ML20149E686
| ML20149E686 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 03/09/1994 |
| From: | Amarasooriya W, Meyer J SCIENCE & ENGINEERING ASSOCIATES, INC. |
| To: | NRC |
| Shared Package | |
| ML20149E682 | List: |
| References | |
| CON-NRC-05-91-068-0, CON-NRC-5-91-68 SCIE-NRC-212-92, SCIE-NRC-212-92-R01, SCIE-NRC-212-92-R1, NUDOCS 9406210222 | |
| Download: ML20149E686 (18) | |
Text
SCIE-NRC-212-92 OYSTER CREEK INDIVIDUAL PLANT EXAMINATION llACK-END TECilNICAL EVALUATION REPORT Revision 1 l
l l
W. H. Amarascoriya J. F. Meyer l
l l
Prepared for the U.S. Nuclear Regulatory Commission Under Contract NRC-05-91-068-03 March 9,1994 SCIENTECH, Inc.
I1821 Parklawn Drive Rockville, Maryland 20852 h/06210ZZZ"Xkh
OYSTER CitEEK INDIVIDUAL PLANT EXMINATION ll A C K-EN D TECilNICAL EVALUATION REPORT l
l Preparer
" f+f7k %
W.ll. Amarasooriya U Safety Analysis Group Technical Review Approval l
AA Ss< S1bi
/
Mohf(nmad hkxlants Jamesg. Meyer Marfager, Safety Analysis Group Manager, Safety Analysis Group SCIENTECil, INC.
March 9,1994
TABLE OF CONTENTS l' age t
1.
INTRODUCTION 1
2.
CONTRACTOR REVIEW FINDINGS.
2 l
l 2.1 Review and Identification ofIPE Insights 2
[
2.1.1 General Review ofIPE Back-End Analytical Process 2
2.1.1.1 Completeness.
2 2.1.1.2 Description, Justification, and Consistency 2
2.1.1.3 Process Used for IPE 2
2.1.1.4 Peer Review ofIPE.
2 2.1.2 Containment Analysis / Characterization 3
2.1.2.1 Front-end Back-end Dependencies 3
2.1.2.2 Sequence with Significant Probability 3
2.1 2.3 Failure Modes and Timing.
4 I
2 1.2.4 Containment Isolation Failure.
4 2 1.2.5 System /Iluman Response 5
2.1.2 6 Radionuclide Release Characterization 5
2.1.3 Quantitative Core Damage Estimate.
6 2.1.3.1 Severe Accident Progression 6
2.1.3.2 Dominant Contributors: Consistency with IPE Insights.
7 2.1.3.3 Characterization of Containment Performance.
7 2.1.3.4 Impact on Equipment Behavior 8
2.1.4 Reducing Probability of Core Damage or Fission Product Release.
8 2.1.4.1 Definition of Vulnerability.
8 2.1.4.2 Plant Improvements.
8 2.1.5 Responses to CPI Program Recommendations
.10 2.2 IPE Strengths and Weaknesses
.10 2.2.1 IPE Strengths
.10 2 2.2 IPE Weaknesses.
.I1 3.
OVERALL EVALUATION.
.12 4.
REFERENCES
.13 APPENDlX Oyster CwelIPE Ilack-End Review ii Revison 1/ March 9,1994
1.
INTRODUCTION This technical evaluation report (TER) documents the results of the SCIENTECil Review of the Oyster Creek Individual Plant Examination (IPE) Back-End submittal [1] This technical evaluation report complies with the requirements of the U.S. Nuclear Regulatory Commission i
contractor task order for Step 1 reviews, and adopts the NRC Step 1 Review objectives, which include the following:
To determine if the IPE submittal provides the level of detail requested in the Guidance Document," NUREG-1335 To assess the strengths and the weaknesses of the IPE submittal To pose a preliminary list of questions about the IPE submittal, based on this i
limited Step 1 review To complete the JPE Evaluation Data Summary Sheet.
In Section 2 of the TER, we s.ammarize our fmdings, and brie 0y describe the Oyster Creek IPE submittal as it penains to the work requirements outlined in the contractor task order. Each portion of Subsection 2.1 corresponds to a speci6c work requirement. In Subsection 2.2, w( set out our assessment of the Oyster Creek submittal strengths and weaknesses. In Section 3, wo i
present our evaluation of the Oyster Creek IPE overall, as well based on the Step 1 review.
Appended to this repon is the IPE Evaluation Summary Sheet, which we completed on the Oyster Creek IPE.
j i
i Oyster Creek IPli13ack-!!nd Review 1
Revison 1/ March 9,1994
2.
CONTRACTOR REVIEW FINDINGS 2.1 Review and identification of IPE Insights This section is structured in accordance with Task Order Subtask 1.
2.1.1 General Review ofIPE Hack-End Analytical Process 2.1.1.1 Com niet eness l
The Oyster Creek Individual Plant Examination (IPE) Back-End submittal is essentially complete l
with respect to the level of detail requested in NUREG-1335. The IPE submittal meets the NRC l
sequence selection screening criteria described in Generic Letter 88-20, and summarizes how this was done (see Table 8-1, page 8-5, of the Level 2 PRA).
2.1.1.2 Descrintion, Justification, and Consistency The IPE methodology used is clearly described and its selection isjustified. The approach followed is consistent with Generic Letter GL 88-20, Appendix 1.
2.1.1.3 Process lised for IPE As noted in Subsection 1.3, page 1-1, of the IPE. submittal report, "The analysis of Oyster Creek containment performance was accomplished in the context of extending the Level I study [2] to Level 2, as defined in NUREG/CR-2300."[3] The Level 1 model quantification led to 4
identification of 19 plant damage states (PDSs) with a frequency of lx10 per year or greater.
l'or the Level 2 analysis, these PDSs were condensed into a set of seven key plant damage states (KPDSs) for which containment event trees (CETs) were developed. Representative sequences were selected for each KPDS. MAAP3.0B, Rev. 7.03, was used to calcelate severe accident event timing and containment loads for each of the representative sequences.
2.1.1.4 Peer Review of IPE l
Independent peer review of the Oyster Creek IPE by the six-member Independent In liouse l
Review Group (IIIIRG) is discussed in Subsection D.3.2, Appendix D, of the Level 2 PRA. The l
1111RG review comments and their disposition are described in Subsection D3.3. Of the nine IlllRG comments listed, only one led to a textual revision. It was concluded that the other IlllRG comments could be adequately addressed by responding only to the reviewers who made them. It appears that the Oyster Creek IPE did receive adequate and appropriate peer review.
Oyster Creek IPE Hack-End Review 2
Revison 1/ March 9.1994
2.1.2 Containment Analysis / Characterization 2.1.2.1 Front-end, Hack-end Dependencies The interface between the Level 1 system analysis and the Level 2 containment analysis consists of a set of plant damage states (PDSs), as discussed in Section 5 of the Level 2 PRA. A PDS is the result of one or more of a number of physical conditions, which were analyzed in Level 1, including the following: pressure inside the reactor at the time of vessel breach, presence of water on the drywell floor, containment pressure boundary integrity status, availability of water to cool the core debris, suppression pool cooling, and containment venting. Other secondary conditions were also considered. The Level 1 analysis of Oyster Creek ended, in principle, at the onset of significant core damage, which was defined as the time when ". the top of the active fuel is uncovered, and vessel water level is continuing to drop " The Level 1 event trees identified 19 plant damage states with a frequency of 1.0E-8 per year or greater. For the purpose of Level 2 analysis, these were reduced to a set of seven key plant damage states, selected on the basis of the Generic Letter reporting criteria. Each of the seven KPDSs received consideration during the subsequent Level 2 analyses. It appears that proper account was taken of front-end to back-end dependencies, and, overall, the analysis of front-end, back-end dependency is logically and clearly presented.
2.1.2.2 Sequences with Sienificant Probability Accident sequences with a significant probability of occurrence were evaluated, as noted in the Level 2 PRA, Sections 5 and 8. A number of KPDSs with a higher annual frequency rate than 1.0E-8/ year received further consideration, using the containment event tree (CET). These KPDSs were ". a factor of 10 lower that the NRC sequence frequency criteria," and are listed in Table 8-1 (on page 8-5 of the Level 2 PRA). The CET events used to funher analyze the KPDSs were selected to address in-vessel core degradation, the potential for in-vessel recovery, the phenomena associated with ex-vessel progression, containment integrity challenges, containment failure, its timing, and the effectiveness of other safeguard systems to mitigate offsite releases. The CET end-states were binned together into a number of release categories. Because only one CET was actually developed, it had to be quantified for each KPDS, and therefore the CET branching probabilities (split fractions) varied, in most cases, for each KPDS. The CET used in the Level 2 PRA was developed to resemble the NUREG-1150 (and NUREG/CR-4551) accident progression event trees (APET) developed for the Peach Bottom examination.
Ilowever, fewer events were analyzed during the Oyster Creek IPE (The examiners asked 145 questions at Peach Bottom, as compared viith 16 at Oyster Creek). The Oyster Creek IPE submittal explains that answers to many of the questions asked at Peach Bottom were " implicit in the definitions of the Oyster Creek plant damage states" and therefore not included in the Oyster Creek CET. (in actuality, however, some of the questions whose answers were considered implicit were asked as part of the Oyster Creek CET.)
The Oyster Creek CET, shown in Figure 7-1, page 7-11, of the Level 2 PRA. See Subsection 21.3.3 for a further description of CET top events.
Oyster Creek IPli Back{nd Review 3
Revison 1/ March 9,1994
The quantification of the CET for each KPDS was carried through a number of split fractions defined for each applicable CET top event. The process appears reasonable, but it is difficult to follow the nature of each split fraction, the terminology employed, and the split fraction logic used (see Table 10-1 of the Level 2 PRA). The final results of the quantification of the CET sequences are not given. The results were used to define the release categories and to calculate their frequencies directly The leading release fractions are discussed in Subsection 2.1.2.6.
2.1.2.3 Failure Modes and Timine The Oyster Creek containment failure characterization is described in Section 6 of the Level 2 PRA, and it is detailed in an EQE Engineering calculation report, which is appended to the Level 2PRA Page 15 of the EQE Engineering report notes the following:
[TJhe potential failure modes of the containment shell are investigated The loads considered include temperature, pressure, and dead load Failures of both the dryvell and the suppression chamber (torus) are considered he faihire modes examined include.
s 1.
Membrane failures of the drymell shell (sphere. eylinder, head shell) 2.
Failure of the drymeli head flange seal 3
Failure of the vent line from the dnuell to the suppression chamber 4
Failure of the suppression chamber.shel!
5.
Failure at penetrations For each of the failure modes examined, the probability of failure was calculated as a function of internal pressure within containment metal temperatures ranging from 300 F to 1,200*F. A log-normal failure model was used to perform these calculations. An overall uncertainty value was estimated, using material-strength cnd modeling uncertainties, whose rationale for selection as the basis of the overall estimate was not given.
The containment capacities of each failure mode at temperatures of 300*F and 700*F are shown in Tables 6 I and 6-2 of the Level 2 PIUL The failure made with the least containment pressure was ti;rt ofleakage through the bolted drywell head flange connection: 121 psig at 300*F after 70 hours8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br />. Ilowever, as 70 hours8.101852e-4 days <br />0.0194 hours <br />1.157407e-4 weeks <br />2.6635e-5 months <br /> is a relatively long time for accident progression, the subsequent i
failure mode-the membrane failure of the drywell shell (134 psig also at 300*F)-was more significant (from the source term viewpoint).
The liner melt-through (a consequence of direct contact of the containment she'l with fuel debris) was also analyzed. The results appear reasonable when compared with NUREG/CR-5423 (Table 10-10 of the Level 2 PRA). A difference factor of 2 cited in the Oyster Creek results is attributed to a 6-inch-high and 1-foot-wide curb on the drywell floor. The analysis of this difference from NUREG/CR-5423 the appears to have been adequate.
2.1.2.4 Containment Isolation Failure Containment isolation failure is considered part of a plant damage state. One of the pnmary conditions taken into account before binning plant damage states is the " Containment Pressure Oyster Creek IPE Back-End Review 4
Revison 1/ March 9,1994
Boundary Integrity Status." This condition addresses the containment isolation failures and potential containment bypass, as well as early or late containment failures.
As stated on page 10-6 of the IPE submittal report, " Pre-existing leaks were not specifically addressed in this study. Because the Oyster Creek containment is continuously monitored for oxygen content to insure inerting, plant operating staff would be alerted to such leaks and would respond accordingly." This is reasonable and consistent with other back-end assessments of facilities with inerted containments.
2.1.2.5 System /Iluman Response The CET considers possible methods for arresting the core under the top event VB of CET. Each KPDS is screened to identify the ones with reasonable recovery potential.
Top Event Number 11 in the CET (Emergency Crew vents Containment in Core Damage Scenarios) modeled the intentional venting of the suppression pool air space. Success of this event means that the vent flow capacity is adequate and that long-term containment failure would be precluded. Depending on the KPDS, the probability of this top event occurring (probability of i
failure to vent) ranges from 0 017 to 1.0.
t It is assumed that the probability that the crew vents the containment does not change from " pre-core damage" venting in the Level 1 analysis to the " post-core damage" venting in the Level 2 analysis. In addition to wetwell (toms) venting, venting via the drywell is available and is
" procedure-activated" for the representative sequence in the KPDS OJAU (note Section 10.11 on page 10-12) i About half of the CDF is allocated to "No Vessel Breach," resulting in no radionuclide releases.
Vessel breach (afler core damage) is prevented by either introducing fire protection water when the vessel is under low pressure or providing suflicient " control rod drive hydraulic system" flow when the vessel is under high pressure. For both vessel injection modes, operator action is required. (Note page 10-2 of the Level 2 PRA ) Thus it appean that this operator-controlled cooling function has a key bearing on the radiological release profile for Oyster Creek.
Radionuclide Release Charecterization 2.1.2.6 l
The radionuclide release characterization is described in Section 11 of the Level 2 portion of the submittal. Release categories (qualitative descriptions of the containment event tree end-state bins) and associated source terms (quantitative descriptions of the CET end-state bins, including release timing and release fractions) are generated. As an aid in defining the release categories, a source term event trec (STET) is used, as stated in the submittal: "The purpose of such a tree is to define the different release categories for which the source term characteristics could be sufTiciently different to warunt a separate source term definition."
The STET is shown in Figure 11-1. The characterization of the radionuclide releases from the containment are a function of the seven top events in the STET. A seven-character end-state Oyster Creek IPE Back End Review 5
Revison 1/ March 9,1994 M.+
,s,
identifier was developed, as described on pages 11-2 and Il-3. The seven STE1 top events were keyed to either a plant damage state group or to the status of certain CET top events. The rules for binning CET sequences to release categories are listed in Tables Il-1 and I l-2. The results of the release category binning are shown in Table Il-3. Each release category was assigned to an enveloping release category "according to the rules of conservative condensation." From this, six key release categories (KRCs) were generated, as a simplified and conservative characterization of the releases to the environment The source terms for these KRCs were calculated by selecting representative sequences and using h1AAP to model the behavior and release of 12 radionuclide groups, as listed on page Il-5 of the Level 2 PRA. The analysis process and assumptions are described for each KRC starting on page Il-6. In Table Il-4, page 11-16, the frequency of each KRC is listed, while in Table 11-5, page Il-17, source term information (release fractions and release times) for each KRC is provided. In addition, Table 11-6, page i1-18 summarizes the distribution of Csl within the vessel and containment at the end of the h1AAP runs.
The discriminator interrogatories used to define the release categories were the following:
Reactor coolant system pressure at vessel breach?
Drywell sprays available?
Core damage arrested in-vessel?
Time of containment failure?
Size of containment failure?
Containment bypassed?
Suppression pool sembbing prior to containment faliure?
Accident-mitigation in reactor building?
Generic Letter 88-20 states that the following should be reported: "any functional sequence that 4
has a core damage frequency greater than lx10 per reactor year and that leads to containment failure which can result in a radioactivc release magnitude greater than or equal to BWR-3 or PWR-4 release categories of WASH-1400." The Oyster Creek IPE submittal meets this reponing requirement. See Table 8-1, page 8-5, for a summary, i
The radionuclide characterization is well developed and portrayed in the submittal. It appears to be reasonable assessment of radionuclide transport and release.
2.1.3 Quantitative Core Damage Estirnate 2.1.3.1 Severe Accident Procression The accident progression analysis, performed with hiAAP computer code, is discussed in Section 9 of the Level 2 PRA. (Note the limitations and assumptions used in the hiAAP analysis, as described in Section 4.3 of the Level 2 PRA.) hiAAP results are discussed and presented in figures and tables for the following KPDSs:
Low Pressure Station Blackout with Stack-Open Relief Valve (PlFW)
Ihgh Pressure Station Blackout (NIFW)
Oyster Creek IPE llack-End Review 6
Revison 1/ March 9,1994
l, r
Large DBA LOCA with No Core Spray (OIAU)
Turbine Trip ATWS with SLC Failure (MKCU)
Reactor Water Cleanup (RWCU) System Failure in Pressure Reducing Station (OJAU)- Bypass Sequence l
Loss of Feedwater with Failure of Scram Dircharge Volume (SDV) to * ' ate l
(MJ AU) - Bypass Sequence Station Blackout with SDV Failwe to Isolate (NJllW)- Bypass Sequence.
A discussion of the phenomenological uncertainties of severe accident progression could not be located in the submittal.
l 2.1.3.2 Dominant Contributors: Consistenes with IPE Insights Table 1-5, page 1-10, of the Level 2 PRA compares results at Oyster Creek and NUREG-1150 Peach Bottom for the dominant contributors to containment failure. These results and those of l
the Fitzpatrick IPE are given in Table 1, baow, where it can be seen that the early containment j
failures at Oyster Creek are less than one-third of those at Fitzpatrick or Peach Bottom.
Correspondingl,", Oyster Creek has a significantly higher fraction of no vessel breach than does l
Fitzpatrick or Peach Bottom.110 wever, the Oyster Creek containment is assumed to always fail, in pan because no AC recovery is assumed.
Table 1. Containment Failure as a Percentage of CDF:
Comparison to Fitzpatrick IPE and Peach Bottom NUREG-1150 Results i
Containment Failure Fitzpatnck IPE Peach Bottom /
Oyster Creek NUREG 1150 CDF (per year) 1.9x10" 4.5x10 3.2x10 4
4 Early Failure 60.4 55.7 15.9 Bypass na na 7.3 Late Failure 26.0 16.0 26.4 Intact 2.5 18.0 0
No Vessel Breach 11.1 10.0 50.4 1
2.1.3.3 Characteri7ation of Containment Performance The containment performance observed during the Oyster Creek IPE was characterized using containment event trees. The top events of these event trees are discussed in Subsection 7.2 and listed in Table 7-1 of the Level 2 PRA. The CET chronologically models core degradation, vessel failure, containment behavior, and reactor building behavior. The first top event analyzed was j
one that would occur in the entry state from the front-end (i.e., a KPDS). The next top five events consisted of phenomena that could occur from the time core damage began until vessel i
Oyster Creek IPE 13ack End Review 7
Revison 1/ March 9,1994
I failure seemed imminent (these events were: vessel breach, safety valve failure, containment j
intact prior to vessel breach, containment leakage, and suppression pool bypass).
The next four events analyzed were phenomena that could occur during and shortly after vessel breach. These four events could affect transient loading conditior.s significantly and could lead to early containment failure (blowdown, direct containment heating, ex-vessel fuel coolant interaction, drywellliner melt-through). The next three events were analyzed with attempts to formulate a long-term containment response, to prevent containment failure by establishing adequate debris bed cooling, and to either remove or vent containment heat. These events were:
occurrence of containment venting, incidence of containment remaining intact late, and occunence of containment leak areas. The last two top events set out in the Oyster Creek CET were simulated to s'udy the phenomena that could affect the reactor building integrity and the ability of the phenomena to reduce an offsite source term if the containment failed.
1 As shown in the summary in Table 2, below, many of the top events addressed containment behavior. The containment loading was calculated using the MAAP computer code.
2.1.3.4 Impact on Equipment Itchavior A discussion of the impact of severe accidents on equipment behavior could not be located in the submittal 2.1.4 Reducing Probability of Core Damage or Fission Product Release 2.1.4.1 Definition of Vulnerability As noted in Subsection 3.2, page 3-2, ofIPE Submittal Repon, "A vulnerability is defined as any core damage sequence that exceeds lx10" per reactor year, or any containment bypass sequence 4
or large early containment failure sequence that exceeds lx10 per reactor year." GPU found no vulnerabilities for the Oyster Creek nuclear power plar.t.
2.1.4.2 Plant Improvements With respect to plant improvements, Subsection 8.2, page 8-5, of the submittal report explained that "Because of the relatively low frequencies associated with the various containment failure modes, no specific hardware modifications or changes to existing procedures beyond those identified in the level 1 analysis are planned at this time. The level 2 PRA will be used as a major input to the development of accident management guidelines."
Although no back-end improvements are planned as such, the submittal does address front-end improvements, which might affect mitigation of the consequences of back-end events. Note Section 2.1.5, below.
A plant improvement, planned as pan of the 14R modifications, is the hard-piped containment vent system. This system is assumed available in the IPE and becomes imponant for back-end Oyster Creek IPE Ikk.End Review 8
Revison 1/ March 9.1994
Table 2. Oyster Creek CET Top Event Descriptions Top Event Top Event Description Designator CET Entry State VB Vessel Breach Prevented ES EMRV(s) or Safety Valve (s) Sticks Open Prior to Vessel Breach in High Pressure Melt Scenarios 11 Containment intact Prior to Vessel Breach Ll Small Leak Area if Containment Fails to Top Event 11 S1 Suppression Pool Not Bypassed Prior to Vessel Breach Events during or Shortly afler Vessel Breach ET Debris Not Entrained 12 Containment Intact after Vessel Breach
_L2 Small Leak Area if Containment Fails in Top Event 12 LM No Significant Release of Fission Products into the Reactor Building due to Drvwell Liner Melt-Through Long-Term Containment Events S3 Suppression Pool Not Bypassed Late DV Emergency Crew Vents Containment in Core Damage Scenarios 13 Containment intact Late L3 Small Leak Area if Containment Fails in Top Event 13 Events Pertaining to Reactor Building Effectiveness HB No Hydrogen Burn in Reactor Building BE Reactor Building Effective assessment, during " dirty venting," that is, venting after core damage. This is addressed under CET Top Event 11, discussed on page 7-7 and page 10-12 of the Level 2 PRA. For the Key Plant Damage State (KPDS) OIAU (note the containment matrix in the TER Appendix), the venting is judged to be effective 98.3% of the time, the same value used for the Level 1 " clean" venting.
Oyster Creek II'E Back{nd Resiew 9
Revison 1/ March 9,1994
2.1.5 Responses to CPI Program Recommendations Generic Letter No. 88-20, Supplement No 1, notes the following recommendations of the Containment Performance Improvement Program (CPI) pertaining to the Mark-1 containments:
Alternate water supply for drywell spray / vessel injection Reactor pressure vessel depressurization system reliability enhancement Emergency procedures and training.
In Section 4 of the submittal ieport, these recommendations are addressed.
An alternate water supply for vessel injection is in place and credit is taken for it in the JPE Specificai y, it is GPU's position that injection is timely enough and of sufIicient quantity to prevent vessel breach. This is addressed in top event VB in the CET. An alternate water supply for &ywell spray was considered to not be cost-effective, although there are situations where, if water was provided to the dry, non-coolable core debris, it would mitigate the consequences of the accident.
Reactor pressure vessel d3 pressurization system reliability enhancement is accomplished by providing an alternate AC source connection scheduled for the 14R refueling outage. This will reduce the likelihood of an extended station blackout, thereby improving depressurization reliability.
GPU has implemented the BWR Revision 4 EPGs and they are reflected in the IPE submittal.
2.2 IPE Strengths and Weaknesses 2.2.1 IPE Strengths 1.
The IPE submittal appears to have addressed most of the important and relevant phenomena in suflicient detail. Most of the leading severe accident hazards, such as liner melt-through, are systematically addressed.
2.
The results of the IPE at Oyster Creek are compared in suflicient detail with the NUREG-1150 results at Peach Bottom, and the differences are well documented.
3.
The back-end analysis is robust, (i.e., it was performed in a way that protects the results from the impact of changes that may occur later as the result of a front-end analysis). A front-end analysis should not significantly affect the conditional calculations of the back-end results.
Oyster Creek IPE Back-End Review 10 Revison 1/ March 9,1994
2.2.2 IPE Weaknesses 1.
A discussion of the impact of severe accidents on equipment behavior could not be located in the submittal.
2.
Although the probabilities of sequence occurrence seem reasonable throughout the submittal, the sources of such estimates are often not cited. For example, CET split fractions probabilities are not well defined.
3.
The Oyster Creek IPE seems to focus on the events and containment characteristics that have early, detrimental effects on health, and on large, early releases to the environment.
Large, early releases are important, but constitute only a small fraction of the probable accident events that should be considered and the subsequent containment responses that make up a back-end assessment. Attention to long-term effects and consequences is wanting i
I Oyster Creck IPE Back End Resiew 1I Revison 1/ March 9.1994
I I
l 3.
OVERALL EVALUATION As discussed in Section 2, this IPE submittal contains a large amount of back-end information, which contributes to the resolution of severe accident vulnerability issues at Oyster Creek. A large segment of the back-end portions of the IPE submittal is well written and directed to addressing Generic Letter 88-20 issues. The issue ofliner melt-through is addressed well. The concrete curb is an attractive design feature to protect the liner.
Ilowever, there appear to be some weaknesses in the submittal, as set out in Section 2 of this report. In summary, our concerns about the submittal follow:
The methodology appears to be geared to providing consequence assessment and comparison to safety goals, as contrasted to understanding the vulnerabilities of i
the containment and the impacts that equipment, EOPs, and phenomenology have on containment performance.
There appears to be no discussion of back-end uncertainties. This coupled with a lack of sensitivity analysis weakens the overall conclusions and could give the wrong impression of the state of knowledge of containment vulnerabilities A large fraction of the core damage frequency results in no vessel breach, that is, the accident is arrested in-vessel. There appears to be no discussion of the uncertainties in this conclusion or of the sensitivity of this result to major assumptions.
The quantification of the CETs could not be traced.
Oyster Creek IPE 11ack-End Review 12 Revison 1/ March 9,1994
l l
4.
REFERENCES l
1.
General Public Utilities Corporation, " Oyster Creek Individual Plant Examination Report,"
August 1992.
2.
GPU Nuclear Corporat;on and PLG, Inc., " Oyster Creek Probabilistic Risk Assessment i
(Level 1)," Vols I th"ough 6, November 1992.
3.
American Nuclear Society and Institute of Electrical and Electronics Engineers "PRA Procedures Guide: A Guide to the Performance of Probabilistic Rick Assessments for Nuclear Power Plants," prepared for the U.S. Nuclear Regulatory Commission, NUREG/CR-2300, Vols 1 and 2, January 1983.
a i
i i
I I
r h
i r
4 1
t Oyster Creek IPE 13ack-End Review 13 Revison 1/ March 9,1994 i
l APPENDIX IPE EVALUATION AND DATA
SUMMARY
SIIEET ilWR Itack-end Facts Plant Name j
l Oyster Creek f
f Containment Type Mark I
~
i i
Unique Containment Features 1
Presence of a drywell floor concrete curb; a thinning of the liner in the sandbed region; 25 percent increased stmetural capability of toms as a result of a backfit performed; and an increased containment cooling capability as a result ofimproving the NPSH limits with j
a rise in drywell pressure l
Unique Vessel Features s
None found l
Number of Plant Damage States 19 Ultiraate Containment Failure Pressure 134 psig at 300*F l
f Additional Radionuclide Transport And Retention Structures l
)
Suppression Pool scrubbing is assumed. However, the containment failure modes appear j
l to preclude the possibility of post-containment-failure scrubbing. Reactor Building i
mitigation does not appear to be credited.
l Conditional Probability That The Containment Is Not Isolated i
i A value could not be found in the IPE submittal, but it is assumed to be very low because i
of the characteristics of the Mark I inerted containments.
l j
Oyster Creek IPE Ilack-End Review A-1 Revision 1/ March 9,1994 i
t i
i
APPENDIX (continued)
IPE EVALUATION AND DATA SUS 151AltY SIIEET important insights, including Unique Safety Features Presence of a drywell floor concrete curb; a thinning of the liner in the sandbed region; I
the earliest release occurs 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> after the accident, and the worst release is caused by a bypass scenario; 25 percent increased structural capability of tonis as a result of a backfit performed, and an increased containment cooling capability as a result ofimproving the NPSil limits with a rise in drywell pressure Implemented Plant Improvements None implemented, but the containment cooling capability increased as a result of improving the NPSH limits with a rise in drywell pressure C-51a t rix Simplified Oyster Creek C-Statrix Key Plant Damage Frequency Early Failure Bypass Late Failure No Vessel State per year Breach PlFW l.13 E-6 0
0 0
1.0 NIFW l.06E-6 0.31 0
0.69 0
l 01AU 5.74 E-7 0.01 0
0.18 0.81 MKCU l.70E-7 1.0 0
0 0
OJAU 164E-7 0
1.0 0
0 MJAU 5 2611-8 0
1.0 0
0 NJHW 1 54E-S 0
1.0 0
0 Oyster Creek IPE lhck End Review A-2 Revision 1/ March 9,1994
~.
s i
ENCLOSURE 4 OYSTER CREEK INDIVIDUAL PLANT EXAMINATION TECHNICAL EVALUATION REPORT (HUMAN RELIABILITY ANALYSIS) 6 s
1 i
l i
}
..