.41 % E 4
Perr:ent of Total Analyzed =
25.35 %
R Delta for this Case l
CDF 51.40 %
0.00 %
LERF 36.37 %
7.41 %
}
y ' ;.y..
'.,b o ;m; ;:\\..\\.
Comments:
)
--p==-
__ _ _ w wu,s.emm-wtwuemammvm.wm vem =" wwm=- c=wem v=_rr----- -
1 1
deferred-03 doc 10/02/97
.--t---r.
+
Strety and Rhi Andpis Page 26 of 48 e
Resision 0
!!.2 Seismic Qualification Modifications-Phase il The scope of the project is to implement modifications w hich address outliers resulting from Oyster Creek's unresolved safety issue (USI) A 46 Program which was performed in response to the NRC's Generic Letter 87-02. Seismic verincation walkdowns performed utilizing SQUG methodology were conducted during 1994 (reference 24). Phase i modifications have been completed.
Project Description and Change The specific work scope for this project includes modi 0 cation to the following:
1.
Anchorage of the Core Spray Main Pumps 2.
Anchorage of the Core Spray Booster Pump (P 20-002A) 3.
Anchorage of the Containment Spray Pumps (P-21-001 B, P-2100lc and P-2100l D) 4.
Anchorage for panet ER 661 100 (Turbine Building Ragems Panel) which is a missile hazard to two Safe Shutdown Equipment List (SSEL) components.
5.
Modify r~chorage for MCC 1 Al2 and 1812. MCC 1 Al2 is an SSEL component and MCC 18 2 is an interaction hazard for cabling for MCC I Al2.
6.
Modifications for the CRD liydraulic Control Units. The existing installed anchorage can be made adequate for the majority of the units provided their stiffness is improved. The individual unit which stands alone (11CU 305-34-51) will require a more signincant modification to provide additional ancho age.
7.
Provide new anchorage for the MSIV solenoid rack in the drywell, 8.
The platform supporting T 22 001 (reactor building equipment drain tank) in the southwest corner room is seismically inadequate. A modification is required to ensure that the core spray pumps and associated cabling is not jeopardized by platform failure.
9 Replacement of relays in the diesel generator control circuits, the rotary inverter control cabinet, ASCO Transfer Switches.
- 10. Modi 0 cation of the 41P VAC Switchgear circuitry.
I 1.
Modi 0 cation of the anchoinge for the diesel generator roof slabs.
Currently this work is scheduled for completion by the end of 1998. Consider the deferral of work until 18R refueling outage.
Risk Impact Evaluation The evaluation for the risk impact of this modification is performed using insights from the Seismic PRA performed in support of Generic Letter 88 20, Supplement 4.
Due to the timing of the IPEEE project and the seismic qualification of equipment (SQUG) project, fragility calculations performed in support of the IPEEE relied, to varying degrees, on the SQUG work packages. The varying degree of reliance on SQUG packages produces fragility calculations which range from those based on the SQUG packages alone (including modifications credited in the SQUG packages, if any) to those based on the seismic IPEEE fragility walkdowns. Where no SQUG package existed when seismic fragility walkdowns were performed, the plant equipment fragility was based on the "as-built" condition. Where a completed SQUG package existed, the fragility value was based on the SQUG package.
In this case the fragility would be based on any planned modifications, if the equipment was a SQUG outlier. If the equipment passed the SQUG evaluation, the SQUG package would be used in the development of the fragility.
The use of the SQUG packages in the fragility evaluation results in several fragilities, which are based on the successful completion of the SQUG modifications. Such is the case with items I,2,3 and 8 from the lis; of planned modifications. Given this fact, as well as the importance of the core spray and containment defened-03 doc 10iO2N7
Scfety aid Risk Ana1 sis 3
Pcge 27 of 40 Resision 0 spray systems in the mitigrtion of transient events at Oyster Creek, it is recommended that items 1,2,3 and 8 be implemented as planned.
SQUG modification items 4 and 5 are m:luded in the SQUG list to provide support for control room ventilation including the requirement for power supplies for recovery of ventilation using ponable fans.
Ventilation studies done is support of the OCPRA, including actual test data, indicate that control room ventilation is not required for a signincant period of time following its loss (reference 25). Therefore, control room ventilation was not required for success in the OCPRA or the Seismic PRA. On this basis, the risk impact of these SQUG modification isjudged to less significant.
SQUG modification items 6 through 10 (excluding item 8) were included in the Seismic PRA fragility analysis. Fragility analysis of these items was based on the "as-built" plant during the seismic fragility walkdowns. As such, these SQUG modifications are expected to have a limited affect on the risk associated with seismic events. On this basis, the deferral of these modification items is considered low.
The final item in the SQUG modification list is the modification of the diesel generator building roof slabs.
This modification is required due to corrosion of the existing anchorage. The fragility analysis did not explicitly include the roof anchorage of the diesel generator building roof slabs. Currently, the diesel generator building is considered operable. Periodic verification of the condition of the diesel generator building roof slab anchorage is performed.
In order to estimate the risk impact of the deferral of this modification, the capacity of diesel generators themselves (assuming that building failure during a seismic event fails both diesel generators) is adjusted.
Three case studies are performed.
In the first case (case 1), a fragility value (i.e., capacity) of 0.18g mean acceleration is used. This estimate provides a 50% probability that the diesel generator building fails during a safe shutdown earthquake (SSE). Two additional case studies are performed assuming that the diesel generator building roof higher capacity, in case 2, the diesel generator building roof is assumed to have a capacity of 2 times the SSE or 0.36g In case 3, the diesel generator building roofis assu aed to have a capacity of 3 times the SSE or 0.54g. These values are chosen since seismically designed systems or structures typically have capacities 2 to 3 times the design capacity. This assertion is logical since at the design acceleration equipment is likely to be successful and the capacity or fragility referred to in this document represent the mean failure accelerations. Table B 6, provides a description and the results of the evaluation.
l I
l Results and Conclusions t
The estimation of the risk impact of the Seismic Qualification Modifications - Phase 11 assumes that due to the importance of the core spray and containment spray systems in the mitigation of transients at Oyster Creek, these modifications (items I,2,3 and 8) proceed as originally scheduled. In addi ion, it is assumed t
that modifications which affect the control room ventilation (items 4 and 5) and those for which fragility evaluation were performed on the "as-built" plant (items 6 through 10, excluding item 8) are of low risk significance.
The core damage frequency increase ranges from 61,1% to 9.9% using assumed capacities of 0.18 to 0.54g for the diesel generator building roof slabs. Provided that the capacity of the diesel generator building roof is verified to meet design, then case 3 (0.54g capacity) is judged to best represent deferral of the Seismic Qualification Modifications. Case 3 is categorized as low based on the percent increase in core damage frequency. It should be noted that the increase in core damage frequency in case 2 is 1.0x10 per year 4
which is typically considered low.
Current compensatory measures include the periodic verification of the condition of the diesel generator building roof slab anchorage. A verification (i.e., testing) of the capacity of the anchorage should be performed to provide less conservative estimates of the true capacity of the anchors.
defened-03 doc 10/02/97 A
Scret) and Risk Ancipis Pege 28 of 48 Res tsion 0 Table 11-6 Sumniary of Seismic Qualification Modification (Phase II) Risk impacts Case Description Seismic EDG Roof Core Damage Acceleration Failure Frequency Percent (g)
Probability (increase)
Increase Case 1: Diesel Generator Building 0.007 - 0.26 1.62x10' 4
Roof Stabs Capacity Equal to 0.18g.
0.26 - 0.46 9.17x 10.i 5.8x10 61.1 %
(2.2x10 )
0.46 - 0.62 9.97x 10
O.62 - 0.82 1.00 Case 2: Diesel Generator Building 0.007 - 0.26 1.23 x 10
4 Roof Slabs Capacity Equal to 0.36g.
0.26 - 0.46 4.13x 10
WO 28 E 4
(1.0x10 )
0.46 - 0.62 8.74 x 10
O.62 - 0.82 9.83 x 10
Case 3: Diesel Generator Building 0.007 - 0.26 1.36x 10" (3.6x 10'j) 4.0x 10 9.9%
Roof Slabs Capacity Equal to 0.54g.
0 26 - 0.46 1.43 x 10.i 0.46 - 0.62 6.28x i O
O.62 - 0.82 9.08 x 10
deferred-03 doc 10/02/97
-)
Sclet) and Risk Ancipis Pcgc 290f 48 Reg ishm 0 11.3 Antleipatory Scram flypass Logie improsement his project was initiated to improve upon actions taken in response to LER 95-005. He actions taken to date include the reset PSil switches to conservative setpoints. This conservatism is required, with the existing plant configuration, to assure that under cenain plant configt. rations, where steam is redirected, that thermal power remains below 4096 when these anticipatory SCRAM signals are bypassed. Because the PSil switches tap off the third stage extraction steam lines, the operation of the switches is not a true indicator of reactor thennal power. The PSil switches are a better indicator of turbine load, Thus the parameters monitored by the PSil switches are not indicative of total plat thermal power except ocring normal " full pow er" plant steam alignments.
Project Description and Proposed Change The modification would replace the current PSil switches with more precise switches, with hysteresis suaicient to lesson contact bouncing and with narrower dead bands. In addition, auxiliary relays would b; installed with local control switches and indicating lamps at the turbine standard to provide indication w he n the PSil switches are closed. The new PSil switches will provide a pennissive signal that will allcw bypassing of the affected anticipatory SCRAM signals. Group annunciation of when the anticipatory senim bypass is permitted will be provided to the control room using existing spare wires to the control ror,m.
(Additional wiring will be required). This will allow return of the setpoints from the current 25% to the 40% power level.
Currently, operators are not aware when the turbine stop valve closure and turbine control valve fast closure scrams are bypassed (i.e., no control room or local indication). Lack of indication of when the scrams are oypassed results in lost generation due to unnecessarily low power reductions when turbine scrams musi be bypassed (e.g., grid work). In addition, without indication of the engaged scram signal j
bypass, operators could assume the scram is engaged when it it fact is not, resulting in an inadvertent scram.
Risk Impact Evaluation The risk of deferring this project from the 17R to the 18R refueling outage is estimated using the insights and results of the Level 1 OCPRA. Since, not performing the modification in the 17R refueling outage could result in the potential for an inadvertent scram (reference 11) and the safety significance of the non-conservative setpoint is considered minimal (reference 12), the turbine trip frequency is increased by one turbine trip over the operating cycle.
Table B-7, Dyster Creek initiating Event Contribution, provides the laitiating event contributions to total core damage frequency. The turbine trip initiating event frequency of 0.897 per year is adjusted to re0cct the potential for an additional turbine trip due to the deferral of the anticipatory scram modification for one cycle. That is, the turbine trip frequency is increase by an additional turbine trip each cycle (t / two years) or by 0.5 per > car for a new turbine trip frequency of 1.397 per year. The results of this model(risk model:
TERIP) are provided on Table B-8, TTRIP Model Initiating Event Contributions.
Results and Conclusion The total core damage frequency increases from 3.80x10 per year to 4.06x10* per year or 6.8%. The 4
turbine trip initiating event (TTRIP) increases in contribution from a 1296 contributor in the base case to an 18% contributor in the TTRIP risk model. The risk impact is categorized according to the ranges specified on Table 2 (main report section). These ranges are those used in the Risk Management of On Line Maintenance Program at Oyster Creek (reference 16). Since the increase in the total core damage frequency is 6.8%, the risk category is Low.
derened-03 doc I0/02M7
Safety andisk Analysis Pcge 30 cf 4R Resision 0 TABLE B 7: OCPRA INITIATING EVENT CONTRIBUTION MODEL Name: OCPRA 13 Initiator Contributions to End State Group : ALL Total Frequency for the Group = 3.7982E-06 initiator Frequency Unaccounted Percent LOSP 1.24 E-06 1.00E-09 32.73 %
TTRIP 4.64E-07 3.64 E-09 12.23 %
RT 2.84E-07 1.80E-09 7.48%
LOFW 2.60E-07 1.40E-09 6.85 %
CMSIV 2.57E-07 3.18E-09 6.76 %
LOTB 1.48E-07 9.61 E-10 3.90 %
LOCV 1.48E-07 2.69E-09 3.89 %
LOIS 1.22E-07 7.26E 10 3.21 %
EPRI.
1.19E-07 2.54E-09 3.13%
LBI 1.09E-07 8.30E 11 2.87 %
LOFC 1.02E-07 2.69E-09 2.69%
SBl 9.46E-08 2.01E 10 2.49%
IEMRV 9.04E-08 1.64E-09 2.38 %
PLOFW 7.83E-08 1.89E-09 2.06 %
LBIO 7.65E-08 2.86E-11 2.01 %
sal 5.24E-08 2.13E 10 1.38 %
SBO 4.57E-08 2.25E-11 1.20 %
LOlA 3.15E-08 1.96E-09 0.83 %
EPRH 2,93E-08 1.52E-09 0.77 %
LOCW 2.15E 08 1.35E-09 0.57 %
IADS 1.59E-08 9.01E 11 0.42 %
SAOTB 2.28E-09 4.89E 11 0.06 %
LAICS 1.48E-09 5.36E-11 0.04 %
LAIMS 1.37E 09 5.30E-11 0.04 %
LAOMS 4.62E-10 3.36E-12 0.01 %
LAOIC 3.99E 11 3.36E-12 0.00 %
SAORB 2.66E-11 1.02E-11 0.00%
SAOIC 8.52E-12 1.24E-11 0.00 %
TOTALS 3.80E-06 2.98E-08 100.00%
deferred-03 doc 10/02/97
Sifsty and Rkk Andpis Page 31 of 48 Rn blon 0 TABLE B-8: TTRIP MODEL INITIATING EVENT CONTRIBUTION Model Name: TTRIP Initiator Contributions to End State Group: ALL Total Frequency = 4.0572E-06 Initiator Frequency Unaccounted Percent LOSP 1.24 E-06 1.00E-09 30.64 %
TTRIP 7.24 E-07 4.12E-09 17.83%
RT 2.84E-07 1.80E-09 7.00 %
LOFW 2.60E 07 1.40E-09 6.42 %
CMSIV 2.57E-07 3.18E-09 6.33 %
LOTB 1.48E-07 9.61E 10 3.65%
LOCV 1.48E-07 2.69E-09 3.65%
LOIS 1.22 E-07 7.26E 10 3.00 %
EPRL 1.19E-07 2.54E 09 2.93%
LBI 1.09E-07 8.30E-11 2.68 %
LOFC 1.02 E-07 2.69E-09 2.52 %
SBl 9.46E-08 2.01E 10 2.33%
IEMRV 9.04E-08 1.64E-09 2.23%
PLOFW 7.83E-08 1.89E-09 1.93%
i LBIO 7.65E-08 2.86E-11 1.89%
SAI 5.24E-08 2.13E-10 1.29 %
SBO 4.57E-08 2.2SE-11 1.13%
LOIA 3.15E-08 1.96E-09 0.78 %
EPRH 2.93E-08 1.52 E-09 0.72 %
LOCW 2.15E-08 1.35E-09 0.53 %
IADS 1.59E-08 9.01 E-11 0.39 %
SAOTB 2.28E-09 4.89E 11 0.06 %
LAiCS 1.48E-09 5.36E-11 0.04 %
LAIMS 1.37E-09 5.30E-11 0.03 %
LAOMS 4.62 E-10 3.36E-12 0 01 %
LAOIC 3.99E-11 3.36E-12 0.00 %
SAORB 2.66E-11 1.03E 11 0.00 %
SAOIC 8.52E-12 1.24E-11 0.00 %
TOTALS 4.06E 06 3.03E 08 100 %
deferredMdoc 10/02/97
_____A
Sciety and Risk Analysis Page 32 of 48 Reshion0 D.l Thermo-Lag Fire Harrier Modifications lhe scope of this project was to install modifications to bring the Thenno-Lag 3301 fire barrier systems installed at Oyster Creek into compliance with 10CFR50 Appsndix R.
Project Description and Proposed Change The fire barriers will be upgraded by overlaying the existing Thermo-Lag 3301 with fire banier material from another vendor. If the plant configuration does not have suf0cient space for the additional material on a specific fire barrier, the Thermo-Lag will be removed and replaced with new fire barrier material. If any power cable cannot accept the additional capacity de rating from the application of additional material, the Thermo-Lag will be removed and replaced with new fire barrier material or the cable rerouted to achieve an acceptable configuration.
The NRC has raised several concems as to adequacy of these systems and has issued infonnation notice 92 46, Bulletins 92 01, and 92-01, Supplement I and Generic Letter 92-08 on this subject The NRC now considers the fire rating of these systems to be indeterminate and is requiring compensatory measures (i e.,
fire watches) until the issue is resolved.
Risk Impact Esaluation The nsk impact of altering the current completion date of the Thenno-Lag project is estimated using the Oyster Creek Individual Plant Examination for External Events (reference 4). Specifically, the risk impact is estimated using the Fire Individual Plant Examination (IPEEE).
The Oyster Creek Fire IPEEE methodology is a modified PRA methodology which uses an iterative screening approach to remove from detailed evaluation (screen) those plant fire areas and zones which 4
present low risk (i.e., less than lx10 per year core damage frequency). More detailed analysis is then performed on those fire areas and zones which do not screen (i.e., core damage frequency greater than
!x 10* per year). As stated above the approach is iterative in nature and involves three steps:
- 1. Tne first step involves the "all engulfing fire" in which the Fire IPEEE models the failure of all equipment within and cables which trensit a given fire zone. in addition, all failure modes are addressed including " hot shorts" and a conservative transient model is chosen (e.g., all EMRVs open for pressure relicf which requires all EMRVs to reclose).
- 2. For those fire zones whose core damage frequency does not screen (i.e., is not less than lx10* per year) a second iteration is performed. In the second iteration, " revised core damagefrequency estimate", the assumptions used in the development of the risk model as well as simple recoveries (such as sentilation restoration) are credited. Fire suppression probabilities, both manual and automatic are modeled only in more detailed evaluations.
- 3. For those fire zones whose core damage frequency does not screen (i.e., is not less than lx10* per year) in the second iteration, a third and final iteration is performed. In the third iteration, the " detailed core damagefrequency estimate", automatic fire suppression probabilities are modeled as well as factors concerning fire growth and propagation.
dererred 03 doc t0/02M7
.. - - - ~ ~. - ~. _. -. ~ ~ ~
- -. - -_..- - ~. ~. ~ _
Safety cnd Rhk Analysis Page 33 of 40 Res klon 0 Using the above approach to the quanti 0 cation of the core damage frequency due to fire events results in j
two Orc zones w hich do not screen. These fire zones are the Cable Spreading Room' and the "A" 480 VAC Switchgear Room. The core damage frequency produced by the initial and revised estimates of core j
damage frequency (i.e., the first and second iterations) remain upper bound estimates since detailed evaluation is not performed for these areas. It is likely that a detailed evaluation would result in lower i;
estimations cf the core damage frequency for the screened fire zones. As such, it is not appropriate to make
]
judgements with regards to core damage frequency or the risk importance ranking of an individual fire zone which was screened without addressing the conservative assumptions made in the analysis. That is, 4
ccmparisons between screened Hre zones with core damage frequencies of, for example,7x10 per year 4
and 3x10 may not be valid without investigating the various assumptions performed in the analysis.
4 Thermo-Lag protection of circuits was not modeled in the Fire iPEEE. That is, circuits protected by Dermo-Lag were as3umed to fait due to the fire event in all iterations performed to evaluate the core damage frequency. A more complete overview of the methodology used in the development of the Fire IPEEE is presented in Appendix C.
Seven (7) fire zones at Oyster Creek use Thermo-Lag to provide a fire barrier. These fire areas are presented in Table 11.9 below.
Table H.9-Summary of Oyster Creels Fire Zones Containing Thermo-Lag Fire Area / Zone Combustible Loading Core Damage Designator Description HTUs / Sq. Ft.
Rating Frequency OB-FZ-06A Office Bldg "A" 480 VAC Swgr Room 176601 liigh 5.!E 6 OB FZ-06B ' Office Bldg "B" 480 VAC Swgr Room 142101 liigh 3.lE 7
- RB-FZ-01 D Reactor Building 51' Elevation 20362 Low 2.7E 7 '"
RB-FZ Olb Reactor iluilding 23' Elevation 24117 Low l.3 E 7 '"
RB-FZ-Ol F2 Reactor Building -( 19') Elevation 964 Low 9.0E-7 ""
TB FZ-1IC Turbine Bldg. Swgr Rm, West End 13575 Low 4.6E-7
- T B-FZ-1 I D Turbine Building - Basement South End 35163 Low 2.lE-7
- Notes:
(a)- Fire zone screened in initial evaluation assuming "all enguinng Gre".
(b)- Fire zone screened in a revised evaluation including refined risk modeling.
(c)- Fire zone was screened following detailed evaluation including application of Gre severity factor, fire detection and suppression.
Results and Conclusion ne core damage frequency estimates produced in the Oyster Creek Fire IPEEE do not model the effect of the Thermo-Lag fire barriers therefore, upgrade of the Thermo-Lag barriers would serve to reduce the current estimates of the core damage frequency. Ilowever, due to the high combustible loading as well as the importance of the 480 VAC system in the mitigation of fire events at Oyster Creek, the modifications of the 480 VAC Switchgear Rooms are scheduled to proceed as planned.
Based on the fact that modeling of the Thermo-Lag fire barriers could result in lower core damage frequency estimations, the risk impact of the deferral of the project, with the exclusion of the 480 VAC
- Switchgear Rooms, is assigned a low category.
The cable spreading room did not screen in the detailed evaluation. This fact is provided for completeness. The cable spreading room does not contain circuits protected by Thermo-Lag.
defenni-03 doc 10/02/97
. -. - -. _. - -.. - ~ _ _ _ - - - - -
-. ~ - - - -
- - ~. - - - - -
e-1
$sfety and Risk Analpis
. Pcge 34 or 18 Revision 0
. H.5 Reactor Water Cleanup Leakage Monitoring, LOCA Detection and Isolation -
The purpose of these modi 0 cations is two-fold. First, the installation of thermocouples on tiie discharge of relief valves in the reactor water cleanup (RWCU) system to allow for the easy determination ofleakage by opemtions or maintenance (reference 17). Second, the installation of temperature sensors at the entrance to j.
the reactor water cleanup recirculating pump room to detect leaks which constitute a LOCA on the high pressure portions of the RWCU system (reference 18).
l Project Description and Proposed Change ne installation of thermocouples on the discharge of the relief valves in the RWCU system to allow easy determination of leakage by operation or maintenance is being performed for " improved radiological conditions" and dose reduction. The risk impact of the deferral of Jiis modi 0 cations is considered low -
based on judgement and the fact that this modi 0 cation was initially proposed for the purposes of dose reduction and convenience.
The installation of temperature sensors to detect leaks in the RWCU which constitute a LOCA on the high pressure portions of the RWCU system is being performed in response to the long term corrective actions for GE Nuclear SIL 604 (reference 19) and Oyster Creek Deviation Report 96-1097 (reference 20). The concern of the GE SIL is that at certain power levels below 100% automatic isolation of the cleanup system on low reactor water level may not occur due to the capacity of the feedwater system to maintain level despite inventory losses out the break Operator action to isolate the break is assumed not to occur for 10 minutes in licensing analysis. Thus, the mass release may be greater than previously analyzed. Concerns on the affect of additional mass release on the environmental quali0 cation of equipment as well as radiological consegaences have arisen. The scope of this modifkation is to install temperature sensors outside the Reactor Water Cleanup Room. These temperature sensors would be used to generate an isolation signal upon indication of a LOCA that would isolate V 161, V 16-2, V-16-14 and V 1641, Actions taken in response to the deviation report include: EQ Evaluations for Potentially AtTected Safety Related Components, Additional Operator Guidance (Alarm Response Procedures), Additional Operator Training, and a Safety Evaluation. Planned actions include the design and implementatica Automatic isolation Modification.
Risk Impact Evaluation The risk impact of the deferral of the proposed modincation is determined using insights from the Level I and Level 2 OCPRAs. The Level 1 OCPRA mcdels a large number of loss of coolant including those outside the primary containment. However, a large loss of coolant from below the reactor core and outside the containment (in the high pressure section of the RWCU system) was not originally modeled. The low pressure section of the RWCU system was modeled in the Interfacing Systems LOCA (ISLOCA) analysis (Appendix B.3 of the Level 1 OCPRA). The failure of the low pressure section of the RWCU system was thought to be dominant and therefore breaks in the high pressure sections were not addressed. The estimate of the risk impact of a break in the high pressure piping of the RWCU, therefore, requires an additional initiating event.
The frequency of this initiating event is determined on Table B.10. Generic data for the failure of piping sections (reference 21) is multiplying the number of hours in a year and by the approximate number of-piping sections between the primary containment and the pressure control valve.
Pipe Break No. of No. of Pipe RWCU LOCA a
Probability hours / year Sections Annual Frequency (in sections / hour) deferred.o3 doc 10/02N7
~_
~
Safety and Rhk Andysis Pagc 35 of 48 Res blon 0 The calculation of the initiating event frequency also includes the failure to isolate probability based on operator response to the event and the failure of the motor operated isolation valves to close on demand.
The individual motor operated valve failure as well as the common mode failures of the valves is modeled.
4 Generic data is.ased for both the individual valve failure as well as the common cause failure probabilities (reference 22). The result of the addition of the operator error probability and the mechanical failure of the motor operated valves is the failure probability to isolate the RWCU line break.
Ope ator Isolates
+
MOVs Fail to Close on Failure to isolate
=
RWCU Break Demand RWCU Pipe Break (including common causes)
De failure to isolate combined (multiplied) with pipe break frequency results in the initiating event for the Unisolated RWCU line break.
4 RWCU LOCA Annual Failure to isolate Unisolated RWCU Line Frequency RWCU Pipe Break Break
}
The initiating event impact conservatively assumes the failure of the all equipment in the reactor building due to the harsh environment following the failure to isolate the break.2 This assumption is conservative since short term early operation of the core spray system will most likely occur. The feedwater system 4
injects into the downcomer of the reactor vessel and since the break is below the reactor core, it will exit through the break without providing core cooling. With the failure of the core spray system (failed as a result of the initiating event impact), core damage is assumed to occur. (With the short term operation of the core spray system, parallel injection valves will be open, allowing injection of the fire protecti?n system using manual valves located outside the reactor building wall.)
It is not necessary to exercise the Level. "PRA since with a large break below the reactor corr. feedwater cannot provide suf0cient cooling inventory That is, the fecdwater system injects into the downcomer of the reactor vessei and will exit the break without providing core cooling. With the failure of the core spray system (failed t s a result of the initiating event impact), core damage is assumed to occur.
Since this initiat ng event also results in the bypass of the primary containment, the large early release fraction is estimated, insights from the Level 2 OCPRA are used to develop Table B.ll, RWCU Line Break LERF Estimation Spreadsheet, in the estimation of the large sarly release frequency increase, the initiating event frequer cy (which is equal to the core damage frequency increase) is assigned to a containment bypass endstate (i.e., designator:
OJAU). The increase in large early release is therefore equal to the initiating event frequency and core damage frequency increase.
4 The increase in core damage frequency due to RWCU line break of the high pressure sections is 8.0x10 4
per year. This equates to a 21.1% increase. The large early release frequency is calculated to be 8.0x10 per year, the same as the core damage frequency increase. The percent increase in the large early release frequency is 106.0%.
From Table I (main report), Categorization of Risk Impacts Affecting Core Damage Frequency, a 21.1%
increase in the core damage frequency corresponds to a category of Medium. Ilowever, the categorization 2
Successful isolation of a RWCU pipe break is assumed to result in an isolation transient which is not modeled due to the low probability of occurrence compared with other isolation transients.
Following successful isolation, no other equipment failures are expected due to initiator.
defened-03. doc 10/02/97
Safety and Risk Analysis Page 36 of 48 -
Resision 0 of the increase in the large early release frequency, from Table 2 (main report) results in a risk impact category ofliigh based on a 106% increase in LERF, Sensitivity Studies it should be noted, that the results above are dominated by the operator action to isolate the RWCU system.
It is therefore prudent to perform sensitivity cases on the issue wi.h attention to the dominant contributor.
The operator action failure rate used in the initial case is lx10.'2 This vr.lue is censervative given the changes made to alarm response procedures and emphasis on operator response training. The sensitivity case (case 2) uses an operator failure rate of lxiO to estimate the frequency of RWCU Unisolated Pipe Breaks and impacts on core damage and large early release frequency. The estimation of core damage frequency and large early release frequency is performc',s above and presented as Case 2 on Table B-ll-and Table B 12.
Case 2 results in a core damage frequency increase of 3.2% and a large early release frequency increase of 16.3%. Base on core damage frequency the risk impact category is low ano based on large early release frequency increase the risk impact category is medium.
Results and Conclusions The risk impact of the deferral of the RWCU LOCA Detection and Isolation Modi 0 cation is presented in Table 13, below. The results indicate that although the core damage frequercy increase remains relatively low the large early release frequency could experience a significant increase. Using conservative values in the evaluation of the risk impact of the RWCU unisolated break, the increase in large early release frequency -is 106%. With less conservative values the large early release frequency increase is approximately 16%. The risk impact is dominated by operator action error rates.
Based on the degradation of the a signincant fission product barrier, and the significant increase in the large early release fraction it is recommended that the RWCU system modi 0 cation to install LOCA detection and monitoring proceed as originally scheduled.
Table B.13 -
RWCU Unisolated Pipe Break Core Damage and Large Early Release Frequency Results Case Description Core Damage Frequency Large Early Release -
Frequency Value Percent Value Percent increase Increase l
Case I: Unisolated Pipe Break, Operator Action 8.0x 10
21.1%
8.0x 10
106 %
Equal to 0.01 Case 2: Unisolated Pipe
- Break, Operator Action 1.2x 10
3.2%
1.2x 10
16 %
Equal to 0.001 i
defcned-Oldoc 10/02/97
~
.-n_
~.. -.
Safety and Rhk Ar alpls Pege 37 of 48 Reg ision 0 TABLE B 10:
ESTIMATION OF THE PROBABILITY OF OF AN UNISOLATED RWCU LINE BREAK l l RWCU LOCA FREQUENCY ESTIMATION
- No.
Event Description g'erence Case 1 Case 2 q
enenc Rpe BreA Nguency (per 1
8.60E 10 8.60E 10 h
A.1 section per hours) g A.2 Number of Hours in a Year n/a 8.76E+03 8.76E+03 i Annual Pipe Break Frequency (per pipe A.3
= A.1
- A.2 7.53E-06 7.53E-06 section)
I A.4 Estimated No. of Pipe Sections 2
10 10 l
A.5 Total F% Break Frequenc
= A.2
- A.4 7.53E 05 7.53E-05 qw.<
.M annw wasut ISOLATION PROBABILITY ESTIMATION No.
Event Desenption Reference Case 1 Case 2 B.1 Operator Isolate RWCU Break 3
1.00E-02 1.00E 03 j
B.2 bingle MOV Operates on Demand 1
4.30E-03 4.30E-03 J
Beta Factor for Two MOVs Fail t B.3 4
7.00E-02 7.00E 02 Operate on Demand (generic) 0 Two MOVs Fallto Operate on Demand
- B.4
= (B.2 - (1 - B.3) ^ 2 1.60E 05 1.60E-05 (non-common cause)
Two MOVs Failto Operate on Demand B.5
= B.2
- B 3 3.01E-04 3.01E-04 j
(common cause)
B.6 sc arge s Fahre
= B.4 + B.5 3.17E-04 3.17E-04 to Isolate on Demand l
nr scarge B.7
= B.6
- 2 6.34E-04 6.34E-04 k
MOVs Fail to isolate on Demand W
B.8 Failure to isolate Probabill
= B.1 + B.7 1.06E-02 1.63 E-03
-& wc; ms E umtpainstatsrum.e uA:qtutw atunaw m C.1
= A.5
- B.8 8.01E-07 1.23E 07 FREQUENCY ws. M am n w >m mm-=--- mnum o
References:
- 1. ' PLG, incorporated, " Database for Probabilistic Risk Assessment of Light Water Nuclear y
Power Plants (Failure Data)", PLG-0500, Volume 2, Revision O. July 1989.
]
2.
GPU Drawings. BR-2143, BR 2144, BR M565, and 3E-215-A2-1001.
3.
Estimated based on remote (convol room) action which procedura'ized and trained with approximately 10 minutes for completion.
4.
PLG. Incorporated. " Database for Probabilistic Risk Assessment of Light Water Nuclear Power Plants (Common Cause Failure)". PLG-0500, Volume 4. Revision 1, July 1989.
namessvetKLWAMm4hmwaw;4 mao--"MWM6 -
- r n %NyMTWOC "M M deferred-03 doc 10/02S 7
Safety and Risk Analpis l' age 38 of 40 Reusion 0 Table B 11: RWCU LERF Estimation Spreadsheet Reference Case: Base Case (Risk Model: OCPRA 13) l Case under study: RWCU Line Break (CASJ 1) l j
Level 1 Initiating Events bjy Level 1 Key Plant Damage States
[-
- 1. t:.
Value Reference Variance j input input Reference Percent {
CMSIV 4.17E-01 4.17 E-01 0%
B KPDS Value Base Case Variance $
EPRH 5.61E 02 5.61 E-02 0%
PIF W 1.16E-06 1.16E-06 0%
EPRL 1.76E-01 1.76E 01 0%
NIFW 1.04 E-06 1.04E-06 0%
LADS 1.33E-03 1.33E 03 0%
OIAU 5.75E 07 5.75E-07 0%
~
IEMRV 3.31 E-02 3.31 E-02 0%
OJAU 9.84E 07 1.83 E-07 438 %
LAICS 8.21 E-05 8.21 E-05 0%
MKCU 1.72E 07 1.72E-07 0%
, LAIMS 1.15E 04 1.15E-04 0%
MJAU 5.88E 08 5.88E 08 0%
c LAOIC 6.96E-08 6.96E-08 0%
NJHW 1.56E 08 1.56E 08 0%
i LAOMS 6.44E-08 6.44E-08 0%
Total CDF 4.60E-06 3.80E-06 0%
k LBI 5.67E-04 5.67E-04 0%
BjB M M %isissiii C -MMMTMT56.-i,-=:q LBIO 8.37E-06 8.37E 06 0%
kl LERF Estimation E
- LOCV 2.24 E-01 2.24E-01 0%
N Percent of CDF Analyzed =
87.08 % [
LOCW 2.71 E-02 2.71 E-02 0%
Total analyzed frequency = 4.00E-06 V
- LOFC 1.71 E-01 1.71E 01 0%
Category TA - Large Early L 0FW 1.51 E-01 1.51 E-01 0%
MKCU 100 %
1.72E-07 1.72E-07 LOlA 4.33E-02 4.33E-02 0%
NIFW 30.85 %
1.04E-06 3.22E-07 LOIS 7.51E 03 7.51E 03 0%
OIAU 0.95 %
5.75E-07 5.47E-09 h LOSP 3.26E-02 3.26E-02 0%
Total 4.99E-07 j LOTB 1.03E-02 1.03E-02 0%
Percent of Total Analyzed =
12.46 %
PLOFW 1.78E-01 1.78 E-01 0%
Category 18 Containment Bypass g, RT 7.21 E-01 7.21 E-01 0%
OJAU 100 %
9.84E-07 9.84E-07
'i j SAI 9.27E-03 9.27E-03 0%
MJAU 100 %
5.88E-08 5.88E-08 h SAOiC 1.59E-06 1.59E-06 0%
NJHW 100 %
1.56E-08 1.56E-08 l SAOR8 7.70E 07 7.70E-07 0%
Total 1.06E 06 l SAOTB 3.64E-04 3.64E-04 0%
Percent of Total Analyzed =
26.43% b SBI 7.81 E-03 7.81E 03 0%
Total LERF (sum of above) =
1.56E-06.
SBO 2.86E-06 2.86E-06 0%
Reference LERF =
7.56E-07 -'
TTRIP 8.97E-01 8.97E-01 0%
Percent Change in LERF = l 105.93% k b !
jgjg;ggd84jfNg/jQgM fd p
Percent of Total Analyzed =
38.88 % [
sainmerenwe 4b wm v-m m#vm m
- ul
]
{3
[ Risk significant cutof fs:
l Risk Significant Cutof f Delta for this Case
- p..
CDF 51.40 %
21.09 %
j-3E LERF 36.37 %
105.93 %
}
f-k.r m m w
- d d i J
wm m ew
~
w 4
% Comments:
High Risk Significance M
N he; ORA %&
W i
r
+
n >
s-r
~
~-:
-~?
~
deferred-03 doc 10/02!97
hafet) and Risk Analpis Pare 39 of 40 e
Retision u Reference Case: Base Case (Risk Model: OCPRA 13)
Case under study: RWCU Line Break (CASE 2)
Level 1 Initiating Events Level 1 Key Plant Damage States E
1.E.
Value Reference Variance E input input Reference Percent CMSIV 4.17 E-01 4.17 E-01 0%
B KPDS Value Case Variance EPRH 5.61 E-02 5.61E 02 0%
PlF W 1.16E 06 1.16E-06 0%
EPRL 1.76 E-01 1.76E 01 0%
NIFW 1.04E-06 1.04E-06 0%
LADS 1.33E 03 1.33E-03 0%
OIAU 5.75E 07 5.75E-07 0%
IEMRV 3.31 E-02 3.31 E-02 0%
OJAU 3.06E-07 1.83E 07 67 %
LAICS 8.21 F-05 8.21 E-05 0%
MKCU 1.72E 07 1.72E-07 0%
LAIMS 1.15E-04 1.15E 04 0%
MJAU 5.88E-08 5.88E-08 0%
LAOIC 6.96E-08 6.96E-08 0%
NJHW 1.56E-08 1.56E-08 0%
LAOMS 6.44 E-08 6.44E 08 0%
Total CDF 3.92E-06 3.80E-06 3%
LBt 5.67E 04 5.67E-04 0%
LBIO 8.37E-06 8.37E-06 0%
LERF Estimation LOCV 2.24 E-01 2.24E 01 0%
l Percent of CDF Analyzed =
84.84 %
LOCW 2.71 E-02 2.71 E-02 0%
q Total analyzed frequency = 3.33E 06 LOFC 1.71 E-01 1.71 E-01 0%
Category /4 Large Early l
LOFW 1,51 E-01 1.51 E-01 0%
MKCU 100 %
1.72E 07 1.72E-07 l
LOlA 4.33E 02 4.33E 02 0%
NIFW 30.85 %
1.04 E-06 3.22E-07 i
LOIS 7.51E 03 7.51 E-03 0%
OlAU 0.95 %
5.75E 07 5.47E-09 LOSP 3.26E-02 3.26E 02 0%
Total 4.99E-07 g LOTB 1.03E-02 1.03E 02 0%
Percent of Total Analyzed =
14.99% B PLOFW 1.78E-01 1.78 E-01 0%
Category 1B Containment Bypass 2
RT 7.21E 01 7.21 E-01 0%
OJAU 100 %
3.06E-07 3.06E-07 ;
sal 9.27E 03 9.27E 03 0%
MJAU 100 %
5.88E-08 5.88E-08 ;:
SAOIC 1.59E 06 1.59E-06 0%
NJHW 100 %
1.56E 08 1.56E-08.-
SAORB 7,70E-07 7.70E-07 0%
Total 3.80E-07 T SAOTB 3.64E 04 3.64E-04 0%
.z Percent of Total Analyzed =
11.43 % 3 SBI 7.81 E-03 7.81 E-03 0%
F Total LERF (sum of above) =
8.79E-07 SBO 2.86E-06 2.86E 06 0%
Reference LERF =
7.56E-07 TTRIP 8.97E-01 8.97E-01 0%
Percent Change in LERF = l 16.27% j$
Percent of Total Analyzed =
26.43 %
r EPRI PSA Applications Guide I'+
J Risk significant cutoffs:
Risk Significant Cutoff Delta for this Case t
CDF 51.40 %
3.24 %
$[
LERF 36.37 %
16.27 %
m s==_s-mnww nwawammmmmd, a hp-2 1 := h e. w;; m a na m wesim Comments:
Less Conservative Human Action Velue Medium Risk Significance meshwwwa+wrm9MswAWamufummiish he - h
( gn@w/de FN 't Fc rma.,
v.m AT deferred-03 doc
{
10/02/97
A
='
Sikty and Rbk Andysis Page 40 cf 48 Res hion 0 APPENDIX C FIRE INDIVIDUAL PLANT EXAMINATION METIIODOLOGY OVERVIEW deferred-03 doc 10/02M7
Safety and Rid Antlysis Page 41 or48 Resision 0 4.0 Oyster Creek Fire Individual Plant Examination The Oyster Creek Fire Individual Plant Examination report presents the methods and results of the fire analysis of the impacts of fire events at the Oyster Creek Nuclear Generating Station (OCNGS),
The study is performed in response to the Nuclear Regulatory Commission's (NRC) Generic Letter 88-20, Supplement 4, " Individual Plant Examination of Extemal Events (IPEEE) for Severe Accident Vulnerabilities" The analysis satisfies the requirement for the Internal Fire Analysis and presents the methods, calculations and results in the suggested NUREG 1407 format.
The analysis is performed using standard probabilistic methods sim; tar to those used in the development of a Level 1 Probabilistic Risk Assessment with several notable sxceptions.
First, all accident sequences developed in this study are initiated by fire events which are internal to the plant.
Second, a cutoff in the frequency of core damage is used to screen fire areas and hence the study is termed the Oyster Creek Individual Plant Examination or a scoping studying. One of the outcomes of using a screening approach is that the core damage frequency reported represents an upper bound since a more detailed evaluation would result in lower core damage contributions of individual fire areas. Thist approach to ar.alyzing interna!!y initisted fire events is a less resource intensive effort while still providing assurance that plant specific vulnerabilities,if any, are determined.
Third, sigriificant portions of the Electric Power Research Institutes (EPRI) Fire Induced Vulnerabuity Evaluation (FIVE) methods are used in the study, The study is comprised of ten tasks and results in the evaluation of the risk of internal fires at the Oyster Creek Nuclear Generating Station (OCNGS). The process is described in overview in the following paragraphs and illustrated in Figure 41, it should be noted that since the report is organized in the suggested NUREG 1407 format, multiple tasks are often documented in a single report section or sub-section. Each task or group of tasks as illus,trated on Figure 4-1 is described as well as illustrated in the associated figure to the right.
Task 1 - Develop Fire Initiating Event Frequencies This task identifies areas of the Oyster Creek Nuclear Generating Station in which the potential for fire initiation, growth and/or propagation can significantly impact plant operation from at power conditions.
The input to this task is from the Level 1 Oyster Creek Probabilistic Risk Assessment (OCPRA),
the Fire Hazard Analysis Report, Oyster Creek Fire Mitigation Procedure and plant walkdowns.
sace 4-1 wwm deferredMdoc 10/01H
~-.
. ~..
W hure 4-1 Oysis-Creek Fire Individual Plant Examination Process Illustration 5
R A
identify Components l
Fire Area
\\
Task 2
\\
Task 3 Mitigation
(
Potential Task 6 iterations _
A Y
Y t
Y Develop Develop Develop and Detailed Presentation Fire initiating Fire initiating Quantify y
Fire of
"' A
Results I
Event Event Fire Propagation A>
Frequencies impact Table Plant Model
>1E4 f Analysis Task 1 Task 4 Task 5 h
Task 7 Task 9 A
I l
l no
,O I_
Y y
Evaluation Fire Growth ya d
and
{_
Multiple Results Propagation Area Fire?
/
andIssues p'
Task 8 Task 10 g
s 5
2 Y
xi s
'f i e
s-t 8
e-5 0
c$ &
40 FIGS.XLS 4-2 10r495
$arcty and Risk Analysis 4
Page 43 or48 e
Revision 0 i
in this task the Fire Hazard Analysis Report fire area and zone designations are used with the OCPRA, fire mitigation procedure I
4 Develop and plant walkdowns to determine plant areas in which a fire Fire initiating l
l event may perturb plant operation sufficiently to result in a Event
+
l demand for a reactor scram. The Electric Power Research l
Frequencies 4
Institutes (EPRI) Fire Induced Vulnerabliity Evaluation (FIVE)
Task 1 l
r6ethodology and database is then used to develop fire initiating event frequencies for each of the identified fire areas and zones (critical fire areas).
Figure 4 2 4
Develop Fire initiating Events Several fire areas are screened from further consideration based on the insignificant impact of the fire event. For example, the Site Emergency Building does not contain any plant equipment and is located some distance from the plant and, as such, a fire in tNs area is not expected to result in a demand for a plant trip or damage to plant equipment.
The list of fire areas and zones with their frequency of fire ignition serves as input to the development of Task 4 (Development of the Fire Initiating Event impact Table) and Task 5 (Developmeni and Quantification of the Plant Model). Details on this task are presented in report Section 4.1, Fire Hazard Analysis.
Task 2 - Identification of Risk Significant Components in this task the Level 1 OCPRA, the Fire Hazard Analysis Report (FHAR) and plant walkdowns as well as the list of critical fire areas (from Task 1) are used to determine the potential risk significant components. These components are then screened on the basis of their susceptibility to fire events. The result is the list of risk significant components. Details on this task are presented in report Section 4.4, Evaluation of Component Fragilities and Failure Modes.
Task 3 - Identification of Risk Significant Component Locations in trils task, reviews of plant information and plant walkdowns are used to determine the location -
of the risk significant components and their supporting cables.
Supporting cables include any required electrical or other functional system support cables.
Supporting cables also include the possibility of component failure due to " hot shorts" which cause the component to go to an active failure position. That is, sypporting cables include those cables whose electrical hot short (i.e., energized) can result in a component changing state inta an undasired state or position. For example, a normally closed valve changing to the open pos; ton due to a fire event which affects the cable in a remote location of the plant.
The result of this task is the Location of Risk Significant Components and Associated Cables Table, which is used in the Development of the Fire !nitiating Event Impact Table (Task 4).
Several fire areas are screened from further consideration since they contain no risk significant compionents or support,ng cables. These areas are screened from further consideration for their individual contribution to the core damage frequency however they are still considered for their asce 4-3 tomm dererred 01 doc 10/02/97
Safety end Rhk An:Jpis Ptge 44 or48 Res kion 0 potentid to be involved in multiple area fires (Task 8).
Identify The information developed in this task is used Identify Component as input for the development of the Fire components Locations initiating Event Impact Table (Task 4) and the Task 2 Task 3 Fire Growth and Propagation (Task 8).
Details on Task 3 are presented in report
/
Section 4.4, Evsluation of Component Fragilities and Failure Modes.
Y Develop Task 4 Develop Fire initiating Event Fire Initiating impact Table Event
+
Impact Table in this task the Location of the Risk Significant Task 4 Components and Associated Cables (Task 3) and the OCPRA are used to develop the Fire y
initiating Event impact Table.
Figure 4 3 Each critical fire area, as defined in Task 1, is identification of Risk Significant Components considered an initiating event.
Using the
)
physical component locations and the locations of supporting cables a five event impact table can be developed. This impact table provides the affected components (an hence system functions) given an "all engulfing fire" within a fire area. The term "all engulfing fire"is used to describe the modeling of a fire which falls all components and cables in the area and does not account for detection, suppression or other area mitigative features. In addition, " hot short" impacts are included in the impact table. The Fire Initiating Event impact Table is therefore the most conservative impacts which a fire e ent within a given fire area can cause.
The impact table is used as input into the Development and Quantification of the Fire' Risk Model (Task 5). Details on Task 4 are presented in report Section 4.2, Review of Plant Information and
- Walkdowns.
l l
Task 5 - Development and Quantification of the Plant Model This task develops and documents the Oyster Creek Fire Risk Model. Actually three sub-tasks are performed in the development and quantification of the fire risk model and these sub-tasks are represented on Figure 4-4 as three separate paths of input and output. All three sub-tasks are documented in report Section 4.6, Analysis of Plant Systems, Sequences and Plant Response.
The first input / output path develops the individual fire area upper bound core damage frequency estimations with input from Tasks 1 and 4 and is described in the " Initial Estimate of Upper Bound Core Damage Frequency" report sub-section.
sece 4-4 uwm deferred-03 doc 10/02/97
,O Sciety and RhL Ana!pis e
Page 45 or43 o
Regision 0 The second input / output path is represented as the iteration loop between the Detailed Fire Propagation Analysis (Task 7) and develops fne refined core damage 4
frequency estimates for those fire areas whose upper bound core damage 4
frequency (UBCDF) was initially greater than 1x10. A single !terations is mado which results in the calculation of the Revised Estimate of Upper Bound Core Damage Frequency. Any fire areas which are not screened (UBCDF less than n
1x10 ) are analyzed in Task 7 and documented in the " Detailed Evaluation of Core Damage Frequency" report sub-se,ction.
The third input / output path develops the " multiple fire area" upper bound core damage frequency estimations.
T Each input / output paths is discussed in detail below, f
initial Estimate of Upper Bound Core Damage Frequency (UBCDF)
The first input / output path used the Level 1 OCPRA, Fire Initiating Event Frequencies (Task 1) and the Initiating Event impact Table (Task 4) to develop and quantify the fire risk model for the
" Initial Estimation of the Upper Bound Core Damage Frequency" as a result of fire events within an individual fire area.
The impacts of a fire event (Task 4) together with the fire initiating event frequency (Task-
- 1) are combined with the Develop and random failure probabilities of Quantify y..
Fire Area system functions modeled in Fire CDF >1E 6 the Level 1 OCPRA to prcduce Plant Model the fire risk model. That is, the Task 5 s
failures produced by the fire initiating event are added to the OCPRA plant model (the I
independent failures) to produce a risk model which Figure 4-4a Input / Output Path 1 initial Estimation of Upper Bound Core Damage Frequency calculates the core damage frequency due to fire events.
Since the fire initiating event impact table represents the most conservative outcome of a fire in a given fire area (i.e., "all engulfing fire" and " hot shorts") and fire growth, propagation, detection, suppression or other fire area mitigative features are not modeled, the quantification of this fire risk model produces an upper bound core damage frequency for each fire event. Fire areas 4
whose UBCDF is less than 1x10 per year are screened from further consideration. Fire areas 4
whose total UBCDF contribution is greater than 1x10 per year require a Revised Estimate of Upper Bound Core Damage Frequency which is performed as part of the input / output path two, described below.
xc.o 45 mas dercrred 03 doc 10/02/97
Sarcty and Risk Analpis Page 46 or48 Revision 0 Revised Estimate of Upper Bound Core Damage Frequency in the second input / output path the fire areas whose initial upper bound core damagejrequency F
ru was greater that 1x10 per year are evaluated.
Assumptions MM W Potent I regarding the "all engulfing fire, and fire risk modelsimplifications are addressed and potentially
.. iterations 4.._
relaxed to more accurately reflect y
_, _A y
the risk associated with a fire Develop and Detailed event in these particular fire Ouantify y
Fire areas.
Fire Fire Area Propagation Plant Model CDF >1E-/6 Analysis Following t.% adjustment of the Task 5
/
Task 7 conservative asumptions the fire i
risk modelis requantified, in the y"*
l case where the total fire area l
UBCDF is less than 1x104 per Figure 4 4b. Input / Output Path 2 l
year the fire area is screened Revised Estimation of Upper Bound Core Damage Frequency from further consideration.
i 4
Where the total fire araa UBCDF is greater than 1x10 per year the output is directed to Task 7, Detailed Evaluation of Fire Corn Damage Frequency. This sub-task is documented in report Section 4.6.2, Revised Estimation of Upper Bound Core Damage Frequency.
Upper Bound Core Damage Frequency Estimation for Multiple Area Fires The third input / output path develops and quantifies the fire risk model for multiple fire Develop and area events. Input is from the Fire Growth Quantify and Propagation Task (Task 8), the Fire Development of the Fire Initiating Event Plant Model Frequencies (Task 1) and the Fire Initiating Task 5 Event impact Table. For each multiple fire A
area event the frequency of the initiating event is calculated as the sum of the individual fire y..
areas which comprise the event. The impacts Multiple Area Fire?
of the newly defined initiators are also the sum of the impacts of the individual fire areas which comprise the multiple area fire. The no impacts and frequencies are factored into the Y
Level 1 OCPRA. The quantification of this fire risk model produces an estimation of the Figure 4-4c - Input / Output Path 3 upper bound core damage frequency as a Multiple Fire Area Quantification result of multiple fire area events.
This sece 4-6
==
dererred4)3 doc 10/01H
streiy end nist nai;iis Resisnon 0 input / output path is documented in report Section 4.3.
Task 4 - Critical Fire Area Mitigation Potential This task documents the Oyster Creek Nuclear Generating Station's fire detection and
- suppression systems. Input to the task is from the Fire Hazard Analysis Report and the Fire Mitigation Procedure. The information developed in this task serves as input to the Detailed Fire Propagation Analysis (Task 7). Details on this task are contained in report Section 4.5, Fire Detection and Suppression.
Task 7 Detailed Fire Propagation Analysis 4
Those fires areas whose upper bound core damage frequency is greater than 1x10 serve as -
input into the Detailed Fire Propagation Analysis. The Fire Area Mitigation information collected in Task 6 is used to adjust the conservative assumptbns made in the risk model for these areas.
The model is then re-quantified. The result of this task is revised risk model impacts and/or adjusted severe fire frequencies. Details on Task 8 are provided in report Section 4.6, Analysis of Plant Systems, Sequences and Plant Response.
Task 8 - Fire Growth and Propagation This task investigates the potential for fire growth and propagation of fires beyond individual fire areas. Evaluations of fire growth and propagation within a fire area are addressed in the Detailed Fire Propagation Analysis (Task 7) which is presented in report Section 4.6. This task Fire Growth and Propagation beyond individual fire areas is addressed qualitatively using the Electric Power Research Institutes (EPRI) Fire induced Vulnerability Evaluation (FIVE) assumptions regarding the effectiveness of fire barriers are applied.
l The input to the task is from the
- Identification of Critical Fire Areas (Task 1) and the Fire Initiating b
Event impact Table (Task 4). The Fire Growth result of this task is an evaluation and
- T_
Multiple of - the potential " multiple ' area Propagation
~
' Area Fire?
fires". 'in the ekse where a multiple Task 8 area fire-is assumed to occur, a new initiating event is developed.
This initiating event is. equal in I
frequency of occurrence to the Figure 4-5 sum of the frequency of - fire Multiple Area Fire Evaluation initiation of the fire areas involved,
~
The impacts of this new initiator is equal to the combined _ impacts of the fire areas involved. This new initiating event is input into the Development and Quantificatibn of the Fire Risk Model (Task 5), Details on Task 8 are presented in report Section 4.3, Fire Growth and Propagation, asce 47 men deferred-03 doc 10/0197
______J
._..m.
\\ s#
Sarety cnd Risk Antipis Prge 48 or43 i
Regision 0 Task 9 Presentation of Results This task assembles, summarizes and presents the overall results of Presentation the Oyster Creek Fire Individual Plant Examination including a of summary of containment performance. Details are presented in Results report Section 4.7, Presentation of nesults.
Task 9 i
Task 10 - Evaluation of the Results and Fire issues y
Evaluation This task applies the results and lessons learned to the Sandia og issues, A-45 and others. Details are presented in the following Results report sections:
and issues Task 10 Section 4.7, Containment Failure Modes due to Fires Section 4.8, Treatment of Sandia Fire Risk Scoping Study issues Figure 4-6 Results and Conclusions Section 4.9, USl A-45 and Requirements of NUREG 1407.
Each of the sections of this report begins with a detailed description of the task including the input to the task, output of the task and the steps which are used in the analysis. Taken together, the introduction to each section provides the detailed methodology of the performance of the Oyster Creek Fire Individual Plant Examination.
sece 4-8 deferred-Oldoc 10/02N7