ML20204J827
| ML20204J827 | |
| Person / Time | |
|---|---|
| Issue date: | 10/14/1997 |
| From: | Callan L NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO) |
| To: | |
| References | |
| SECY-97-234, SECY-97-234-01, SECY-97-234-1, SECY-97-234-R, NUDOCS 9903300167 | |
| Download: ML20204J827 (85) | |
Text
_
m g y,
i&" 8) i % d99 _px :
k dote insalSV g
,o e
w
,,,,e
>~some..
.....se.seeees POLICY ISSUE (Information)
October 14.1997 SECY-97-234 ERE:
The Commissioners FROM L. loseph Callan Executive Director for Operations
SUBJECT:
CUARTERLY STATUS FOR THE PROBABILICTIC RISK ASSESSMENT IMPLEMENTATION PLAN PURPOSE:
This quarterly report presents the status of activities for the Probabilistic Risk Assessment (PRA) imp!ementation Plan, including the development of risk-informed standards and guidance. The report also serves to provide responses to Staff Requirements Memoranda (Attachment 1) dated May 28,1997, June 5,1997, and June 13,1997, which include, respectively:
(1) actions the staff has taken to expedite (a) the use of IPE results to prioritize inspection activities; (b) improvements in regional capabilities for the use of PRA and risk insights; and (c) provision of related inspector training; (2) the staff's plans for training NRC staff on (a) the risk-informed regulatory approach (es) contained in the regulatory guidance and standard review plan documents and (b) overall PRA methods and techniques; and
,p (3) an update on the staff's efforts to work with industry to address shortfalls and limitations in the data on reliability and availabllity of risk-significant systems to be provided to the,0 /
staff voluntarily.
30000 CONTACT.
FOTE: TO BE MADE PUI LIC11 AVAILABLE IN Ashok Thadarii, OEDO 5 WORKING DAYS FROM THE DATE OF THIS PAPER 415-1705 2
9903300167 971014 p)/, ' ' '
//
,s f
(
/
ff,
/.
ec e'
The Commissioners 2
BACKGROUND-In a memorandum dated January 3,1996, from the Erefive Director for Operations to Chairman Jackson, the staff committed to submitting quarterly reports on the status of its development of risk-informed standards and guidance. Previous quarterly reports were sent to the Commission on March 26, June 20, and October 11,1996, and on January 13, April 3, and July 22,1997. This quarterly report covers the period July 1,1997 to September 30,1997.
DISCUSSION:
Attachments 2 and 3 provide this quarter's implementation plan update. Significant achievertionts in the past quarter include the following:
The staff incorporated proposed resolutions of the policy, technical, and process issues
=
in drafts of the application-specific Regulatory Guide (RG) and Standard Review Plan (SRP) for inservice inspection (ISI), and discussed these new drafts, with the Advisory Committee on Reactor Safeguards (ACRS) and th? Committee to Review Generic Requirements (CRGR). Both the ACRS and the CRGR have reviewed the guidance and concurred in the staff's proposal to issue the guidance for comment by the public.
On August 20,1997, the staff forwarded the draft guidance documents to the Coni.tission and requested its approval for issuing the documents for comment.
Commission approval was received in an October 1,1997, SRM.
To facilitate solicitation of public comments on the ISI RG and SRP, the staff will conduct a workshop during the comment period to explain the draft documents and answer questions. The workshop will be held late November or early December, at the Marriott Hotelin Bethesda, Maryland.
In completing the draft RG and SRP for risk-informed inservice inspection, the staff has found that a greater than expected effort was required to incorporate all points of view and gain a consensus on draft guidance. With this experience, the staff projects that the schedule for issuing the final ISI RG and SRP will slip from February 1998 to April 1998.
For risk-informed ISI programs, the industry had identified three pilot plants that would submit requests for authorization for the use of risk-informed ISI methodology these applications have not yet been received. In a letter from NEl, dated August 29,1997, the industry requested to add two additional plants to the list of pilot applications and identified an aggressive schedule for all the pilot plants. The scope for the two new pilot plants is limited to Class-1 piping (primary coolant system piping only). To date, only one application has been submitted by one of the pilot plants with a limited scope RI-ISI program.
j Due to industry's delays in submitting the applications, and the addition of two plants as pilots, the staff is unable to develop an integrated review schedule for the pilot plants et this time. This schedule is contingent on information regarding actual timing of submittals, the quality of the submittals, and the ability of the pilot plant licensees to commit the resources necessary to respond to the staffs requests for additional
l The Commissioners 3
information (RAls). The staff continues to hold working meetings with industry to
' facilitate the development of regulatory guidelines.
The staff completed ten more maintenance rule baseline inspections, which included inspection of licensee methods for using PRA in maintenance programs and in inspection of safety assessments performed by licensees when removing equipment from service for maintenance in accordance with Paragraph (a)(3) of the Maintenance Rule. As of September 30,1997, the staff has completed 36 inspections.
The NRC staff briefed the Commission in May 1997 on the individual Plant Examination (IPE) insights report, draft NUREG-1560, " individual Plant Examination Program:
j Perspectives on Reactor Safety and Plant Performance." As a result of this briefing, the staff received a SRM dated May 28,1997, requesting the staff "to expedite activities in the following areas: (1) using IPE results to prioritize inspection activities; (2) improving regional capabilities for the use of PRA and risk insights; and (3) providing related inspector training." The staff has been active in accomplishing each of the three items as described in item (3) of Attachment 2.
In a June 5,1997, Staff Requirements Memorandum (SRM), the Commission requested information on the plans for training NRC staff on 1) the risk-informed regulatory approach (es) contained in the regulatory guidance and Standard Review Plan documents and 2) overall PRA methods and techniques. Attachment 4 provides the staff's response to the SRM. The attachment describes the training that will be necessary to implement the initiatives discussed in the draft RGs and SRPs for risk-informed regulation.
The staff has developed responses to all the public comments it has received on draft l
NUREG-1560 and where appropriate, draft NUREG-1560 has been revised. Attached for information are present drafts of the executive summary of NUREG-1560 (Attachment 5) and Appendix C of NUREG-1560 which documents resolution of comments from the public (Attachment 6). The final version of NUREG-1560 will be I
published in November 1997.
i A draft interim report has been developed that provides preliminary perspectives and summarizes the informa' ion presented in the first 24 Indi Ja! Plant Examination for Extemal Events (IPEEE) submittals reviewed by the staff. This interim report will be sent to the Commission by the end of November 1997. A summary of the significant preliminary perspectives frorn the first 24 IPEEE reviews is preserued in Attachment 7.
In an SRM dated June 13,1997, the Commission requested that the staff periodically report on their efforts to work with industry to address shortfalls and limitations in the data on reliability and availability of risk-significant systems to be provided to the staff voluntarily. The staff's quarterly report on this activity is provided in item (10) of.
1 i
The Commissioners 4
i 4
A three-day public workshop was held on August 11-13,1997, on the following draft Regulatory Guides, Standard Review Plans, and NUREG report:
l
. General Guidance (DG-1061 and SRP) 2 i
Inservice Testing (DG-1062 and SRP)
)
Graded Quality Assurance (DG-1064)
Technical Specifications (DG-1065.and SRP), and The Use of PRA in Risk-Informed Applications (NUREG-1602) l i
The workshop was well attended by industry representatives. They offered a number i
of constructive comments, some criticisms, and some suggestions for changing the 1
guidance. Overall, the comments indicated general support for pursuing risk-informed l
regulation but in a manner which would necessitate modifications to the draft j
guidance. The significant issues raised at the workshop are summarized in item (1) of In a lett9r to the NRC dated August 21,1997, the Nuclear Energy Institute (NEI) made a l
proposal for three new risk-informed pilot applications of PRA in support of changes to the licensing basis of operating nuclear power plants. The staff met with NEl on j
' September 17,1997, to discuss the proposal, including potential NRC activities. The pilots would use a full scope PRA to assess risk versus the regulatory requirements and plant operating and maintenance costs. The staf' has concluded that, in concept, the initiative is worthwhile and plane to meet with NE! h November to discuss plans for i
j pursuing the initiat /e.
In June,1997, NRC staff met with representatives of the American Society of Mechanical Engineers (ASME) to discuss cooperation with both industry and professional societies to develop new codes and standards, as directed in the SRM on Direction Setting Issue (DSI) 13, dated March 7,1997 (see SRM in Attachment 1). The development of PRA sta ards was one subject of this meeting. At the meeting ASME indicated their interest and is convening an ad hoc committee that will have the responsibility to develop such a standard. This committee will be comprised of ASME personnel, NRC staff, national laboratory, academic, and industry personnel.
A charter for this committee is now being drafted, and will describe the goals and objectives of the committee, comm"Me membership and associated responsibilities, i
schedules, and milestones. An addinon, the charter will include anticipated scope of the standard (e.g., Level 1,2 and 3 PRA, including core damage accidents initiated by internal and extemal events during full power operation) and the level of PRA modeling and analysis appropriate for different PRA uses. The Commission will be informed of progress on this development work in the quarterly updetes of the PRA Implementation i
Plan.
.. ~. - - -..
The Commissioners 5
COORDINATION:
The Office of the General Counsel has reviewed this paper and has no legal objections to its issuance.
b'
[
.A.
~
L. Joseph Callan Executive Director for Operations Attachments:
As stated DISTRIBUTION:
Conunissioners OGC OIG OPA OCA CIO CFO EDO SECY 4
l e
l 1
I i
Staff Requirements Memoranda Addressed in October 1997 l
PRA implementation' Plan Update i
I i,
9 4
4
?:m i en
+-
nn.v r.oi u o i i r.cuvi ww7voz-wr w w nmfgntgaggr yW
~ I Action: (Morrison, xtS/
Ollins, NRR l
'e,,
UMED SWES gys;,gg))gn
~
p NUCLEAR RCOULATOftY COMBCSCJON Jurtjer:
3 I
wAsoctoN.3A ENOM1 Th0Epson
/
IN Kesecase, etxAsi::
\\.,~.
- May 28, 1997 ps,.a To.
Mo70507 r
Norry ascasim Blaha Ross, AE00 MEMORANDUM TO:
L. Joseph Callan Exec,d'iv~ Director for Operationa Joh.'C.Hoy%e,(Secretary FROM:
l
SUBJECT:
STAFF REQUIREMENTS - BRIEF N3 ON IPE INS!GHT REPORT, 2:00 P.M, WEDh'ESDEV, MaY 7,
- 1997, COMMISSIOh?RS' CCNFERENCE ROOM, ONE WHITE FLINT NORTH, ROCKVILLE, MARYLAND (OPEN TO PUBLIC ATTENDANCE)
The couutissivu was briefed by the NRC stef f en the Individual Slant Examination (IPE) insicht repott.
The C mission asked the staff no expedite activl les c.n Le !valu. lug areas:
(u using.
IPE results to priorinize inspection activities; ;2) improving regional capabilinies for the use of PRA and risk insights; and
{2) providin
_ ted inspecter training.
fEDCr)
RES)j iSECY susp=nse:
tir $/39 9700205 4
W v
The Commission asked the stati to prov:.ce tne scope anc senedule of activities related to using IPE results to assess reculatory effectiveness in resolving major safety issues.
The Commissien specifically requested that the staff provide an estimate cf the average cost to respond to the Staticn Blackout rule per persen-rem avo.rted in achieving an average reduction in core damage frequency of 2E-5/RY.
These activities should be coordinated with the regulatory effettiveness organization.
fEBCr)
(NRR)
(SECY Suspense:
6/27/97) 9700207 After the IPE database has been placed on the Internet, the staff should consider allowing licensees to upde.te their IPEs voluntarily to reflect changes i a plant configuration.
(RES)
S a
s sec,**
UNITED STATES o
a o,
NUCLEAR REGULATORY COMMISSION j' '
g V, ASHINGTON D C 20555 j
i 3,
j
%, ' s.'
,cf t
June 5, 199e orricE or rHE
.$E CR ET A R Y f
f MEMORANDUM TO:
L. Joseph Callan Execu iv Dir ctor for Operations AGw -
f FROM:
Joh
. Hoyl Secretary STkFREQUIREMENTS-SECY-97-077-DRAFT
SUBJECT:
REGULATORY GUIDES, STANDARD REVIEW PLANS AND NUREG DOCUMENT IN SUPPORT OF RISK INFORMED REGULATION FOR POWER REACTORS The Commission has approved publication of the draft regulatory guides, standard review plans and NUREG document for a 90-day public comment period.
(EDO)
(SECY Suspense:
6/13/97) i i
The staff should provide the Commission information on its plans for conducting public workshops.
The public workshop (s) to be i
conducted during the public comment period should be of sufficient duration and depth to provide meaningful insights into the approaches described in the documents.
1 the staff should provide the Commission information In addition, on the risk-informed on its plans for training the NRC staff 1) contained in the regulatory guidance and regulatory approach (es) in overall PRA methods and standard review plan documents and 2) dcular attention should be given to increasing techniques.
Part basic user-level knowledge of PRA methods at the regional level.
(EDO)
(SECY Suspense:
9/30/97)
The staff should ;_.itinue to evaluate the proposed decision criteria and the methods of ensuring conformance to the criteria.
The staff should also develop guidance on how to confirm the licensing basis assumptions and analyses used to justify current changes.
SECY NOTE:
THIS SRM, SECY-97-077, AND THE COMMISSION VOTING RECORD CONTAINING THE VOTE SHEETS OF ALL COMMISSIONERS WILL BE MADE PUBLICLY AVAILABLE 5 WORKING DAYS FROM THE DATE OF THIS SRM.
. the staff should explcre the fcilowing areas to In particular.,
add clarity and consistency :c :he process.
The feasibility of assigning assurance levele for 1.
conformance to decision criteria.
values for comparisons with decision The use of point consideration of 2.
criteria, without any explicithow consideration of uncertainty uncertainty (i.e.,
should be explicitly considered in conjunction with using point values -- for example, use of probability limits).
The implications of small increases in core damage and large early release frequency 3.
f requency (CDF) as a codified in the guidance documents, function of the uncertainty associated with the PRA (LERF) results.
Codifying in the guidance documents the experience gained f rom r'3e pilots to provide additional guidance 4.
attention" process when on the " increased management proposed changes approach the guidelines.
Clarifying the distinction between risk-informed and risk-inftrmed, performance-based regulatory approaches.
5.
/
The staff should continue to pursue the long range l
by different licensees by promoting high quality standards, in a timely The staff should continue its efforts to complete, manner, the i
to complete the draft regulatory guidance and standard rev ew plan for inservice incpection.
Chairman Jackson cc:
Commissioner Rogers Commissioner Dicus Commissioner Diaz Commissioner McGaffigan OGC CIO CFO OCA ASLBP (via E-Mail)
OIG Office Directors, Regions, ACRS, ACNW, PDR DCS
Action:
Ross, AE0D/
/
\\
UNITED STATES
~
NUCLEAR REGULATORY COMMISSIO Cys:
Callan
~.
j
,E O
Thompson W A$HINGTON. D.C. 20555 Jordan g
p,e q
Norry June la,, 1997 Blaha N
Lieberman, OE OFFICE OF THE SECRETARY
]g@
Halman, ADM 1 (Aed G
Allison, AEOD Shelton, IRM MEMORANDUM TO:
L. Joseph Callan Meyer, ADM Exec tive D* rect.or for Operations La-- -
FROM:
Jo C.
oyl Secretary
SUBJECT:
STAFF REQUIREMENTS - SECY-97-101 - PROPOSED RULE, 10 CFR SECTION 50.76, " REPORTING RELIABILITY ANE AVAILABILITY INFORMATIOh FC.,
RISK-SIGNIFICANT SYSTEMS AND EQUIPMENT" The Commission has approved the staff's recommendation to accept the voluntary approach proposed by NEI for obtaining reliability and availability data for key safety systems.
The staff should continue to work with industry representatives to improve the content of the voluntary data.
These improvements should seek to expand the voluntary program to minimi::e 1) uncertainty in data and 2) the use of compensatory measures to derive needed parameter estimates.
The staff should periodically update the Commission on its efforts to work with industry to address shortfalls and limitationo in the data, and advise the Commission on whether the voluntary approach remains a viable
{
method of meeting regulatory needs.
This periodic update may be incorporated, as appropriate, in the quarterly updates to the PRA implementation plan.
(EDO-) (AE0D/RES)
(SECY Suspense:
9/26/97)
- 9500047, rec SECY NOTE:
THIS SRM,.SECY-97-101, AND THE COMMISSION VOTING RECORD CONTAINING THE VOTE SHEETS OF ALL' COMMISSIONERS WILL BE MADE PUBLICLY AVAILABLE 5 WORKING DAYS FROM THE DATE OF THIS SRM.
l
~
QUARTERLY STATUS UPDATE OF THE AGENCY-WIDE -
IMPLEMENTATION PLAN FOR PROBABILISTIC RISK ASSESSMENT (PRA)
(from June 30,1997 to September 30,1997)
J
SUMMARY
OF SIGNIFICANT PROGRESS
'(1)
Reaulatorv Guide (RG) and Standard Review Plan (SRP) Develooment (Tasks 1.1 and 2.dl On April 8,1997, the staff sent to the Commission SECY-97-077, " Draft Regulatory Guides, Standard Review Plans and NUREG Document in Support of Risk-Informed Regulation for Power Reactors." SECY-97-077 requested Commission approval to publish for comment four draft Regulatory Guides (RGs), three draft Standard Review Plan (SRP) sections, and one draft i
NUREG series report that support implementation of risk-informed regubtion for power reactors. By Staff Requirements Memo andum (SRM) dated June 5,1W7, the Coraminsion approved publication of the draft documents. A notice was placed in the Federal Register announcing availability of the documents and requesting public comment on them.
Pub'ic Workshoo on Reaulatorv Guides and Standard Review Plans To facilitate solicitation of public comments, the staff held a workshop on Augu:,t 11,12, and 13,1997, at the DoubleTree Hotel in Rockville, Maryland to explain the draft documents and a.1swer questions. The workshop was well attended by industry representatives. They offerad a number of constructive comments, some criticisms, and some suggestions for changing the guidance. Overall, the comments indicated general support for pursuing risk-informed regulation but in a manner which would necessitate some modifications to the draft guidance. The more significant issues raised during the workshop regarding the general regulatory guidanen included:
how the guidelines on CDF and LERF would be applied when proposed increases in i
risk are very small; the conditions under which a full scope PRA would be necessary; I
what constitutes a " quality PRA" and the role of NUREG-1602 in judging the quality of the PRA supporting an application; having separate acceptance guidelines for accident sequences initiated during power operation and sequences initiated during low-power and shutdown operations;
- having new industry /NRC pilot programs to ensure the effectiveness of the guidance issued for use.
i The staff is reviewing the comments provided at the workshop and those formal written public comments it has received.
i A2i
Draft Reaulatorv Guide and Standard Review Plan for Inservice insoection The staff completed new drafts of the application-specific RG and SRP for inservice inspection (ISI) and discussed them with senior agency manageinent, the Advisory Committee on Reactor Safeguards (ACRS) and the Committee to Review Generic Requirements (CRGR)in a number of meetings held over the past three months. Both the ACRS and the CRGR have completed their reviews of the guidance and concur with the staffs proposal to issue the guidance for comment by the public. On August 20,1997, the staff sent to the Commission SECY-97-190,
" Draft Regulatory Guide and Standard Review Plan on Risk-Informed Inservice Inspection of Piping." SECY-97-190 requested Commission approval to publish for comment the RG and SRP that supports implementation of risk-informed inservice inspection programs. Commission approval was obtained in an October 1,1997, SRM. In completing the draft RG and SRP, the staff has found that a greater than expected effort was required to incorporate all points of view and gain a consensus on draft guidance. With this experience, the staff projects that the schedule for issuing the final ISI RG and SRP will slip from February 1998 to April 1998.
(2)
Pilot Acolications frask 1.2)
For risk-informed ISI programs, the industry had identified three pilot plants that would submit requests for authorization for the use of risk-informed ISI methodology; these applications have not yet been received. In a letter from NEl, dated August 29,1997, the industry requested to add two additional plants to the list of pilot applications and identified an aggressive schedule for all the pilot plants. The scope for the two new pilot plants is limited to Class-1 piping (primary coolant system piping only). To date, only one application has been submitted by one of the pilot plants with a limited scope RI-ISI prcgrarn. Due to industry's delays in submitting the applications, and the addition of two plants as pilots, the staff is unable deve:cp an integrated review schedule for the pilot plants at this time. This schedule is contingent on information regarding the actual timing of submittals, the quality of the submittals, arid the ability of the pilot plant licensees to commit the resources necessary to respond to the staffs RAls.
As noted in an August 21,1997, memorandum to the Commission, completion of the RI-IST pilot plant safety evaluation has been delayed. Nevertheless, between July 14 and 18,1997, the NRC staff and its contractors reviewed PRA models, backup calculations, and data at Comanche Peak Steam Electric Station (CPSES). The review was conducted as part of the staffs evaluation of Texas Utilities Electric Company's (TUE's) proposed RI-IST program and was aimed at determining whether the CPSES PRA is consistent with the quality and scope guidelines in draft Regulatory Guide DG-1061, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Current Licensing Basis." While the review team identified some minor weaknesses with the CPSES PRA for the RI-IST application (e.g., missing success paths, limited documentation of human error probabilities, optimistic recovery factors for equipment repair, plant-specific performance data not having already been incorporated into the PRA), the review team feels that these issues can be addressed adcquately by the licensee. The staff also identified an area in the calculation of sequence success that needs further clarification. The calculated core damage frequency from the licensee's base PRA will approach 1x10d per year when extemal event initiators and shutdown operations are taken into account. Thus, the licensee's proposed A2-il
~
j-.
- RI-IST program is receiving increased NRC technical and management review in accordance with guidance in DG-1061.
i In response to staff questions and concems, South Texas Project (STP) submitted for staff review another Operational QA Program (OQAP) revision, revised procedures for implementing facets of the graded QA program, a proposed Final Safety Analysis Report (FSAR) revision that would invoke 10CFR50.59 change controls on the GQA implementing procedures, and responses to staffinformation requests. The staff has prepared a safety evaluation for graded QA based on the reviews performed, which will be sent to the Commission via a separate Commission paperin October 1997.
The staff has been working with the Combustion Engineering Owners Group (CEOG) to develop a TS administrative control for a configuration risk management program (CRMP). The CRMP constitutes the third tier of the three-tiered approach the staff has used in reviewing risk-informed TS allowed outage time (AOT) changes. As discussed in SECY 97-095, the staff is requiring licensees to incorporate a commitment to imp lement a CRMP in the TS'as part of the basis for its approval of risk-informed TS AOT changes. Once the staff reaches agreement with the CEOG on a TS administrative control for the CRMP, and commitments are received from the individual pilot licensees, the staff will issue amendments to the lead plant and the other CE pilot licensees that have review results comparable to those for ANO-2, as discussed in SECY-i 97-085 and endorsed by the Commission in its May 28,1997, SRM.
(3)
Insoections (Task 1.3)
Significant PRA-related technical support has been provided for the agency's Maintenance Rule baseline inspections. ' As of September 30,1997, the staff has performed 36 full inspections.
These inspections were performed with the support of experienced staff and contractor
)
personnel trained in the use of PRA, using an inspection procedure that focuses on the inspection and assessment of the relevant PRA-related technical aspects of the NRC-approved j
industry guideline for implementing the rule (i.e., NUMARC 93-01).
New technical guidance on the use of PRA in the power reactor inspection program has been issued with the revision of Inspection Manual Chapter 2515, Appendix C.
The NRC staff briefed the Commission in May 1997 on the Individual Plant Examination (IPE) irisights report, NUREG-1560. As a result of this briefing, the staff received a SRM requesting j
the staff "to expedite activities in the following areas: (1) using IPE results to prioritize inspection activities; (2) improving regional capabilities for the use of PRA and risk insights; and (3) providing related inspector training." The staff has been active in accomplishing each of the three items as discussed below, i
Since June 1995, briefings on IPEs have been made by the Office of Nuclear Regulatory Research (RES) staff to all four regional offices. To date, the majority of IPEs reviewed by the staff have been covered in the briefings. In addition, a detailed briefing (tailored for each l
region) of the results and insights from NUREG-1560 was presented at each region. These briefings (both types) have been attended by the resident inspectors, regional personnel, and j
plant inspection teams (where applicable). The briefings have been specifically structured to
)
A2iii
.__=
__ _ _ _ _ _ _ _. _ _ _ _ _ _ _ _ ~..
aid in prioritizing inspection activities, and to provide guidance on how to use PRA resu many cases, Senior Reactor Analysts (SRA), on assignment in RES, participated in the preparation and presentation of the briefings as part of their developmental training.
Coasequently, the SRAs have gained a solid knowledge of the variety of information c ir a PRA.
The briefings on IPEs have provided valuable insights, particularly in plant-specific inspect activities. However, since many of the licensee's IPEs are out of date, inspectors will need to use supplementary information, including current licensee PRAs, as available,' to draw appropriate inspection insights. Consequently, plant-specific briefings based on the IPEs have been discontinued. Instead, the SRAs in each Regional office and in NRR, who are now either fully trained and certified, or are in training, will continue to provide ongoing PRA advice for site-specific activities with support from headquarters offices as needed. Regiona and headquarters SRA activities include: providing risk-based inspection prioritization, event assessment and inspection follow up, maintenance rule inspection support, inspection procedure guidance development, and maintenance of the SRA homepage on the NRC intranet.
(4) Accident Manaaement (Task 1.9)
The staff ieview of the IPE submittals included an assessment of licensee responses to the requests in GL 88-20 and NUREG-1335 related to accident management. Based on IPE insights, the staff has not identified any areas where immediate industry actions related to accident management appear necessary. However, the following accident management area <,
raised in the IPE submittals warrant further staff evaluation:
Inhibiting ADS in boiling water reactors (BWRs)
Use of drywell sprays to prevent Mark I containment liner failure Preclude terminating injection to the reactor from extemal sources Effectiveness of extemal reactor vessel cooling These follow-up items will be addressed in the staffs evaluation of the BWROG Emergency Procedure and Severe Accident Guidelines (FP/ SAG) described in SECY-97-13 (5) Evaluatina IPE Insiahts To Determine Necessarv Follow-uo Activities As part of finalizing NUREG-15G0, " Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance," the staff has defined an initial set of follow-up acti This initial set of activities consists of:
'Since there is no requirement for licensees to submit or update PRAs, actual availability to staff in inspection or other activities is on a case-by-case basis. The staff will investigate options for addressing this issue.
A2iv
l I
i Additional analysis of plants identified in NUREG-1560 as having risks approaching the Commission's quantitative health objectives (QHO's), based on a preliminary screening analysis. This additional analysis will use updated l
. information and refined methods to make a better comparison with the QHOs; Analysis of plants with accident sequence frequencies greater than 1x10-5 per e-reactor year and/or conditional containment failure probabilities greater than 0.1.
The analysis will evaluate whether thr is plants have features which merit backfit consideration. This will be done in manner consistent with the Safety Goal screening assessment in the Commission's Regulatory Analysis Guidelines; e
Analysis of selected generic issues which may merit further staff evaluation, such as:
Contributors to station blackout including grid unreliability RCP seal LOCA and its associated contribution to core damage and large early release frequency, Steam generator tube rupture; e
Follow-up on whether the actions licensees stated they were taking as a result of their IPE have, in fact, been taken; 4
i e
Follow-up on selected licensee responses to Containment Performance Improvement questions included in GL-88-2), Gupplements 1 and 3.
In conjunction with this effort, the staff is deyeloping a pian for audit of licensee-identified improvements credited in IPE analyses, to determine the effectiveness of licensee actions to reduce risk. The schedule for finalizing the list of items and a program plan to address those items is scheduled for completion in November 1997.
In a May 21,1996, Staff Requirements Memorandum (SRM), the Commission requested that the staff track the regulatory uses of IPE/IPEEE results. Additionally, the Commission noted that consideration should be given to linking the resulting IPE/IPEEE databases together in a e Je, integrated, coherent program. This task was placed under item 1.10 of the PRA tc,e.ementation Plan in the October 1996 update, and a structure and linking process is under development. The staff will discuss the database content in the next quarterly implementation Plan update. Due to other staff priorities, such as support for pilot applications and risk-informed regulatory documents, and delays in authorizing contract funds, the target schedule for defining uses for risk information, clarifying regulatory use, and assessing methods of data collection has been revised from December 1997 to May 1998.
(6)
Methods Development and Demonstration (Task 2.4)
The Seabrook nuclear power plant is participating in the first trial PRA application of ATHEANA (A Technique for Human Ever.t Analysis). ATHEANA is a human reliability analysis (HRA) i A2v
method under development in RES which addresses errors of commission as well as omission.
It focuses on combinations of performance shaping factors and plant conditions which increase 4
the likelihood of certain human errors. In addition to identifying unsafe acts that will be
]
considered for quantification within the plant PRA model, ATHEANA is showing promise for i
identifying strategies for improving human reliability, 3
in response to concems over the influence of management and organizational factors, safety culture, and downsizing and deregulation on human performance and safe plant operations, RES held a workshop in August 1997 to discuss these issues with nationally and intemationally recognized leaders in management and safety issues (including expeits from academia, utilities, nationallaboratories, consulting cornpanies, the NRC, DOE, and NASA). The experts presented papers and results of current research and participated in working sessions on these topics. The products of this workshop will be used to suggest research methods and/or to
~
l assess the influences of management and organizauonal factors, safety culture, anci the effects of downsizing and deregulation.
Individual Plat Examination (IPE) and IPE of Extemally Initiated Events (IPEEE)
(7)
Reviews (Task 2.5)
Status of IPE Reviews The reviews of all 75 original IPE submittals (i.e., not including Browns Ferry, Unit 3) have been completed with a staff evaluation report (SER) issued by RES to NRR for each submittal. With the exception of the Crystal River and Susquehanna IPEs, all IPE submittals have now been found to meet the intent of Generic Letter 88-20. The licensees for Crystal River and Susquehanna plan to submit revised IPEs that would address the staff concems. It is expected that these two revised IPEs will be submitted to the staff by December 1997 and staff review will be completed by June 1998.
Preliminary review of the recently submitted IPE for Browns Ferry, Unit 3, and responses to a staff request for additional information have been completed. It is expected that RES will issue its SER for the Browns Ferry, Unit 3, IPE by December 1997.
IPE Insichts Reoort in October 1996, the staff issued draft NUREG-1560, " Individual Plant Examination Program:
Perspectives on Reactor Safety and Plant Performance," for public comment. Comments were received from numerous licensees, individuals and other government organizations. Overall, the comments received were positive in nature. The staff has developed responses to all the comments received and, where appropriate, draft NUREG-1560 has been revised. Attached for information are the executive summary of NUREG-1560 (Attachment 5) and Appendix C of NUREG-1560 which documents resolution of comments from the public (Attachment 6). The final version of NUREG-1560 will be published in November 1997.
A2 vi l
-=
-,___m
-.__ -, _ _ _ _ _ -+---
w
.wrw--
ey--
r--
y
-y
1 1
Status of IPEEE Reviews i
i Of the 74 expected IPEEE submittals, the staff has received 63; four of which were not complete. Currently,49'submittals are under various stages of review. Nine additional submittals are expected to be received bylhe end of December 1997, one by June 1998, and the submittal date of one iPEEE has yet to be determined. The staff will complete all IPEEE reviews and Staff Evaluation Reports (SERs) by June 1999. Similar to the IPE program, the staff will take prompt action should any sigmficant vulnerabilities or safety insights be identified a
in these reviews.
An interim report has been developed that provides preliminary IPEEE perspectives and summarizes the information presented in the first 24 IPEEE submittals reviewed by the staff.
This interim report will be sent to the Commission by the end of November 1997. A summary of the significant preliminary perspectives from the first 24 IPEEE reviews is presented in
,. In addition, a draft report that summarizes the findings and perspectives from all IPEEE reviews will be sent to the Commission in June 1999 and released for public comment.
After receipt and review of comments, the staff will issue the final IPEEE insights report in -
December 1999.
(8)
Risk-Based Trends and Patterns Analvsis Task (3.1)
The Common Cause Failure (CCF) database has been updated with events through 1995. The database and its associated technical reports are being provided on CD ROM to all nuclear utilities in accordance with the INPO agreement regarding distribution of NPRDS proprietary data. Initial draft reports for the initiating event update, loss of offsite power study, the auxiliary feedwater system study, and the Westinghouse reactor protection system study have been received by AEOD from its contractor and reviewed with the ACRS. In July 1997, the BWR high pressure core spray system draft report was distributed to NRC staff for intemal peer review. Comments have been received and are being incorporated into the final report.
(9)
Accident Seauence Precursor (ASP) (Task 3.2) 4 The last 1996 preliminary ASP analysis has been sent to the licensee for review. Four 1996 final analysos are completed and have been sent to the respective licensees and have been made publicly available. Seven precursor analyses are under licensee or AEOD final review.
Events of significance, i.e., those with a conditional core damage probability (CCDP) greater than 1x104, include a Catawba event (loss of offsite power with diesel failure), a Wolf Creek event (frazil icing of the ultimate heat sink), a Seabrook event (long term unavailability of an emergency feedwater turbine-driven pump), a Prairie Island event (loss of offsite power to safeguards buses at both units), and a Haddam Neck event (potential inadequate residual heat removal pump net positive suction head following a medium or large loss of coolant accident).
(10). Comoile Ooeratina Exoerience Data (Task 3.5)
INPO submitted a revision to the Memorandum of Agreement with the NRC regarding access to the EPIX database. The EDO indicated his agreement with minor modifications and sent it to i
INPO on August 21,1997, for signature.
4 A2 vii
]
l'
The Sequence Coding and Search System (SCSS) conversion from a mainframe computer to a PC-based computer has been completed as well as the beta testing of the new system. Direct access capability to SCSS via the Intemet is now functional. Training and direct use for all NRC staff will be implemented by December 31,1997.
(11)
Staff Trainina (Task G.6)
A new course, "PRA for Technical Managers," has been added to the curriculum and two presentations were held in FY 1997. This course is designed to provide all levels of staff j
managers with a basic understanding of PRA methods, strengths, and limitations needed to implement risk-informed, performance-based regulations. Current plans are to present this course seven times in FY 1998 in headquarters.
PRA Level 2 and Level 3 courses have been added to the PRA curriculum. The first presentation of the new PRA Level 2 course, " Accident Progression Analysis," was held in February 1997. This three-day course addresses accident phenomenology under post core damage conditions and development of PRA models for this severe-accident regime. Based on feedback from the first presentation of the course, the course is undergoing significant modification. The PRA Level 3 course, " Accident Consequence Analysis," was " dry-run" in early 1997 and a first presentation was given in September 1997. The three-day course addresses environmental transport of radionuclides and the estimation of offsite consequences from core damage accidents. Current plans are to present each course twice a year.
A new course on extemal events has been completed. This three-day course addresses extemal events (such as fires, floods, earthquakes, high winds, and transportation accidents) and the development of external-event PRA models such as those used in the IPEEEs. The first presentation of this course was held August 5-7,1997.
A new course, "PRA Technology and Regulatory Perspectives", is under development and i
scheduled for first presentation in January 1998. A pilot presentation of the course was given on September 22-26,1997. The course was originally scheduled to start in October 1997.
- However, based on the pilot presentation, further development and refinement of the course necessiated its delay to January 1998. The course will replace the PRA Basics for Regulatory Application course and the Insights into IPEs course for some basic level users.
In a June 5,1997, Staff Requirements Memorandum, the Commission requested information on the plans for training NRC staff on 1) the risk-informed regulatory approach (es) contained in the regulatory guidance and Standard Review Plan documents and 2) overall PRA methods and techniques. Attachment 4 of this Commission paper provides the staff's response to the SRM, and describes the training that will be necessary to implement the initiatives discussed in the draft RGs and SRPs for risk-informed regulation.
l
)
l A2 viii
E' REVISIONS TO THE EXISTING PRA IMPLEMENTATION PLAN (1)
Risk-Iniormed Reaulatorv Guides and Standard Review Plans (Tasks 1.1 & 2.1) in completing the draft RG and SRP for risk-informed inservice inspection (ISI), the staff -
has found that a greater than expected eRort was required to incorporate all points of
. view and gain a consensus on draft guidance. With this experience, the staff projects
_ that the final ISI RG and SRP wtii be delayed from February 1998 to April 1998.
(2)
Risk-informed Pilot Aeolications (Task 12) inservice insoection With respect to the risk-informed ISI programs, the staff expected but has not received a formal submittal from the three pilot plants (Surry, ANO-2 and Fitzpatrick). Based on an NEl letter, dated August 29,1997, the staff anticipates receiving pilot applications to implement RI-ISI programs through the winter of 1998. This includes applications from two new pilot plants (ANO-1 and Vermont Yankee). The staff continues to hold public working meetings with the industry and with Virginia Power on the Surry pilot, in anticipation of receipt of a formal application and to facilitate the development of regulatory guidelines.
- Due to industry's delays in submitting the application for Fitzpatrick an'd ANO-2, and the addition of two plants as pilots, the staff is unable to develop an integrated review schedule for the pilot plants at this time. The schedule is contingent on information regarding the actual timing of submittals, the quality of the submittals, and the ability of the pilot plant licensees to commit the resources necessary to resNnd to the staff's j.
RAls.
i i
With respect to the EPRI method (EPRI-TR-106706), the staff has not received l
' responses to its RAls. The EPRI method is used by all of the pilots except Surry.
Delays in receiving responses to the staff's RAls could also impact the schedule for the review of the pilot plants.
j inservice Testina 4
i in a June 17,1997, memorandum to the Commission, the staff stated that it exp?cted to l
be able to issue the safety evaluation (SE) on the Com#ie Peak RI-IST program in October 1997. The October 1997 completion date fori s Comanche Peak SE was l
based on the assumption that TU Electric Company (TUE) would respond satisfactorily to both the second and final round requests for additional information by August 8,1997.
The staff's final round RAI asked the pilot plant licensee to describe how their proposed Rl-IST program comports with the draft Rl-IST RG and to explain their rationale for any 2-differences.
In a letter to the NRC dated July 31,1997 (amended by letter dated September 12,
{
1997), TUE stated they need additional time to determine how Comanche Peak's RI-IST A2-ix
program comports with the NRC draft guidance. TUE indicated that the resources required to complete the RAls are also being used to provide support for (1) the third refueling outage of Comanche Peak Unit 2 in the fall of 1997, and (2) the NRC Maintenance Rule baseline inspection scheduled at Comanche Peak in October 1997.
TUE plans to respond to the second and third round RAls by September 30,1997. This will delay issuance of the SE on the Comanche Peak RI-IST program until at least late December 1997.
In a letter to the NRC dated August 1,1997, Arizona Public Service Company (APS) informed the staff that its resources must be diverted from the Palo Verde Rl-IST program development effort in order to complete other activities (e.g., the 10 year IST program update and improved technical specification implementation). Therefore, due to the resource constraints and operational priorities discussed above, APS indicated that they will not be in a position to resume supporting the RI-IST implementation effort until mid-1998. At that time, APS will provide the NRC staff with a schedule for responding to the third RAl.
Graded Quality Assurance Task 1.2 of the PRA Implementation Plan states that the target schedule for completing the graded QA safety evaluation (SE) for STP is July 1997. The South Texas Project (STP) is the only graded QA volunteer plant that submitted a revised graded QA program for staff review and approval. The staff has prepared safety evaluation for the STP program that will be transmitted to the Commission in a separate Commission paper in October 1997. Dialogue with STP on several issues as well as competing priorities for staff resources delayed completion of the safety evaluation from July to September 1997. Staff monitoring of activities at all three volunteer plants (STP, Grand Gulf, and Palo Verde) will continue in order to observe the results of equipment categorization for additional systems, and the results of the application of graded QA controls and to assess the integrity of the corrective action and operational performance feedback programs. This monitoring effort is expected to continue for an extended period (several years) to provide the staff with lessons teamed.
For the purposes of the PRA Implementation Plan, this phase of volunteer plant interactions will be considered complete when the GQA RG and inspection procedure (IP) are issued in final form. In the future, the staff will continue to monitor the vo!unteer plant GQA implementation, gain feedback to revise the RG and IP as warranted, and evaluate GQA implementation strategies for other licensees who choose to pursue GQA. Although issuance of the SER for the STP GQA program is expected in October 1997, the completion date for the GQA pilot application remains March 1998 to reflect j
the expected schedule for issuance of the final GQA inspection procedure.
j New Pilot Anolications In parallel with the NEl initiatives to study the risk and cost of regulated activities (see Task 2.7 "Whole-Plant" Risk Studies), the pilot plants will be submitting license A2 x
~
amendment applications related to diesel generator start time and hydrogen control.
Specific schedules will be established when each application is received.
(3)
Insoections (Task 1.3)
As discussed just above, work has been initiated on en inspection procedure for Graded Quality Assurance (GQA). However, because of higher priority work on the South Texas GQA program safety evaluation, the schedule for completing a draft inspection procedure for GQA has been changed from September 1997 to December 1997 with completion of the final guide in March 1998.
Due to personnel being reassigned to higher priority activities, such as development of the PRA for Regulatory Applications course, the completion date for the review of core inspection procedures has been revised to October 1997.
(4)
Acolication of IPE/IPEEE to Generic issue Resolution (Task 1.6)
The completion target for identifying generic issues to be audited and selection of plants to be audited has been revised from "TBD" to December 1997.
(5)
"Whole-Plant" Risk Study (Task 2.7-New)
In a letter to the NRC, dated August 21,1997, the Nuclear Energy Institute (NEI) made a proposal for three new risk-informed pilot applications of PRA in support of changes to
)
the licensing basis of operating nuclear power plants. The staff met with NEl on September 17,1997, to discuss the proposal, including potential NRC activities. The pilots would use a full scope PRA to assess risk versus the regulatory requirements and plant operating and maintenance costs. The staff has concluded that, in concept, the initiative is worthwhile and plans to meet with NEl in November to discuss plans for pursuing the initiative. In parallel with the "whole-plant " risk studies, the pilot plant licensees will be submitting license amendment applications on issues such as diesel generator start time and hydrogen controls. These items will be tracked under PRA Implementation Plan Task 1.2 " Pilot Applications for Risk-informed Regulatory initiatives."
(6)
PRA Standards Develooment (Task 2.8-New)
In June,1997, NRC staff met with representatives of the American Society of Mechanical Engineers (ASME) to discuss cooperation with both industry and professional societies to develop new codes and standards, as directed in the SRM on Direction Setting issue (DSI) 13, dated March 7,1997 (see SRM in Attachment 1). The development of PRA standards was one subject of this meeting. At the meeting ASME indicated their interest and is convening an ad hoc committee that will have the
)
responsibility to develop such a standard. This committee will be comprised of ASME personnel, NRC staff, national laboratory, academic, and industry personnel.
A2-xi
A charter for this committee is now being drafted, and will describe the goals and objectives of the committee, committee membership and associated responsibilitie.s, schedules, and milestones. An addition, the charter willinclude anticipated scope of the standard (e.g., Level 1,2 and 3 PRA, including core damage accidents initiated by intemal and extemal events during full power operation) and the level of PRA modeling and analysis appropriate for differentPRA uses. The Commission will be informed of progress on this development work in the quarterly updates of the PRA implementation Plan.
(7)
Low Power Shutdown Rid Reevaluation (Task 2.9-New)
RES has been assigned responsibility to further investigate methods for estimating the a
risk of severe accidents initiated during low power and shutdown operational states.
The results of this investigation could include, for example, staff activities such as the development of new analysis methods or performance of experiments.
The staff intends to complete planning for this investigation in FY1988. Consistent with agency resources allocations, defined work will begin in FY1999.
(8)
Revision of Safety Goal Poliev Statement (Task 2.10-New)
SECY-97-208 discusses a number of issues relating to possible revision of the Safety Goal Policy Statement, including the possible elevation of core damage frequency to a fundamental safety goal. The staff recommended that additional discussions with ACRS be undertaken, with a goal of providing a Commission paper by March 31,1998, which would include the staff's analysis, conclusions, and recommendations. This item has i
been inserted into the implementation Plan; however no specific actions will be taken until the SRM on SECY-97-208 is received.
)
(9)
Risk Basad Trends and Pattems (Task 3.1) 4 i
The date for the component studies (Task 3.1) has been delayed because the cognizant l
engineer has been detailed to the Millstone Project. The dates for the systems studies have been delayed due to difficulties in applying models to the various system designs in a manner consistent with the reportability of failures and demands in multiple train systems. The delay in the initiating events update is due to difficulty in interpreting the extent of loss of offsite power and the nature of some initiating events from LERs.
(10)
Accident Seauence Precursor Proaram (Task 3.2)
Schedules for development of low power / shutdown models and extemal events (earthquake and fire) models for use in the Accident Sequence Precursor Program are currently being revised to reflect NRC staff comments on the initial models and staff assigned to higher priority work.
i A2 xii
i ia j -
-(11)
Risk-Based Performance Indicators (Task 3.'4)
The delay in the development and implementation of risk-based performance indicators j
(Task 3.4) is due to the delays in the component and system studies. The outputs of I
these tasks serve as basic inputs for risk-based Pls.
j
_ (12) - Risk Assessment of Material Uses (Task 4.4)
'The work for developing PRA methods (Task 4.1) for use in evaluating medical devices i
containing nuclear material has been subsumed into the larger risk assessment of material uses (Task 4.4). A working group of NRC and Agreement States personnel has been chartered to:
identify and document a technical basis for a risk-informed approach to the regulation of nuclear byproduct material, and develop plans for a graded regulatory approach for nuclear byproduct materials, based on risk information.
There was an initial meeting of the working group in mid-June 1997. Additional meetings were held in July, August, and September 1997 and are expected to continue about monthly through September 1998. Contractor support is planned to be available by November 1997 to assist the working group in its activities. The expected completion date of working group activities is September 1998.
(12)
Nuclear Material Licensina and Hich-Level Waste issues (Task 4.5) in the SRM of April 15,1997, about risk-informed, performance-based regulation (DSI-12) the Commission directed the staff to (1) reexamine its risk-informed, performance-based or risk-informed, less prescriptive (RIPB) approaches with regard to nuclear material licensees and to high-level waste issues, to ensure that the needs of those licensees and those areas receive adequate consideration; (2) review the basis for nuclear materials regulations and processes to identify and prioritize those areas that are or, with minimal additional staff effort and resources, could be made amenable to RIPB regulation; and (3) develop a framework for applying PRA to nuclear material uses, similar to the one developed for reactor regulation (SECY-95-280), where appropriate, in a paper that will be transmitted to the Commission in October 1997, the staff will reexamine preliminarily the RIPB approaches that it has identified in the PRA Implementation Plan, primarily those for nuclear materials licensees and high-level waste issues, but also those for low-level wastes, spent fuel storage facilities, and transportation (the other activities included in the PRA implementation plan). Also, the staff will identify preliminarily other NMSS areas that are or, with minimal resources, can be made amenable to RIPB approaches. Finally, the staff will provide a plan for developing a framework for applying RIPB approaches in NMSS regulation.
A2 xiii
I l
REVISED TASK TABLES provides updates to reflect the progress and revisions to the PRA implementation Plan from July 1 to September 30,1997.
l
~
1 l
l l
l l
\\
1
)
f A2 xiv l
w
_m ATTACIIMENT-3 REVISED PRA IMPLEMENTATION PLAN TASK TABLE (September 1997) 1.0 REACTOR REGULATION l
Regulatory Activity Objectives Methods Target lead Schedule Ofrwe(s) 1.1 DEVELOP STANDARD
. Standard review plans for NRC staff to use in risk-
- Evaluate available industry guidance.
NRR REVIEW PLANS FOR informed regulatory decision-making.
/RES RISK-INFORMED
- Develop a broad scope standard review plan (SRP) chapters REGULATION and a series of application specific standard review plan r
chapters that correspond to mdustry initiatives.
5
- These SRPs will be consistent with the Regulatory Guides developed for the industry.
ISI 8S7C TS 4M7C
- Issue final SRP i
i Gemral 12M7 IST 12 S 7 ISI 4SS
{
TS 1267 i
i 1
l t
' C = Task Completed i
A3-i t
t i
I
k i
i Regulatory Activity Objectives Methods Target Lead Schedule OfTice(s)
NRR j
1.2 PILOT APPLICATION
- Evaluate the PRA methodolony and develop staff
- Interface with industry groups.
FOR RISK-INFORMED positions on emergin6. risk-inTormed initiatives, REGULATORY mcluding those ass %:sted with:
- Evaluation of appropriate documentation (e.g.,
I.2M6C-INITIATIVES
- 1. Motor operated valves.
10 CFR, SRP, Reg Guides, inspection procedures end 2.12M7 (TUE)
- 2. IST requirements.
industry codes) to identify elements entical to acitieving the TBD (APS) l
- 3. ISI ap......ts.
intent of existing requirements.
3.TBD l
- 4. 3/9g
- 4. Graded quality assurance.
- 5. Maintenance Rule.
- Evaluation ofindustry proposals.
5.965C
- 6. Technical specifications.
6a. Commission Approval
- Evatussion ofindustry pilot program implementation.
6a. $M7C r
6b.1267 6b. Pilot Amendmentsissued
- 7. Other applications to be identified later.
- As appropriate, complete pilot reviews and issue staff (applications related to diesel generator start times findings on regulatory requests.
and Ilydrogen Control are expected) t 1.3 INSPECf10NS
- Provide guidance on the use of plant-specific and
- Develop IC 9900 technical guidance on the use of PRAs in 6S7C NRR generic mformation from IPEs and other plant-the power reactor inspection program.
specific PRAs.
power reactor inspection program.
{
- Propose guidance options for inspection procedures related 10/97 r
to 50.59 cvaluations and regular mamtenance observations.
- Review core inspection procedur s and propose PRA 10/97 guidance where needed.
- Complete revision to proposed core inspection procedures 12M7
- Issue drall Graded QA Inspection Procedure for public comment 12M7 l
- Issue final Graded QA Inspection Procedure 3Mg
- Provide PRA training for inspectors.
methods., as input to AEOD/TTD for their development l
and refmement of PRA training for inspectors.
s
- Develop consolidated / comprehensive 2-3 week PRA for 10/97 NRR/
AEOD regulatory applications traming course.
- Conduct training for Maintenance Rule baseline RM6C NRR inspections
- Provide PRA training for Senior Reactor Analysts
- Rotationa: 9nments for SRAs to gain working Ongoing expenence i
A3-ii j
1 i
1
Regulatory Activity Objectives Methods Target Lead Schedule Ottice(s)
- Continue to provide expertise in risk assessment to
- Monitor the use of risk in inspection reports.
Ongoing NRR support regional inspection activities and to communicate inspection program guidance and
- Develop new methodologies and communicate examples of its implementation.
appropriate uses of risk insights to regional ofUccr.
- Update inspection procedures as needed.
- Assist regional offices as needed.
- Conduct Maintenance Rule baseline inspections 768 I A OPERATOR LICENSING Monitor insiehts from IIRAs and PRAs (including
- Revise the Knowledge and Abilities (K/A) Catalogs BS5C NRR tPEs and IPLIEEs) and operating experience to (NUREGs I122 and i123) to incorporate operating identify possible enhancements for mclusion in experience and risk insights.
planned revisions to guidance for operator licensing activitics (initial and requalification)
1.5 EVENT ASSESSMEN1
- Continue to conduct quantitative event assessments
- Continue to evaluate 50.72 events using ASP models.
Ongoing NRR of reactor events while at-power and during low power and shutdown conditions.
t
- Assess the desirability and feasibility of conducting.
- Define the current use of risk analysis methods and insights TDD NRR quantitative risk assessments on non-power reactor in current event assessments.
events.
- Assess the feasibility of developing appropriate risk assessment models.
- Develop recommendations on the feasibility and desirability ofconducting quantitative risk assessments.
l.6 EVALUATE USE OF PRA
- Audit the adequacy oflicensee analyses in IPEs and
- Identify generic safety issues to be audited.
1267 NRR/RES IN RESOLUTION OF IPEEEs to identify plant-specific applicability of
- Select plants to be audited for each issue.
1267 GENERIC ISSUES generic issues closed out based on trE and IPEEE
- Describe and discuss licensecs' analyses supporting issue TBD programs.
resolution.
- Evaluate results to determine regulatory response;i.e., no TilD action, additional audits, or regulatory action.
L r
A3-iii t
r 1
Methods Target lead Regulatory Activity Objectives Schedule Offu:e(s)
I 1.7 REGULATORY
- Assess the efTectiveness ormajor safety issue
- Develop proc:ss/ guidance for assessing regula*.ory ongoing NRR&
RES EFFECTIVENESS sesolution efforts for reducing risk to public health effectiveness.
EVALUATION and safety.
- Apply method to assess reduction in risk.
ongoing i
- Evaluate result, effectiveness of rules.
ongoing
- Propose modifications to resolution approaches, as needed ongoing
~
(SBO rule implementation and RCP sca* issue)
- Identify other assues for assessment if appropriate.
TBD t
i I
1.8 ADVANCED REACTOR
- Continue staff reviews of PRAs for design
- Continue to apply current staff review process.
Ongoing NRR REVIEWS certification applications.
\\
- Develop draR SRP to tech staff for review and concurrence.
6M8 NRR
[
certification reviews of evolutionary reactors 12/99 l
- Finalize SRP.
(ABWR and System 80+).
h
- Develop independent technical analyses and criteria
f RES for evaluating industry initiatives and petitions (NUREG-03%) using insights from NUREG-1150, the regarding simplification of Emergency Preparedness new source term information from NUREG-1465, and (EP) regulations.
availabic plant design and PRA information for the passive and evolutionary reactor designs.
i 1.9 ACCIIENT
- Develop generic and plant specific risk insights to
RES MANAGEMENT s
statiaudits of utility accidents management selected plants to serve as a basis for assessing
(
) programs at selected plants.
comp teness of utility A/M program elements (e.g., severe acci trammg) l i
I t
t t
i A3-iv 6
)
.,,r
~
Regulatory Activity Objectives Methods target trad Schedule Office (s) 1.10 EVALUATING IPE
- Use insights from the stafTreview ofIPIMo
- Review the report"lPE Program: Perspctives on 967C NRR &
INSIGIITS1D identify potential safety, policy, and techmcal Reactor Safety and Plant Performance and identify tne RES DETERhtINE issuce, to determine an appropr, ate course of initial list of required stafT and industry actions (if eny).
i NECESSARY FOLLOW-action to resolve these potential issues, and to includinginsights on A/M.
UP ACTIVITIES identify possible safety enhancements.
Finalize list of required statT and industry actions.
11/97
- Audit licensee improvements that were credited in the TilD NRR
- Determine appropriate approach for tracking the IPEs to determine effectiveness oflicensee actions to t egulatory uses ofIPE/IniEE results.
reduce risk.
- Define use for information, clarify " regulatory use", and
$/98 assess the most effective methods for data collection.
- If appro riate, develop approach for linking IPE/IPEEE 12/98 data bases.
A3-v
=
2.0 REACTOR SAFETY RESEARCII Target Schedule Lead Methods Otlice(s)
Regulatory Activity Ob.iectives i
- Draft PRA Regulatory Guides transmitted to Commission for 2.1 DEVELOP REGULATORY Regulatory Guides for industry to use in risk-informed approval to issue for public comment.
C GUIDES regulation.
General C
IST C
ISI C
GQA C
TS
- Issue final PRA Regulatory Guides.
12M7 General 12M7 IST 4M8 ISI 12M7 GQA 12M7 TS
- Provide technical support to agency users of risk users.
Continuing RES 2.2 TECllNICALSUPPORT assessment in the form of support for risk-based
- Expand the database of PRA models evailable for staff use, regulation activities, technical reviews, issue risk expand the scope of ciailable models to include external assessments, statistical analyses, and develop event andlow power and shutdown accidents,and rerme guidance for agency uses of risk assessment.
the tools needed to use these models, and continue maintenance and user support for SAPillRE and MACCS computer codes.
- St.pport agency etTorts in reactor safety improvements in Continuing RES former Soviet Union countries.
t l
A3-vi
m I Regulatory Activity Objectives Methods Target Lead Schedule OfUce(s) 2.3 SUPPORT FOR NRR
- Modify 10 CFR 52 and develop guidance on the use of
- Develop draft guidance and rule.
SM8 RES STANDARD REACTOR PRA updated PRAs beyond design certification (as described
- Solicit public comment.
1168 RES REVIEWS m SECY 93-087).
- Finalire staff guidance and rule.
1269 RES 2A METIlODS DEVELOPMENT
- Develop, demonstrate, maintain, and ensure the quality
- Develop and demonstrate methods for including aging efTects 9S8 RES AND DEMONSTRATION of methods for performing, reviewing, and using PRAs in PRAs.
and related techniques for existing reactor designs.
- Develop and demonstrate methods for fire risk analysis 968 RES
- Develop and demonstrate methods for assessing reliability / risk 6/99 of argital systems RES 2.5 IPE AND IPEEE REVIEWS
- To evaluate IPE11EEE submittals to obtain reasonable
- Complete reviews ofIPE submittals.
9M7 RES assurance that the licensee has adequately analyzed the
- Complete reviews ofIPEEE submittals.
6/99 RES plant design and operations to discover vulnerabilities;
- Contmue regional IPE presentations.
C RES and to document the significant safety insights resulting
- Issue IPE insights report for public comment.
1066C RES from IPE/IPEEEs.
- FinalIPliinsights report 9M7 RES
- Issue draft final IPEEE insights report 12M9 RES 2.6 GENERIC ISSUES PROGRAM
- To conduct Seneric safety issue management activities,
- Continue to prioritize and resolve generic issues.
Continuing RES including pnoritization, resolution, and documentation, for issues relating to currently operating reactors, for r
advanced reactors as appropriate, and for development or revision of associated regulatory and standards t
instruments.
2.7 NEl INITIATIVE TO CONDUCT
- Review NEl initiative to conduct three pilot"whole
- Agree on ground rules for study 168 RES/NRR "WIIOLE PLANT" RISK STUDY plant
- risk-informed studies of requirements vs. risk
- work with industry to develop national consensus
- Initiate activity 9M7C Ri DEVELOPMENT standard for PRA scope and quality
' Collect studies of LP&S risk as a benchmark for
' Collect and review existing LP&S risk information (domestic 988 RES BENCIIMARK RISK STUDY assessing the need for further staff activities and foreign) 10S 8
- Initiate additional work 2.10 SAFETY GOAL REVISION
' Assess need to revise Commission's Safety Goal to make
- Initiate discussion with ACRS TBD RES core damage frequency a funamental goal and make other changes
)
A3-vii
3.0 ANALYSIS AND EVALUATION OF OPERATING EXPERIENCE, AND TRAINING 3
Methods Target lead Regulatory Objectives Schedule OtBcc Activity 12/9g AEOD 3.I RISK-BASED
- Use tractor operating experience data to assess
- Trend performance of risk-important w...m /a TRENDS AND the trends and patterns m equipment, systems,
- Trend performance of risk-important systems.
12/98 PATIERNS initiating events, human performance, and important ANALYSIS accident sequence.
- Trend frequency of risk-important initiating events.
3/9g
- Trend human performance for reliability characteristics.
- Trend reactor operating experience associated with specific safety As Needed AEOD
- Evaluate the effectiveness oflicensee actions taken to issue and assess risk implications as a measure of sa'fety resolve risk significant safety issues.
perf" nance.
- Develop trending rnethods and special databases for use in
- Develop standard tiending and statistical analysis procedures for C
AEOD l
identified areas for reliability and statistical applications, AEOD trendy activities and for PRA applications in CCF-C other NRC ofhces.
- Develop special software and databases (e.g. common cause failure) for use in trending analyses and PRA studies.
Periodic updates
- Identify and rank risk significance of operational events.
other sources to obtain ASP events.
Sttfut.fu.t PRECURSOR (ASP)
- Perform independent review ofeach ASP analyses. Licensees and Annual PROGRAM NRC staff peer review of each analysis..
- report, AEOD Ongoing 4
- Complete quality assurance of Rev. 2 rimplified plant specific 3N7C RES models.
II/96C RES
- Complete feasibility study for low power and shriown models.
C RES
- Complete initial containment performance and consequence models.
7/99 RES
- Complete development of the Level 2/3 models II/0I RES
- Complete the Rev. 3 simplined plant-specific m;dels.
- Complete low power / shutdown modcIs l
t l
- Provide supplemental information on plant specific
Regions.
performance.
l A3-viii l
l
m Regulatory Activity Objectives Methods Target Lead Schedule O(Tice 33 INDUSTRY RISK
- Provide a measure ofindustry risk that is as complete as
AEOD TRENDS possible to determine whether risk is increasing, activities which use design and operatin6 experience to assess the decreasing, or remaining ccastant over time.
Implied level of risk and how it is changmg.
- Update plan for risk-based analysis of reactor operating exp-rience 3/98
- Implement program plan ele.nents which will include plant-6/99 specific models and msights from IPEs, component and system reliability data, and other risk-important design and operational data in an integrated frame work to periodically evaluate industry trends.
3.4 RISK-BASED
- Establish a comprehensive set of performance indicators
- Identify new or improved risk-based Pts which use component and C
AEOD PERFORMANCE and supplementary performance measures which are more system reliabilitv models & human and organizational performance INDICATORS closely related to risk and provide both early indication evaluation meth' ds.
o and confirmation of plant performance problems.
- Develop and test candidate Pts / performance measures.
9/00
- Implement risk-based Pts with Commission approval.
1/01 3.5 COMPILE
- Compile operating experience information in database
- Manage and maintain SCSS and the PI data base, provide oversight Ongoing AEOD OPERATING systems suitable for quantitative reliability and risk and access to NPRDS, obtain INPO's SSPI, compile IPE failure i
EXPERIENCE DATA analysis applications. Information should be scrutable to data, collect plant-specific reliability and availability data.
the source at the event level to the extent practical and be sufTicient for estimatmg reliability and availability
- Develop, manage, and maintain agency databases for Ongoing parameters for NRC applications.
reliabihty/ availability data equipment performance, initiating events,CCP, ASP, and huma.? performance data).
6/98
- Determine need to revise LER rule to climinate unnecessary and less safety-significant reporting.
6/98
- Determine need to revise reporting rules and to better capture ASP, CCF, and human performance events.
10/99
- Publish revised LER rule.
6 A3-ix
i Target Lead Methods Schedule Ofrece(s)
Regulatory Activity Objectives
- Continue current contracts to present courses as scheduled.
Ongoing AEOD
- Present PRA curriculum as presently scheduled for FY
- Maintain current reactor technology courses that include PRA Ongomg 3.6 STAFF TRAININO 1996 insights and applications.
Ongoing
- Improve courses via feedback.
- Review cunent PRA course material to ensure consistency with Complete Appendix C.
C RES and
- Prepare course material based on Appendix C.
C AEOD
- Develop and present Appendix C training courses.
- Present courses on Appendix C.
C AEOD
- Review JTAs performed to date.
- Determine staff requirements for training, including C
- Perform representative JTAs for staff positions (11 A Pilot analysis of knowledge and skills, needed by the NRC C
Program).
- stalT,
- Evaluate staff training requirements as identified in the PRA Implementation Plan and the Technical Training Needs Survey C
(Phase,2) and incorporate them into the training requirements analysts.
C
- Analyze tha results of the JTA Pilot Program and determine C
requirements for additional JTAs.
Ongoing
- Complete J FAs for other stalTpositions as needed.
- Solicit a review of the proposed training requirements.
- Finalize the requirements.
Ongoing AEOD
- Revise current PRA curriculum and develop new training
- Prepare new courses to meet identified needs.
Ongoing
- Revise current PRA courses to meet identified needs.
9/97C program to futfill identified statineeds.
- Revise currer:t reactor technology courses as necessary to include Ongoing additional PRA insights and applications.
- Establish contracts for presentation of new PRA curriculum.
Ongoing AEOD
- Present revised PRA training curricutum.
- Present revised reactor technology courses.
Ongomg Ongomg
- Improve courses based on feedback.
A3-x r
' ' ~ ^ - - - - - - ---.
~
4.0 NUCLEAR MATERIALS AND LOW-LEVEL WASTE SAFETY AND SAFEGUARDS REGULATION Regulatory Activity Objectives Methods Target Lead 1
Schedule Office (s) 4.1 Validate risk analysis
- Validate risk analysis methodology
- lloid a workshop consisting of experts in PRA and 8S4 NMSS methodology developed to developed to assess the relative profile of IIRA to examine existing work and to provide C
assess most likely failure most likely contributors to misadministration for recommendations for further methodo!ogical modes and human the gamma stereotactic device (gamma knife).
development.
performance in the use of 965 mdustrial and medical
+ Examine the use of Monte Carlo simulation and its C
radiation devices, application to relative risk profiling.
965
- Fsamine the use of expertjudgement in devsloping C
err a rates and consequence measures.'
- Continue the development of the relative risk
- Develop functionah based generic event trees.
TBD RES/
methodology, with the addition of event trec NMSS modeling oTthe brachytherapy remote afterloader.
- Extend the application of the methodology
- Develop generic risk approaches.
- RES, and its further development into additional NM5S devices, including teletherapy and the pulsed high dose rate afterloader.
4.2 Continue use of risk
- Develop decision criteria to support
- Conduct enhanced participatory rulemaking, to 864 PR RES &
assessment of allowable regulatory decision making that establish radiological critena for decommissioning C
NMSS radiation releases and doses incorporates both deterministic ard risk-nuc! car sites; tecanical support for rulemaking Final Rule associated with low-level based engineeringjudgement.
including comprehensive nsk based assessment of Published radioactive waste and residual contamination.
7S7C residual activity.
- Develop guidance for implementing the radiological 268 criteria for license termination Ongoing
models for evaluating risks and attemative remediation methodologies. (Risk harmoniration).
4.3 Develop guidance for the
- Develop a Branch Techr.ical Position on
- Solicit public comments SS7 C.
NMSS &
review of risk associated with conducting a Performance Assessment of a
- Publish final Branch Technical Position TBD.
RES waste repositotics.
LLW disposal facility.
Dependent on I
Resources A3-xi
Regulatory Activity Objectives Methods
. Target Lead Schedule OITsce(s) 4.4 Risk assessment of material
- Develop and demonstrate a risk assessment
- Develop and demonstrate methods for determining 7/98 uses.
for industrial gauges containing cesium-137 the risk associated with industrial gauges containing and cobalt-60 using PRA and other related cesium-137 and cobalt-60.
techniques.
- Final report as NUREG
- 10/98
- De assessment should allow for modification based on changes in regulatory requirements.
- Use empirical data as much as practicable.
i
- Desclop and demonstrate risk assessment
- Working Group with contractor assistance to identify and methods for application to medical and document a technical basis for a risk-informed approach to industriallicensee activities.
the regulation of nuclear byproduct material and to develop plans for a graded approach to nuclear byproduct maecrial regulation based on risk informaitan.
- develop a framework for applying PRA to Regulating Nuclear Materials nuclear material uses similar to the one developed
- Complete Framework TBD for reactor regulation (SECY-95-280), where appropriate.
I t
i i
A3xs
5.0 tilGil-LEVEL NUCLEAR WASTE REGULATIDN Regulatory Activity Objectives Methods Target trad Schedule OfTice(s) 5.1 REGULATION OF filGll-
- Develop guidance for the NRC and CNWRA
- Assist the staffin pre-licensing activitics and in Ongoing NMSS LEVEL NUCLEAR WASTE staffs in the use of PA to evaluate the safety of license application reviews.
11LW programs.
- Develop a technical assessment capability in total-system and subsystem PA for use in licensmg and pre-Incensmg reviews.
- Combine avecialized technical disciplines (errth sciences andengineerin h those of system modelers to improve me logy.
- Identify significant events, processes, and
- Perform sensitivity studies of key technical issues Ongoing NMSS parameters affecting total system performance.
using iterative performance assessment (IPA).
- Assist the statito maintain and to refine the Ongoing NMSS evaluate proposed changes to regulations regulatory structure in llLW disposal regulations that governing the potential repository at Yucca pertain to PA.
Mountain.
- Apply IPA analyses to develop a site-specific regulation for a Yucca Mountain site
- Continue PA activities during interactions with
- Provide guidance to the DOE on site
~
Ongoing NMSS DOE daring the pre-licensing phase of repository characterization requirements, ongoing'sdesign work, and development, site characterization, and repository licensing issues important to the DOh design.
development of a complete and high-quality license application.
- Compare results ofNRC's iterative performance assessment to DOE's VA to identify major differences / issues.
5.2 APPLY PRA TO SPENT FUEL
- Demonstrate methods for PRA of spent fuel
- Prepare user needs letter to RES 4/97C RES/NMSS l
STORAGE FACILITIES storage facilities.
- Conduct PRA of dry cask storage 9/99
[
1 53 CONTINI;E DE OF RISK
- Use PRA methods, results, and insights to
- Update the database on transportation of radioactive End of FY NMSS ASSESSM' INT IN SUPPORT evaluate regulations governing the transportation of materials for future applications 99 i
OF RADIOnCTIVE MATERIAL radicactive material.
- Revalidate the results of NUREG-0170 for spent fuel TRANSPORTATION shipment risk estimates 6/99 f
A3 xiii t
f
c PRA Training to Support Risk Informed Regulatory Initiatives i
BACKGROUND l
In SECY-97-077, dated April 8,1997, the staff requested Commission approval to publish risk-informed regulatory guides and standard review plans for public comment. In a June 5,1997, i
Staff Requirements Memorandum (SRM), the Commission approved publication of these draft documents and directed the staff to " provide the Commission information on its plans for training NRC staff 1) on the risk-informed regulatory approach (es) contained in the regulatory guidance and standard review plan documents and 2) in overall PRA methods and techniques."
The Commission noted that "particular attention should be given to increasing basic user-level knowledge o' prob %ilistic risk assessment (PRA) methods at the regional level." The staffs response to this SRM is provided below.
DISCUSSION-Risk-informed regulation uses data and insights derived from probabilistic risk analyses to complement and support the traditional engineering analysis approach. To support risk- -
t informed licensing decisions, it is essential that the staff and inspectors be familiar with Commission policy and expectations as well as various aspects of PRA analysis methodologies and results. These aspects include, but are not limited to, strengths and limitations of PRA analysis, the scope of PRA analyses, the use of importance measures, and the effects and i
sources of uncertainty. Furthermore, the staff also must be familiar with the regulatory framework being established to support risk-informed applications from industry. W.th these objectives in mind, the stafi has designed specific minimum mandatory training programs for technical staff in the Office of Nuclear Reactor Regulation (NRR), inspectors in both NRR and the Regional Offices and NRR, and Regional technical managers that are discussed below.'
NRR Technical Staff All NRR technical staff will be required to attend a newly developed seminar on responsibilities
]
associated with risk-informed regulation. This seminar will orient the staff on the uses risk-informed regulatory initiatives ano will be led by a NRR senior manager. The seminar covers the PRA Policy Statement, the scope of risk-informed regulation, staff expectations, responsibilities and acceptance criteria.
'With much of the initial focus of risk-informed regulatory activities being on reactor j
applications, the initial training focus has been on NRR and regional staff. Training programs for managers and technical staff in other offices are still under development.
1 A4-i i
d i
All NRR technical staff will be required to complete the four day *PRA Basics for Regulatory l
Applications" (P-105) course or its equivalent.2 The Technical Training Division (TTD) of the Office of Analysis and Evaluation of Operational Data (AEOD) has updated this course to include information contained in the risk-informed RGs and SRPs. The target schedule for completion of this training is the end of fiscal year 1999. The staff is currently evaluating resource needs to meet this target schedule.
NRR and Reaion-Based Insoectors Regional and NRR Inspectors associated with the regulation of power reactors will be required to complete the "PRA Technology and Regulatory Perspectives" (P-111) course..This is a new l
basic user PRA course targeted to the specific needs of inspectors. The course curriculum includes extensive practical workshops and case studies applicable to the needs of the inspector. The first offering of this course is scheduled to begin in October 1997. Resident inspectors will be given the highest attendance priority with the goal of having at least one resident at every site complete the training by the end of 1998.
NRR and Reaional Technical Manaaers Regional and NRR Technical Managers associated with the regulation of power reactors will be required to complete the three day "PRA for Technical Managers" course (P-107). TTD has updated this course to include information contained in the risk-informed regulatory guides and standard review plans. Seven sessions of P-107 are scheduled for fiscal year 1998 and will be sufficient to train two thirds of the agency's technical managers. Additionally, sufficient courses will be available during fiscal year 1999 to permit remaining technical managers, associated l
with the regulation of power reactors, to complete P-107 by the end of fiscal year 1999. Senior management will establish attendance priority as required to support implementation of risk-informed regulatory activities.
Additional Aaencv-WideTechnicalTrainino The training plan described above will provide sufficient training to support implementation of the risk-informed RGs and SRPs; but due to resource limitations, it will not provide staff with all of the basic user level courses and prerequisites recommended in NUREG/BR-0228, "Guldance for Professional Development of NRC Staff in Regulatory Risk Analysis."
Consequently, if additional PRA training is needed to support specific risk-informed regulatory applications, NRC managers will be expected to define such training for their staff. For advanced users of PRA, the NRC's current PRA training curricula includes eleven advanced technology courses.
2NRR technical staff members who have completed basic user level PRA training within the last three years will be exempted frcm requirement to complete the P-105 course. To ensure that these staff members receive adequate training on the risk-informed documents, they will be L
required to receive training based on the newly developed risk-informed modules in addition to the risk-informed regulation seminar.
A4il
4 t
Executive Sum. mary NUREG-1560 (final version)
EXECUTIVE
SUMMARY
Introduction criteria used to define a vulnerability. The wording used in some submittals is such that it is not always On August 8,1985, the U.S. Nuclear Regulatory clear whether a licensee is identifying a finding as a Commisjon (NRC) issued its " Policy Statement on
" vulnerability" or as some other issue worthy of Severe Accidents Regarding Future Designs and attention. Therefore, a problem considered to be a Existing Plants"(Federal Recister.50FR32138). That vulnerability at one plant may not have been policy statement introduced the Commission's plan to specifically identified as a vulnerability at another address severe accident issues for existing commercial plant. In fact, less than half of the licensees actually nuc! car power plants.-
identified " vulnerabilities" in their IPE submittals; however, nearly all of the licensees identified other ne Commission formulated an approach for areas warranting investigation for potential systematic safety examination of plants to study improvements. Thus, the IPE program has served as particular accident vulnerabilities and desirable, cost-a catalyst for further improving the overall safety of effective changes to ensure that the plants do not pose nuclear power plants.
~
any undue risk to public health and safety. To implement this approach, the NRC issued Generic Letter (GL) 88-20 in November 1988, requesting that Only four licensees with boiling water reactor (BWR) lP ants and 15 licensect with pressurized water reactor all licensees perform an Individual Plant Examination (IPE) "to identify any plant-specipe vulnerabilities to (PWR) plants explicitiy stated that their plants had severe accidents and report the results to the vulnerabilities. The following vulnerabilities were Commission. " The purpose and scope of the IPE identified by the four BWR licensees with no effort includes examining internal events occurring at common vulnerabilities cited:
full power, including those initiated by intemal flooding. In response, the staff received 75 IPE failure of water supplies to isolation condensers covering regarding 108 nuclear power plant units.
The staff 'then examined the IPE submittals to failure to maintain high-pressure coolant injection e
determine what the collective IPE results imply about and reactor core isolation cooling when residual the safety of U.S. nuclear power plants and how the heat removal has failed IPE program has affected reac':.r safety.
The following sections summarize the rest!ts of the IPE failure to control low-pressure injection during an Insights Program examination.
anticipated transient without scram (ATWS)
Impact of the IPE Program on drywell steel shell melt-through as a Mark 1 Reactor Safety containment issue ne primary goal of the IPE Program was for Some of the PWR licensees identified common licensees to "identifyplant-specife vulnerabilities to vulnerabilities. Some of the vulnerabilities in the severe accidents that could be fixed with low-cost submittals include:
improvements."
However, GL 88-20 did not specifically define what constitutes a vulnerability; loss of reactor coolant pump (RCP) seals leading hence, the IPEs exhibit considerable diversity in the to a loss of coolant accident (LOCA) i NUREG-1560
f-j l
Executive Summary design and maintenance problems that reduce improvements and to support decisions regarding their
)
1 turbine-driven auxiliary feedwater pump implementation. The specific improvements vary from plant to plant.
However, m m nous reliability improvements that had significant impact on plant internal flooding caused by component failures -
safety include changes to AC and DC power, coolant injection systems, decay heat removal systems, failure of the operator to switchover from the heating, ventilating and air conditioning, and PWR coolant injection phase to the recirculation phase RCP seals.
\\
\\
Ioss of critical switchgear ventilation equipment IPE Results Perspectives leading to loss of emergency buses (Core Damage Frequency) need to enhance operator guidance for in many ways, the IPE results are consistent with the depressurization during steam generator tube
.esults of previous NRC and industry risk studies.
ruptures The IPE results indicate that the plant core damage frequency (CDF) is often determined by many inadequate surveillance of specific valves leading different sequences (in combination), rather than e
to interfacing system LOCAs (ISLOCA) being dominated by a single sequence or failure mechanism. The largest contributors to plant CDF loss of specific electricalbuses and the dominant failures contributing to those a
sequences vary considerably among the plants (e.g.,
compressed air system failures some are dominated by LOCAs, while others are e
dominated by station blackout). However, for most inability to cross-tie buses during loss of power e
plants, support systems are important to the results because support system failures can result in failures conditions of multiple front-line systems. Further, the support In addition, almost all of the licensees identified plant improvements to address perceived weaknesses in system designs and dependency of front-line systems on support systems vary considerably among the plant design or operation. (Over 500 proposed improvements were identified by the plants.) Most of plants. That variation explains much of the variability observed in the IPE results.
these plant improvements are classified as procedural /
operational changes (approximately 45 percent),
design / hardware changes (approximately 44 percent),
Consistent with previous risk studies, the CDFs or both.
Few of the improvements involve reported in the iPE subminals are lower, c average, maintenance-related changes.
Typically, the f ar BWR plants than for PWR plant:, as shown in procedural or design changes indicate revised training Figure E.1. Although both BWR and PWR results are strongly affected by the support system in order to properly implement the actual cnange.
considerations discussed above, a few key differences j
Approximately 45 percent of the plant improvements are identified by the licensees as implemented, with between the two types of plants contribute to this approximately 25 percent implemented and credited tenGncy for lower BWR CDFs and cause a difference in the IPEs. Other improvements are either planned in the relative contributions of the accident sequences or under evaluation.
Some improvements are to plant CDF. The most significant difference is that associated with other requirements (primarily the BWRs have n' ore injection systems than PWRs and i
station blackout rule) and utility activities. However, can depressurize more easily to use low-pressure although these improvements were not necessarily injection (LP1) systems. This gives BWRs a lower identified as a result of the IPE, in some cases, the average mntribution from LOCAs. However, the licensee is using the IPE to prioritize the results for individual plants can vary from this general NUREG-1560 jj
Erecutive Summary variability in modeling assumptions (including trend. As shown in Figure E.1, the CDFs for many BWR plants are actually higher than the CDFs for whether the models accounted for alternative many PWR plants. The variation in the CDFs is accident mitigating systems) primarily driven by a combination of the following differences in data values (including human error factors:
probabilities) used in quantifying the models.
j plant design differences (primarily in support Table E.1 summarizes the key observations regarding systems such as cooling water, electrical power, the importance and variability of accident classes ventilation, and air systems) commonly modeled and discussed in the IPEs.
1E4
- . 3 e
> iE4
.ses.
A 3
- AtA.
l.!!!'.
o E
1A I**l.
I 1
t
' A'
- l
- 1E4 a
g
>t h
A A A A
e a J
E (llF' A
$ 1E4 A
- m cn 85 E
E m
$ 1E 7 i
u 1E4 BWRs PWRs Figure E.1 Summary of BWR and PWR CDFs as reported in the IPEs.
Executive Summary Table E.1 Overview of key CDF observations-Accident class.
Key observations important contributor for most plants because ot' reliance on support systems; failure of such Transients (otbar than station systems can defeat redundancy in front line systems blackouts and
~
Both plant-specific design differences and IPE modeling assumptions contribute to variability in ATWS) results-use of alternative systems for injection at BWRs variability in the probability that an operator will fail to depressurize the vessel for LPI in BWRs availability of an isolation condenser in older BWRs for sequences with loss of decay e
heat removal (DHR) susceptibility to harsh environment affecting the availability of coolant injection capability following loss of DHR capsbility m use feed-and-blesd cooling for PWRs susceptibility to RCP seal LOCAs for PWRs ability to depressurire the reactor coolant system in PWRs affecting the ability to use LPI ability to cross-tie systems to provide additional redundancy Station blackouts Significant contributor for most plants, with variability driven by:
number of redundant and diverse emergency AC power sources availability of alternative offsite power sources j
length of battery life availability of firewater as a diverse injection system for BWRs susceptibility to RCP seal LOCAs for PWRs ATWS Normally a low contributor to plant CDF because of reliable scram function and successful operator responses BWR variability mostly driven by modeling of human errors and availability of altemative boron injection system PWR variability mostly driven by plant operating characteristics, IPE modeling assumptions, and assessment of the fraction of time the plant has an unfavorable moderator temperature coefficient Small contributor for most plants because of the separation of systems and compartmentalization Intemal floods in the reactor building, but significant for some because of plant specific designs Largest contributors involve service water breaks LOCAs Signi4
,t contributors for many PWRs with manual switchover to emergency core cooling (other than recirculation mode ISLOCAs and steam generator BWRs generally have lower LOCA CDFs than PWRs for the following reasons:
BWRs have more injection systems tube ruptures BWRs can more readily depressurize to use low-pressure systems (SGTRs))
ISLOCAs Small contributor to plant CDF for BWRs and PWRs because of the low frequency of initiator Higher relative contribution to early release frequency for PWRs than BWRs because oflow early failure frequency from other causes for PWRs OTR Normally a small
.eritator to CDF for PWRs because of opportunities for the operator to isolate a break ano terminate an accident, but important contributor to early release frequency NUREG-1560 iv
Executive Summary IPE Results Perspectives of the containment, containment bypass, containment isolation failures and, for some BWR plants, (Containment Performance) deliberate venting of the containment.
For the most part, when the accident progression As a group, the large dry PWR containments analyzed analyses in the IPEs are viewed globally, they are in the IPEs have sigmficantly smaller conditional j
consistent with typical containment performance Probabilities of early structural failure (given core analyses. Failure mechanisms identified in the past melt) than the BWR pressure suppression as being important are also shown to be important in the IPEs. In general, the IPEs confirmed that the containments analyzed.
On the other hand, large volume PWR containments are more robust than c ntainmentbypassandisolation failuresaregenerally the smaller BWR pressure suppression containments m re significant for the PWR containments. As seen in meeting the challenges of severe accidents; but the in Figure E.2, however, these general trends are often IPEs also showed containment bypass was more likely not true for individual IPEs because of the with PWR systems.
considerable range in the re-!ts.
For instance, conditional probabilities for bcth early and late Because of the ri.<'c importance of early releases, the containment failure for a number of large dry PWR containment performance analysis descriptions found containments were higher than those reported for in the IPE submittals emphasized the phenoment, some of the BWR pressure suppression containments.
mechanisms, and accident scenarios that can lead to Table E.2 summarizes key observations regarding such releases. These involve early structural failure containment perfonnance.
1.0 g 0.9 :
3 J 0.s :
2
I 5
t rr
.A f, 0.6 :
AA c
a l 0.5 i
'A 64 F
g CA :
AA 3,
c sa 48 afg 0.3 -
h 4'
ase 10.1-'
s,4, a
a aa o
g, a.
9.1 -
g.
d b'
0.0 Bypass Earfy laikre Late failure Bypass Early failure
- 1. ate failure PWRs g)ygg Figure E.2 Summary of conditional containment failure probabilities for BWRs and PWRs as reported in the IPEs.
~
Executive Summary Table E.2 Key observations regarding containment performance.
?
Failure Key observations mode Early failure On average, the large volume containments of PWRs are less likely to have early structural failures than the smaller BWR pressure suppression containments Overpressure failures (primarily from ATWS), fuel coolant interaction, and direct impingement of core debris on the containment boundary are important contributors to early failure for BWR containments The higher early structural failures of BWR Mark I containments versus the later BWR containments are driven to a large extent by dsywell shell melt-through*
In a few BWR analyses, early venting contributes to early releases Phenomena associated with high-pressure melt ejection are the leading causes of early failure for PWR containments' isolation failures are significant in a number of large, dry and subatmospheric containments The low early failure probabilities for ice condensers relative to the other PWRs appear to be driven by analysis assumptions rather than plant features For both BWR and PWR plants, specific design features lead to a number of unique and significant containment failure modes Bypass Probability of bypass is generally higher in PWRs, in part, because of the contribution from SGTRs Bypass, especially SGTR, is an important contributor to early release for PWR containment types Bypass is generally not important for BWRs Late failure Overpressurization when containment heat removal is lost is the primary cause of late failure in most PWR and some BWR containments
)
High pressure and temperature loads caused by core-concrete interactions are important for late failure in BWR containments j
i Containmen* venting i mportant for avoiding late uncontrolled failure in some Mark I containmeats The larger volumes of the Mark 111 containments (relative to Mark I and Mark II containments) are partly responsible for their lower late failure probabilities in comparison to the other BWR containments The likelihood of late failure often depends on the mission times assumed in the analysis
- There has been a significant change in the state-of knowledge regarding some severe accident phenomena in the time since the IPE analyses were carried out.
The results for BWRs, grouped by ct..t.nment type, severe accident than the later Mark 11 and Mark Ill follow expected trends and indicate that, in general, designs. However, the ranges of predicted failure Mark I containments are more likely to fail during a probabilities are quite large for all BWR containment NUREG-1560 vi
Executive Summary designs and there is significant overlap of the results, radioactivity decay and natural deposition, as well as given core damage. A large variability also exists in for accident response actions (such as evacuation of the contributions of the different failure modes for tb population in the vicinity of the plant). What is each BWR containment group.
However, a considered to be a significant release varies among the significant probability of early or late structural licensees. For many, significant release includes tailure, given core damage, was found for plants in all instances involving a release fraction of volatile three BWR containment groups. The containment radionuclides equal to or greater than ten percent of performance results for PWRs indicate that most of core inventory. Using this definition, the reported the containments have relatively low conditional conditional probability for significant early release probabilities of early failure, although a large varies from less than 0.01 to 0.5 for the BWR IPEs variability exists in the contributions of the different and from less than 0.01 to 0.3 for the PWR IPEs.
failure modes for both large dry and ice condenser Frequencies of significant early release are shown in containments.
Figure E.3 and vary from negligible to 2E-5 per reactor year (ry) for BWRs and from IE-8hy to The results presented i 'he IPE submittals are also 2E 5/ry for PWRs. In the BWR IPEs, significant consistent with previous stadies regarding early releases are almost exclusively caused by early radionuclide release. A significant early release is or containment failure, while containment bypass particular concern because of the potential for severe (especially SGTR), plays an important role in the consequencesas a result of the short time allowed for reported PWR releases.
IE-4 3
3 A
A 1E-5,
4
_a_g_
A A
A m
A E
A 1E-6j A
3 eA l
A MA m
b A
- , IE A a
AM&
g A
8 M
IE-8.
sr 7
i 1E IE a IE-11 Mark I Mark 11 Mark III Large Dry plus Ice Subatmos.
Condensers Figure E.3 Frequencies of significant early release (by containment type) as reported in the IPEs.
vii NUREG 1560 n.
Executive Summary IPE Results Perspectives variability in the results from the BWRs is caused by valid rather than inappropriate factors,the HEPs from man Pe&mance) several of the more important human actions 8PPearing in the IPEs were examined across plants.
Only a few specific human actions are consistently important for either BWRs or PWRs as reported in The results from the examination indicated that much the IPEs. For BWRs, the actions include manual eo va a ry in t H s and in k depressurization of the vessel, initiation of standby results f the HRAs across IPEs was due to liquid control during an ATWS, containment venting, 8PPropriate factors. Thus, the staff concluded that in and alignment of containment or suppression pool 8'"8 the licensees attempted to consider relevant cooling. Manual depressurization of the vessel is fact rs m obtaining the HEPs for operatoractions and more important than expected, because most plant that the results of the HRAs performed by the operators are directed by the emergency operating different licensees were generally consistent, and procedures to inhibit the automatic depressurization therefore, useful. On the other hand, it was also system (ADS) and, when ADS is inhibited, the con luded that not all of the variability in the operator must manually depressuiae the vessel, examined HEPs was easily explained. That is, after "accePtaW reasons for vadadon wem codered, Only three human actions are important in more than there still appearedto be some degree of" random" or 50 percent of the IPE submittals for PWRs. These uneXP ained variability, Potential reasons for this l
include the switchover to recirculation during LOCAs, variability include the lack of precision in existing initiation of feed-and-bleed,and the actions associated HRA methods and the fact that some licensees did not with depressurization and cooldown. Plant specific Perform as thorough HRAs as possible.
features, such as the size of the refueling water storage tank and the degree of automation of the switchover to recirculation, are key in determining the While the degree of consistency in HEPs obtained for importance of these actions.
. similar human actions in similar contexts suggests that in paneral the HRA results from the IPEs were useful While the IPE results indicate that human error can be in terms of meeting the intent of Generic a significant contributor to.QDF, nuinerous factors Letter 88-20, it should be further noted that even can influence the quantification. of human error when reasonable consistency exists, it is not necessary probabilities (HEPs) and introduce significant the case that all the HEPs calculated by a particular P ant were realistic and valid for that plant As noted l
variability in the resulting HEPs, even for essentially identical actions. General categories of such factors in Chapters 5 and 13, reasonable consistency can be include plant characteri"ics nodeling details, obtained in HRA without necessarily producing valid HEPs. An HEP is only valid to the extent that a sequence-specific attributes (e.g., patterns of successes and failares in a given sequence),
correct and thorough application of HRA principles has occurred.
dependencies, performance shaping factors modcled,
'1 application of the human reliability analysis (HRA) method (correctness and thoroughness), and the biases IPE Models and Methods of both the analysts performing the HRA and the plant personnel from whom selectedinformation and Perspectw, es judgments are obtained. Although most of these factors introduce appropriate variability in the results As a result of the IPE program, licensees elected to (i.e., the derived HEPs reflect "real" differences such perform probabilistic risk assessments (PRAs) for as time availability, dependencies and plant-specific their IPEs. Perspectives on the PRAs used in the factors), several have the potential to cause invalid IPEs provides information on where the models and variability. In order to examine the extent to which methods are sophisticated versus where potential NUREG-1560 viii
~.. -. - -. - - - -. -. -. _. - - -
l Executive Summary development may be needed. Developing these Additional IPE Perspectives perspectives requires consideration of several issues:
i 1
What are the characteristics of a PRA?
The Safety Goal Policy Statement established two qualitative safety goals, which are supported by two e
How do the IPEs/PRAs compare to these quantitative health objectives (QHOs) regarding the characteristics?
risk to the population in the vicinity of a nuclear j
power plant. Specifically, the safety goals establish What can be said about the IPE analyses given that the risk of both prompt fatalities and latent j
cancer fatalities that might result from reactor the scope of the IPEs (intent of Generic Letter 88 20) and the scope of the staff's IPE reviews?
accidents should not exceed 0.1 percent of the corresponding risks resulting from all other types of accidents or causes. In responding to GL 88-20, First, the characteristics of each analytical task of a PRA were defined. The IPE models and methods as licensees only considered internal events at full power described in the individual submittals are compared and were not requested to calculate offsite health against the characteristics a PRA. Some general effects. Therefore, the IPE results cannot be directly conclusions are then drawn based on the limited compared against the above health objectives.
reviews that have taken place.
However, it is possible to link the CDF and containment performance results to the safety goals by The CDF analyses of the IPEs generally compare well using surrogate indicators (such as the frequencies of j
to the characteristics defined for a PRA. The IPEs early containment failure and bypass). On the basis
]
are generally robust with respect to the identification of the frequencies of early containment failure and of dominant accident sequences. This is not to say bypass reported in the IPEs, a fraction of the plants that particular accident sequence frequencies have have the potential for early fatality risk levels that j
been verified, but rather that most of the imponant c uld approach the QHOs. This subset of plants was accident types have been captured. Although the funher examined using the frequencies of source CDF' analyses in the IPEs are generally robust, terms (from the IPEs) with relatively high release weaknesses were identified in a number of submittals factors (>0.03 I, Cs, Te) and adjusted for population.
in the areas of plant-specific data, common cause On the basis of this further screening, two BWRs and failure data, and human reliability. An important 12 PWRs remain with the potential for early fatality concem for some IPEs is the HRA, with the most risk level that could approach the QHOs.
. significant concem being the use of invalid HRA assumptions not producing consistently reasonable Many of the BWR and PWR plant improvements results.
address station blackout (SBG) concerns and originated as a result of the SBO rule. These The IPEs exhibit greater variation in the methods and improvements had a significant impact in reducing the scope of the accident progressiori ' ontainment SBO CDF (an average reduction of approximately performance) analyses than that found in the CDF 2E-5/ry, as estimated from the CDFs reported by analyses. This is commensurate with the guidance of licensees in the IPEs).
With the SBO rule GL 88-20 and NUREG 1335 which allowed implemented,the average SBO CDF is approximately significant diversity in the ways licensees could 9E-6/ry, ranging from negligible to approximately 3 E-conduct their containment performance analyses. In 5/ry. Although the majority of the plants that many of the IPEs, the containment performance implemented the SBO rule have achieved the goal of analyses and the source term calculations were more limiting the average SBO contribution to core damage simplified than the characteristics identified for a to about IE-5/ry, a few plants are sliscr/ above the PRA.
- goal, ix NUREG-1560
-e
,.w-
Executive Summary i
in NUREG-1150, the NRC assessed risk for five accidents. Further, the IPE program has served f
l nuclear power plants representing both PWR and as a catalyst for further improving the overall BWR designs. While these five plants represent only safety nuclear power plants, and therefore, the l
a small samole of designs, it is possible to consider generic letter initiative has clearly been a success.
whetherthe NUREG-1150 results and perspectives are Areas and issues have been identified where the consistent with those found in the IPEs. The average CDFs estimated for both BWRs and PWRs in staff plans to pursue some type of follow-up NUREG-il50 fall within the ranges of the CDFs activity. Areas under consideration are plant estimated in the IPEs. The relative contributions of improvements, containment performance accident sequences in the IPE results is also consistent improvement items either not implemented or not with the NUREG-ll50 results. For PWRs, SBO, addressed in the IPE submittal, and plants with transients, and LOCAs are usually the more imponant relatively high CDF or conditional containment contributors; for BWRs, LOCAs and ATWSs are failure probability (greater than IE-4/ry and 0.1, generally less important than SBO and transients.
respectively).
Unresolved safety issue (USI) A-45 (" Shutdown The conditional probabilities of early containment failure reported in NUREG-1150 (mean values) also Decay Heat Removal Requirements") and cenain fall within the range of the IPE results for each other USIs and generic safety issues (GSis),
containment type. The IPE results indicate that primarily GSI-23 (" Reactor Coolant Pump Seal conditional probabilities for early containment failure Failures"), GSI 105 (" Interfacing System LOCA are generally higher for BWR containments than for in Light Water Reactors") and GSI-130 PWR plants. On the basis of absolute frequency,
(" Essential Service Water System Failures at early containment failures for BWRs are similar to Multi-Unit Sites"), were proposed by licensees those of PWRs because the higher conditional early for resolution on a plant-specific basis. Other containment failure probabilities for BWR safety issues resulting from the IPEs were containments are compensated by the lower values for identified as candidates for further investigation.
Areas where further research regarding both severe accident behavior and analytical Overall Conclusions and techniques would be useful and should be Observations considered were identified.
information from the IPEs/PRAs has the potent;al In considering the paspectives discussed above, and the results reported in the IPE submittals, certain to support a diversity of activities such as plant conclusions and observations can ce drawn as inspection, accident managemer t sta.. ies, summarized below:
maintenance rule implementation, and risk-informed regulation.
As a result of the IPE program, licensees have IPE results can be used to indicate areas in PRA generally developed in-house capability with an increased understanding of PRA and severe where standardization would be useful.
4 4
Appendix C - Public Comments and NRC Responses on Draft NUREG-1560 l
a.m.,
a.
e.ua-.
M A
.ua4-.,a
,a a 4yiA._.J-
,a__4 ma
_w_ama-l.
~,e J_,u a
4-,4 a
i l
APPENDIX C l
PUBLIC COMMENTS AND NRC RESPONSES ON DRAFT'NUREG-1560 ta 0
i 1
j a
4 1
TABLE OF CONTENTS VOLUME 3 Chanlu Ease 1
LIST OF FIGURES...............
.............................................c-ii LIST OF TABLES...........................
... c il ABBREVIATIONS
.......... c iii C.1 Introduction..........
... C-1 C.2 Chapters 2 and 9: Impact of the IPE Program on Reactor Safety.....
..................C-4 C.3 Chapters 3 and 11: IPE Results Perspectives: Core Damage Frequency...
............. C-5 C.4 Chapters 4 and 12: IPE Results Perspectives: Containment Perfonnance...
C-11 C.5 Chapters 5 and 13: IPE Results Perspectives: Human Performance....
... C 13 C.6 Chapters 6 and 14: IPE Models and Methods Perspectives............
C-16 C.7 Section 7.1 and Chapter 15: Safety Goal implications...........
... C-17 C.8 Section 7.2 and Chapter 16: Impact of Station Blackout Rule on Core Damage Frequencies.... C-19 C.9 Section 7.3 and Chapter i7: Comparison with NUREG 1150 Perspectives................ C-20 C.10 Chapter 8: Overall Conclusions and Observations....
................ C-2 0 C.11 Chapter 10: Background for Obtaining IPE Perspectives........
.............. C-22 c-i NUREG-1560
Table cf Contents, Figures, Tables LIST OF FIGURES VOLUME 3 Pass F aure
.. C-9 Comparison of NUREG-1150 and Westinghouse seal LOCA models - old o ring elastomer..
C.I i
LIST OF TABLES VOLUME 3 Page IaMs
....... C-1 C.1 Submitted comments on draft NUREG-1560.............
............. C-3 C.2 Relationship of draft NUREG-1560 to the final NUREG 1560...........
NUREG-1560 c.;;
ABBREVIATIONS BWR Boiling W.ater Reactor CCFP Conditional Containment Failure Probability CDF Core Damage Frequency Cs Cesium ECCS Emergency Core Cooling System HEP Human Error Probability HRA Human Reliability Analysis I
lodine IPE Individual Plant Examination LERF Large Early Release Frequency LOCA Loss of Coolant Accident ~
MAAP Modular Accident Analysis Program NRC U.S. Nuclear Regulatory Commission NSSS Nuclear Steam Supply System PRA Probabilistic Risk Assessment PSF Performance Shaping Factor PWR Pressurized Water Reactor QHO Quantitative Health Objective RCP Reactor Coolant Pump SER Staff Evaluation Report SRV-Safety Relief Valve SBO -
Station Hlackout Te Tellurium VP Vice President c iii NUREG 1560
APPENDIX C PUBLIC COMMENTS AND NRC RESPONSES C.1 Introduction Commission (NRC) and their contractors, representativeof the owner's groups, vendors, utilities and their contrr.. tors, consultants, and Federal and NUREG 1560, Volumes 1 and 2 were initially issued State agencies. A report summarizing the workshop in October and November 1996, respectively as a was prepared and is available for inspection in the draft report for public comment with the comment NRC Public Document Room (Ref. C.3).
period ending May 9,1997. At that time, notices The report incluces presentation material distributed were published in the Federal Register announcing at the meeting and summarizes the discussion periods the availability of the report and requesting comment during which questions were raised and responses (Ref. C.1).
Distribution wa* made to over provided in addition, three sets of written comments i
500 people and organizations in the United States and were submitted at the meeting. These comments are abroad.
included in Appendix C or the Workshop Summary Report. The authors and organizations submitting To assist readers of the document, a 3-day public these comments are also listed in Table C.1 (items workshop was held in April 1997 on the contents of
- 23-26).
draft NUREG 1560. A notice of this workshop was published in the Federal Register (Ref. C.2)
In response to the request for comments, the NRC and notification of the workshop was sent to all staff received 23 letters.
T1.e authors and persons receiving the draft report. The workshop organizations submitting these letters are listed in took place in Austin, Texas and was attended by,,
. Table C.I.
All letters received are available for representatives of the U. S. Nuclear Regulatory...
inmeetion in the NRC Public Document Room.
Table C:1 Submitted comments on draft NUREG 1560.
Identification Author (s)
Date Organization received by NRC 1
Commonwealth Edison Company fhomas J. Maiman 2-14-97 Executive Vice President (VP) 2 Niagara Mohawk Martin McCormick, Jr.
2-14-97 VP Nuclear Engineering 3
South Carolina Electric and Gas Gary J. Taylor 2-17-97 Company VP Nuclear Operations Tony Spurgin 2 18-97 4
5 Centerior Energy Lew W. Meyers 2 25-97 VP 6
Duke Power Company M.1 Tuckman 3 3-97 Sr. VP Nuclear Generation 7
New York Power Authority James Knubec 3-4-97 Chief Nuclear Officer 4
I C-1 NUREG-1560 i
i m
l App C. Comments and Responses Table C.1 Submitted comtnents on draft NUREG-1560.
Identifiestion Author (s)
Date Organization received by NRC 8
Entergy Operations Inc.
Jerrold G. Dewcase 3-7-97 VP Operations Support 9_
Illinois Power Company Paul J. Telthorst 3-7-97 Director, Licensing P
10 Pacific Gas and Electric Company Gregory Rueger 3-10-97 St VP & General Manager C.A Kukielka, 3 12-97 11 Eric R. Jebsen 12 Carolina Power and Light Company William Orser 3 14-97 Ex.VP Energy Supply 13 PECO Nuclear G.A. Hunger 3-14-97 Director. Licensing 14 TU Electric C L. Terry 3 14-97 Gwup VP 3 14-97 15 BWR Owner's Group 16 Westinghouse Owner's Group Louis F. Liberatori, Jr.
3-25-97 Vice-Chairman 17 Northeast Utilities Services Company Suni! Weerokkody 4-10-97 Supervisor PRA 18 GPU Nuclear, Inc.
J.C. Fornicola 4-29-97 Director, Licensing & Regulatory Affairs 19 Baltimore Gas and Electric Ce.npany Charles H. Cruse 3-27-97 VP Nuclear Energy 5-8-97 20 Public Service Electric and Gas Company D.R. Powel; 5-9-97 Manager, Licensing & Regulation 21 IES Utilities,Inc.
John F. Franz 5-9-97 VP Nuclear 22 Nuclear Energy Institute Anthony Pietrangelo 5-9-97 Director, Licensing Nuclear Generation 23 Environmental Prot: an Agency T. Margulies 24 Virginia Power K. Tuley 25 New York State Department of Health J. Dunkleberger 26 NRC-IPE Workshop
- Written comments submitted at NRC-IPE workshop.
- Verbal comments discussed at NRC-IPE workshop.
NUREG-1560 C-2 i
l
App C. Comments and Responses In addition to these reviews and comments, as part of comments received on the draft, some of the chapters the normal review process, the staff discussed the were rearranged or renamed in the final report.
approach and results of dre.ft NUREG-1560 with the Table C.2 shows the relationship of the draft report to Advisory Committee on Reactor Safeguards on the final report on a chapter by chapter basis. The several occasions (Ref. C.4).
comments received were reviewed and categorized As discussed in Chapter I of this NUREG, the report acc rding to the various chapters. Comments related is comprised of two volumes, with Volume 1 as a to the " summary" chapter (from Volume 1) and the summary of the more detailed information contained associated detailed chapter (s)(from Volume 2) are in Volume 2.
However, due to the nature of the grouped together.
Table C.2 Relationship of draft NUREG-1560 to the final NUREG-1560 Volume I chapters Volume 2 corresponding detailed chapters Final report Draft report Final report draft report
- 1. Introduction same se no corresponding chapter no corresponding chapter
- 2. Impact of the IPE same so
- 9. Plant Vulnerabilities same Program on Reactor and Plant improvements safety no corresponding chap *r no corresponding chapter so
- 10. Background for
- 10. Background for Obtaining Obtaining IPE Results Reactor and Containment Perspectives Design Perspectivea -
- 3. IPE Results
- 3. Core Damage no
- 11. IPE Core Damage II. Reactor Design Perspectives Perspectives: Core Frcquency Frequency Persp:ctives Damage Frequency Perspectives
- 4. IPE Results
- 4. Containment m.
12.11 E Containment
- 12. Containmen' Design Perspectives:
Performance Performance Perspectives Containment Perspectives Perspectives Performance a IPE Results
- 5. Human Action so
- 13. IPE Human
- 13. Operational Perspectives Perspectives: Human Perspectives Performance Perforr innce Perspectives l
- 6. IPE Models and
- 6. IPEs with Respect to no
'
- Sctspectives on PRA
- 14. Attributes of a Quality PRA Methods Perspectives Risk-Informed Models and Methods Regu: an Used in the IPEs
- 15. Comparison ofIPEs to a Ouahty PRA
- 7. AdditionalIPE same so
- 15. Safety Goal
- 16. Safety Goal Implications Perspectives implications
- 16. Impact of Station
- 17. Impact of Station Blackout Blackout Rule on Core Rule on Core Damage Damage Frequencies Frequencies
- 17. Comparison with
- 18. Cor, arison with NUREG-NUREG-il50 1150 Perspectives Perspectives
- 8. Overall Conclusions same se no corresponding chapter no corresponding chapter and Observations C-3 NUA6G-1560
~
J App C. Comments and Responses All of the written comments sent directly to the NRC C.2 Chapters 2 and 9: Impact of (Items 1-22 in Table C.1) and submitted at the e
Fam on Reach workshop (Items 23-25 in Table C.1) together with all of the verbal comments provided at the workshop Safety (Item 26 in Table C.1) have been addressed in the final version of NUREG-1560. The comments fell in addition to comments identifying factual errors in i
into three broad categories:
these chapters which were corrected, the following general comments were received. These :omments and the NRC response are provided below.
. (1) A number of comments either were editorial in nature or address the accuracy of the information provided in draft NUREG-1560. For these 1.
Comment:
Numerous erroneous claims of general applicability of vulnerabilities are made comments, corrections were made to the text in the eport. Implying generic applicability of
]
where appropriate. These comments are not vulnerabilities is inconsistent with the Individual j
reproduced in this appendix with staff response.
The comments are available in the NRC Public Plant Exarnination (IPE), ntpose which is to
)
identify plant-specific vulnerabilities and cost-Document Room.
cffective improvements. (
Reference:
see Table
.1,#,15,20,22)
(2) Some comments wer observations in nature and did not appear to solicit a response nor seek a
Response
revision to the text of the report.
These comments are also not reproduced in this It is true that the generic applicability of appendix with staff response. The comments are identified vulnerabilities cannot be ascertained.
available in the NRC Public Document Room.
In addition, there is no consistent definition of vulnerability used in the IPEs.
- Further, (3) Other comments address insights, interpretations variability in plant design and operation, as well and perspectives drawn in the draft NUREG-as different modeling assumptions, can make a 1560. In some cases, the commentors were vulnerability unique to a particular plant.
concerned that the conclusions were Therefore, statements regarding generic unsubstantiated. In other eases,commentorswere applicability of vulnerabilities have been concerned about policy implications. For these rephrased in the NUREG. The purpose of comments, summaries were developed that presenting the vulnerabilities and associated plant captured the concern and an NRC staff response improvements identified by the licensees is so to the comment is pro ided. These comments that all of the licensees nay benefit from and associated responses are provided in the considering these enhancements as mear.s of following sections. The specific comments are improving the safety at their plant in a cost-available in the NRC Public Document Room.
effective manner.
Some of the comments discussed in the following 2.
Comment: Claims that plant improvements sections are more general in nature and applied to identified by one licensee could be implemented insights, interpretations, etc. discussed in more than by other plants should not be made. Plant one chapter of the report. Comments of this nature improvements should not be implementedwithout can, therefore, appear in several sections of this a full assessment of induced competing risks and appendix, An attempt is made in each section to the expenditure of resources required that may far identify those comments that apply to other parts of outweigh any safety benefit gained. (
Reference:
the NUREG.
see Table C.1, #15)
NUREG 1560 C-4
App C. Comm:nts und Responses
Response
1.
Comment:
The reported core damage frequencies (CDFs) and dominant contributors do All statements about generic application of plant not reflect updated probabilistic risk assessment improvements have been rephrased in the (PRA) results. Many utilities have updated their NUREG.
As with the identification of PRAs one or more times in response to plant vulnerabilities, the purpose for discussing design and procedure changes. In addition, many identified plant improvements is so that all licensees have provided the NRC with revised licensees can benefit by considering their IPE submittals some with extensive modeling potential implementation at their plant to improve changes and changes in the risk contributors and plant safety. A pmdent evaluation by a licensee CDF. To correctly reflect insights from the IPEs of the benefit of plant improvements identified requires consideration of supplementary by other plants would involve both cost-benefit submittals as well. (
Reference:
see Table C.1, and competUng risk considerations.
- 1, 12, 15, 22) 3.
Comment: Listing improvementimplementation
Response
by the licensees as of the date of the iPE submittal is misleading because many plant Because many plant PRAs are being constantly changes have occurred since the initial IPE updated to reflect the current plant design and submittals. (Reference see Table C.I. #1,16) operation, it is not practical to constantly update NUREG-1560 to incorporate new insights.
Response
NUREG-1560 is, and will remain, a compilation of the calculated CDFs and insights obtained NUREG-1560 represents a snapshot in time as far from the original IPE submittals. However, as risk and identified vulnerabilities and plant information from updated IPE submittals is improvements (including their implementation).
provided in Appendix B.
It is recognized that many licensees have updated.a..
their IPEs and the current status of identified- -
'2 Comment:
In comparing the plants, the plant improvements may, be different than from categorization of boiling water reactors (BWRs) what was reported in. the original submittal.
solely by vintage, pressurized water reactors Updated plant improvement status reports are (PWRs) by nuclear steam supply system (NSSS) presented in Appendix B for those licensees who vendor, and Westinghouse PWRs by the number provided updated status information in response of loops is not appropriate and can lead to to the solicitation of womments on Draft misinterpretation of results. It would be valuable NUREG-1560.
to also look at the reeults based an a categorizationof architect /engineerand/or builder C.3 Chapters 3 and 11: IPE and also age of plant to see if vari:tions can be l
exP ained within each NSSS category. mirther Results Perspectives: Core subgrouping of plants according to similar design Darnage Frequency characteristics (e.g., emergency core cooling system, ECCS, designs) could be possible.
Many.:omments were received conceming the
(
Reference:
see Table C.1, #16) accuracyof the information provided in these chapters or the insights that were identified. Corrections were
Response
made to the text where appropriate. In addition, several general comments were provided on the Early in the IPE Insights Program, the plants content of this chapter. These comments and an were grouped by architect / engineer and the IPE associated response are provided below.
CDFs within and among these groups were C-5 NUREG-1560
App C. ' Comments and Responses compared. It was found that comparison of
Response
results on this basis was not productive because Whether plant-specific design / operational there is considerable ' design variability even i
differences or modeling assumptions are among plants designed by the same dominant factors in explaining the variability is architect / engineer.. A decision was made to not always obvious. However,it is believed that i
l perform the analysis using plant groups based gg upon the NSSS vendor to account for basic NSSS variability for censin accident types. In many design differences. The BWRs were funher sub-cases, a judgment is made in the NUREG on
~
categorized by vintage to accou for differences which is the dominant factor for an accident class in ECCS design. se Westinghouse plants were for a plant group. The NUREG identifies that a grouped according to the number of loops since significant amount of variability is due to support i
the ECCS and other general plant features for the system and other plant-specific design / operational differences. Many of these design / operational plants in each of these groups are generally the i
differences are highlighted in the report.
same (see Table 10.3). It is recognized that the However, it is also clear that modeling balance of plant including support systems for assumptions play an important part in the plants in each of the designated groups can be I s me cases, because of limited different and skew any omparison of the results documentat. ".
ion m the IPE submittals, it is not for a plant group. The NUREG identifies that clear if the modeling assumption really reflects a these plant-specific features impact the results design or operational difference. For example, and draws the appropriate conclusions on the many licensees did not credit an alternate coolant resulting insights. Finally, it is recognized that injection system because they did not perform an j
further subcategorization of plants according to a analysis of whether or not it would be successful.
selected parameter could be made. However, The neglect of the potential use of this system is variabilh 5 other parameters would likely.
a model assumption until it is shown that, impact th.a comperison. Because of this fact and because of plant-specific factors, such a system could not be used. For other cases,it is clear also due to resource limitations, funher that a model assumption is being made. For subcategorization was not pursued.
example, many licensees assumed that the DC bus load shedding would always successfully 3.
Comment: ne degree a.vhich a search for occur during a station blackout.
variability associated with plant design differences has been made is onastionable. The 4.
Comment: The choice of success criteria has a NUREG states that important de:,ign features, major impact on the variability of the CDF operator actions, and model assumptions all results in a given category of plants. His is not impact the variability in results. However, few mentioned in the NUREG.
Som. utilities model assumptions are identified. As is well working with smaller PRA vendors had more stringent success (i.e., conservative) criteria than known, substantial differences in PRA results others who worked with reactor vendors and had 3
occur because of balance-of-plant and support access to indonnation that aHowed for less system design differences despite similarities in
- "'"*'^
'* ** '*'I*'
NSSS design. Therefore, it is judged that therc utilities had the resources to perform the is no basis to assert that the basis for observed necessaryanalyses to establish a less cor.ervative variability is anything but dominated by plant success criteria where other utilities did not have l
differences in design, procedures, and training.
such resources and chose to use a conservative
(
Reference:
see Table C.1, #15) success criteria. (
Reference:
see Table C.1, #16)
NUREG-1560 C-6 i
App C. Comments and Responses
Response
Response:
The NUREG identifies where success criteria Because of the variability in the IPE modeling, it
~
assumptions impact the. variability of the is not possible to always ascertain the impact of calculated CDFs. As mentioned 'in the response component failure rates and common cause to the previous comment, because of limited failure rates.
However, these factors were documentation in the submittals, it' was not considered in establishing the parameters j
always ' clear.if differences in success criteria affecting the variability in the reponed CDFs.
2 were due to design differences or modeling Selected comparisons were made and, - as discussed in Chapter 11, these failure rates were assumptions. The basis for not crediting a found to be important to the CDF variability, system (and m. some cases, for crediting a Also, based on a limited survey of data, system) or for the operating requirements of a Chapter 14 indicates that a wide variety of failure credited system : (including support system rates were identified in the IPEs for some requirements)were not always documented in the components. This variability applies not only to submittals. The CDF evaluation thus made no plant-specific data but also to generic failure rates attempt to validate the. differences in success identified in the submittals.
criteria but simply reported its impact on the variability on the results. Also, Chapters 10 and 7.
Comment: Care must be taken when comparing.
14 in the NUREG dise sses the importance of the CDFs from transient events and from loss of success criteria to the results in general terms.
coolant accidents (LOCAs). The IPEs approach the modeling of consequential LOCAs (e.g.,
1
- 5.
Comment: The NUREG should address the reactor coolant pump, RCP, seal LOCAs or j
criteria used to determine what constitutes core stuck-open power-operated relief valves or safety damage. Many IPEs use core uncovery while relief valves, SRVs) differently. Sometimes the others use a peak cladding temperature of CDFs from-these events are reported in the 2200*F. This is important in that it impacts what transient contribution and sometimes in the small equipment can be used to avoid core damage.
or medium LOCA CDF. It needs to be clearly
(
Reference:
see Table C.1, #11,15) stated how this is handled in NUREG-1560.
(
Reference:
see Table C.1, #16)
Response
4
Response
J
'Ihe impact of the definition of core damage on j
It is true that there was considerable van. bility a
success critena is d.iscussed in encral terms in among the IPEs with regard to 1, aping Chapters 10 and 14. Specific impacts on the sequences (for reporting) involving consequential
. variability of the reported CDF definitions were LOCAs. However, the majority of the submittals
- not addressed because insufficient information reported sequences initiated by either a rupture or was provided in the IPE submittals.
an inadvertent open SRV as LOCAs, and sequences with consequential LOCAS occurring 6.
Comment: A discussion on how the component after some other initiatar as transients. This i
failure rates and the common cause failure rates format was chosen 4r categorizingand reporting impact the results is missing from the NUREG.
the esults. For those IPEs that did not provide I
This could be particularly important for assessing
.ne results according to this format, an attempt I
the importance of station blackout (SBO) since was made to regroup the results to allow for the reliability of on-site emergency AC power is comparison with the CDFs for other plants.
i
[
critical.' (
Reference:
see Table C.1, #16)
However, in some cases, insufficient information C-7 NUREG-1560 4
4 r -
-m.,
-e.,
~
w
App C. Comm:nts and Responses was provided in the IPE submittal to distinguish
- 10. Comment: A basis for the key perspective that the CDFs associated with these different accident PWRs with better feed-and-bleed capability sequences. In those instances, the licensee's generally have lower CDFs should be provided.
reported results for transients and LOCAs were There are mcny other plant design features and modeling methods that have a greater impact on used directly, CDF. (
Reference:
see Table C.), #16) 8.
Comment: It is not clear where special initiators fit into the CDF information reported in the
Response
NUREG. Generally loss of component cooling The observation is made in the context of all water and loss of service water can be imponant contributors to the CDF for PWRs due to the PWRs. Within the Westinghouse plant groups, other factors besides feed-and-bleed capabilities potential for an RCP seal LOCA. it would be advantageous to report the transient results in are more important for explaining differencesin terms of CDF due to loss of decac heat removal transient CDFs and are discussed in the report.
and the CDF due to consequential LOCAs.
However, differences feed-and-bleed
(
Reference:
see Table C.1, #16) capabilities are imponant when comparing across all PWRs because of the Babcock & Wilcox and Combustion Engineering plant design differences,
Response
it is agreed that it would be useful to separate the
- 11. Comment: It is not clear from the information contributions from loss of decay heat removal presented that the Westinghouse RCP seal LOCA and consequential LOCAs for the transient model provides a lower contribution to CDF than sequences. However, this information is not the IPEs that used the NUREG-1150 model.
available coWstently from the IPE submittals.
Estimates were made from the reported Since this is very important to many plants, it is information, whenever possible, and used in the recommended that NUREG-1560 provide a repon to identify relevantinsights. The NUREG detailed comparison of the two approaches. One of the dominating factors in the seal LOCA identifies that consequential LOCAs are important contributors to the CDF for many model is the probability of core uncovery occurring within the first hour. IPEs using the BWRs and PWRs.
Westinghouse RCf1 seal LOCA model typically 9.
Comment: The discussion on LOCAs should be use 0.0283 andthe NUREG 1150 modeluses 0.0.
directed at the ability of plants to mitigate small The NUREG il50 model does not consider any LOCAs.
Overall, large LOCAs are not sealleakage for the first 90 m nut s. From these significant contributors to C"F. (
Reference:
see facts it appears that the Westinghouse RCP s..!
LOCA modelis more conservative. (
Reference:
Table C.1, #16) see Table C.I, #16)
Response
Response:
Significant contributions were observed from different sizes of LOCAs in different submittals.
A comparison of the seal LOCA probabilities nerefore,it is not always true that large LOCAs from the two models was not possible due to the are not significant contributors and that the unavailability of the repons documenting the discussion should focus on only small LOCAs.
Westinghouse model (with and without seal The NUREG discussion identifies what sizes of binding and popping open included). However, LOCAs dominate the LOCA contributions in the Point Beach IPE and the response to questions concerning the Farley IPE did provide each plant group and the reasons why.
NUREG-1560 C-8
App C. Comments and Responses an opportunity to campare the core uncovery while Surry and Farley are three loop plants and thus probability as a function of time for cases the core uncovery time for a given leak rate could be involving RCPs equipped with the old o-ring different. However, since the reactor coolant system clastomer with the vessel either depressurized or volumes for the plant, are roughly scaled by the not depressurized,and with the RCPs tripped. A number of coolant loops, the core uncovery times for comparison of the values from these curve fits three plants for the same amount of leakage from with the core uncovery times calculated for each pump should not be substantially different.
identical cases for the Surry plant, as reported.m NUREG-l l50, Volume 3, is provided in Thus, the core uncovery probability comparison in Figure C.I. The curve fit is, i valid over the Figures C.1 provides a reasonable picture of the time frame of 30 minutes to E aours. It should differences between the NUREG-il50 and be noted that Point Beach is a two loop plant Westinghouse seal LOCA models.
5 f U:
I l'
L fM' u,
}
c 4
5 EIS1150 M Z M-U gf g
PUE1150M i u; II U-J
[
Wat$m 506-Wen $m
/
ju:
M (d bohn M(dbalmii!
adpppr4 J 0.5 rulpgeg) l h kei b
l g04-l lUI f
h U-
[i Wenda 1u.
/
01-
/
g.
Wemghe
=
o iM 01-l
..... - ~ ~.... '
n 1 15 15 H 14 16 4.1 6 5.0 0 M 1 13 15 U 14 16 4.1 6 E0 0 H Figure C.1 Comparison of NUREG-1150 and Westinghouse seal LOCA models - old o-ring elastomer.
Figure C.1 indicates that all three models predict smaller probabilities for core uncovery for time small probabilities of leaks and core uncovery periods greater than approximauly 3 to 3.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />, for early times (less than about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />).
particularly for cases where the vessel is Because of this, differences between the three depresst 5ed. For scenarios where the vessel is not models do not have a significant impact on CDF depressurized, however, the probabilities predicted by for this early time period. However, for later the Westinghouse models rise sharply at about 8 times, the differences are more significant. The hours, so that the three models give similar Westinghouse models generally predict much probabilities at that time.
C-9 NUREG 1560
App C. Comments and Rssponses The fact that seal LOCAs occur in all three 7.7E 7/yr). Accounting for this error would models does not mean that the impact on the slightly widen the difference between the CDF will be the same in both cases. As noted Westinghouse and NUREG-ll50 models.
earlier, none of the models result in a significant contribution to CDF in the first three hours
- 12. Comment: The NUREG discusses uncertainty However, unlike the Westinghouse models, the ass ciated with the Byron-Jackson N-9000 seals NUREG 1150 model can result in significant and " infers that the IPEs (for plants with these contributions to CDF based on core uncovery in Pumps) are suspect in their RCP seal LOCA the 3_ 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> time frame. For example,in this conclusions." Details concerning this technical time frame during a station blackout, the core is issue have been provided to the NRC in various likely being cooled by auxiliary feedwater, given that battery power is still available. Therefore, forms in the past. Please modify the NUREG to without a seal LOCA, core damage would not be reflect the technical information provided and expectedduring this time frame. For times past remove the inference that the IPEs are suspect in 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, all three models predict a high their RCP seal LOCA conclusions. (
Reference:
probability of a seal LOCA leading to core see Table C.1, #8)
However, for these longer times, uncovery.
battery depletion would have occurred at most
Response
Westinghouse plants, leading to loss of heat removal and boiloff. Therefore, if AC power NUREG 1560 reflects the information provided recovery does not occur, core damage will result in the IPE submittals which indicate that the whether or not a seal LOCA is present. In this contribution from RCP seal LOCAs is generally situation, the station blackout CDF is not affected by small seal LOCAs that would result in core small for plant with Byron-Jackson pumps. The uncovery at times greater than 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. The NUREG reiterates the statements made in the precise impact of the model differences is submittals that there is little or no potential for plant-specific, depending on battery depletion seal LOCAs in these pisnts if the RCPs are times and AC power recovery attematives.
tripped. The submittals cite the design of the Similar impacts occur for non-station blackout pumps some limited analyses, test, and actual scenarios (e.g., loss of component cooling water experience as the basis for this argument events) where the seal leakage rate impacts the pr viding some references. No judgement is time available for other recovery actions such as made in NUREG-1560 abou; the accuracy of the arranging alternate charging pump cooling.
RCP seal LOCA modeling for these plants based on the information in the submittal. However, The documented NUREG-ll50 seal LOCA model indicates no seal failure prior to 90 the potential for RCP seal LOCAs in these plar.ts minutes. However, after most of *e IPEs were is still being reviewed as pet of NRC's Generic completed, an error in the NUREG-IISO model Safety Issue 23. The information cited in the was identified which indicates that there should submittals as the basis for the RCP seal LOCA be some probability of seal failure immediately modeling is being examined as part of the after loss of seal cooling. Thus, the contribution of RCP seal LOCAs in the IPEs that utilized the NUREG-1150 model is likely underestimated.
- 13. Comment:
The reported CDFs have been An evaluation for the NUREG-1150 study for the rounded to one significant figure. The NUREG Sequoyah plant indicates that the seal LOCA should report the actual CDFs reported in the contribution was underestimated by 18 %
(corresponds to a core damage frequency of IPEs. (
Reference:
see Table C.1, #8)
NUREG-1560 C-10
App C. Comments and Responses
Response
performance in the same way that isolation or structural failure of the containment is.
A decision was made to report the CDFs to one Therefore, the NUREG separates the conditional significant figure (to provide consistency) and are probabilities of containment bypass end based on the actual values reported in the IPE containment " failure" when making comparisons.
submittals.
The imponance of containment failure frequency is acknowledged in Chapter 12 of the NUREG C.4 Chapters 4 and 12: IPE where comparisons of containment failure frequencies as well as release frequencies are also Results Perspectw.es:
presented.
The NUREG does not draw Containment Performance conclusions or make implications rega ding overall plant safety based on CCFPs.
1.
Comment:
Conditional containment failure Containment failure probabilities are used only to probability (CCFP) is not a good measure of compare the containment performance among safety performance. The use of conditional plants with the same type of containment and measures implies an independence between the among different containment types. For this systems which prevent core damage and the purpose the CCFP is the best suited parameter.
systems which prevent containment failure which is part of the design of the current generation of 2.
Comment: The report utilizes at least five light water reactors. Plants with relatively higher different figures of merit in characterizing CCFPs are not necessarily less safe than those containment performance, it is never clear which with relatively lower CCFPs. The measures figure is most appropriate or why. The figures which impact public safety are related to the include: total conditional containment failure frequency of releases from the containment.
probability, conditional probability of various
(
Reference:
see Table C.1, #11,12,22, 23) containment release types (bypus, early failure, late failure), frequency of bypass and early
Response
release, conditional probability of "significant early release," and frequency of releases with the One of the main objectives of the chapters in potential to causc early fatalities. (
Reference:
see NUREG-1560 related to containment Table C.1, #22) performance is to obtain perspectives on the performance of the various containment types
Response
independent frc.n other plant features. For this purpose, the CCFP is a useful parameter since it NUREG 1560 uses various parameter" elated to decouples containment failure from core damage containment performance in different ch.; rs of frequency. This was also retognized by the the report depending on the purpose of the majority of licensees since CCFPs are reported comparisons to be made and the perspectives to directly in most of the IPE submittals. Ideally, be obtained.
There is no single "most the comparison of containment performance appropriate" containment performance figure of among different IPEs would be accomplished by merit for the whole report, nor should there be.
comparing CCFPs for individual plant damage Those parameters which best served to illustrate i
ststes. However, such a comparison is not the points to be made for the issues at hand were possible since the dvinition of the plant damage chosen in different sections of the report. Total states was left to the individual analyst and thus conditional containment failure probability is not varies from IPE !o IPE. NUREG-1560 also used in the NUREG. For purposes of obtaining recognizes that the probability of containment perspectives on containment performance, bypass is not a measure of containment conditional probabilities of containment bypass, NUREG-1560 C-11 1
App C, Comments and Responses the CCFPs for early and late failure are used in Regarding the industry position papers, their Chapters 4 and 12. Conditional probabilities of application in an IPE to qualitatively dismiss a significant or large early release, defined as early number of accident progression phenomena, releases where releases of Cs, I and Te exceeded without any sensitivity considerations, or without 0.1 of core inventory, are also compared in these any understanding of the uncenainty associated chapters since this type of release was singled out with the different phenomena,is not in line with in many IPE submittals. Finally, frequencies of the intent of Generic Letter 88-20.
'this early reler.se from bypass and early containment approach was less helpful in fostering a licensee's failure were used in Chapters 7 and 16 since this appreciation and understanding of severe accident parameter was the one which allowed an indirect behavior than a proper application of MAAP.
comparison of the IPE results with the safety goals.
4.
Comment: Results are presented by reactorand contaimrent type and NSSS.
It would be 3.
Comment: While there have been some mis-valuable to also look at the architect /engineeror applications of MAAP, any implication that the builder to explain the v".ation in reponed MAAP code is inadequate is wrong. It is results. (
Reference:
see Tabic C.1, #16) misleading to state that MAAP does not have a comprehensive treatment of severe accident
Response
4 phenomena. A more problematic item involve.
the utilities which did not properly apply MAAP Early in the IPE Insights Program a decision was and/or relied on the industry position papers.
made to group the containment performance
(
Reference:
see Table C.1, #8,11,12, 22)
- esults under the five commoa containment classes used in the United States. Containment
Response
response to severe accidents has been found to correlate to these five contamment classes as MAAP as well as other system level codes do not illustrated in the NRC's Containment cover the range of postulated severe accident Performance improvement program.
In phe nena (e.g.,
steam i ',losions, direct discussing ( ntainment. eformanceperspective 1
con. 2nent heating, shell met t.sugn, NUREG 1560 identifies those architect / engineer
- *e wn detonation). This is what is meant by specific containment construction features which s statement that the MAAP code does not have play a significant role in the IPE analysis, as j
a comprehensive treatment of severe accident reponed in the lit submittals. These features i
phenomena.
The EPRI repon on MAAP include the containment material, layout of acknowledges "one should recognise that AfAAP reactor cavay, and location < f snmps and drain cannot and does not contwr 1etailed modelsfor lines.
all phenomena" As noted above, other system level codes share this limitation, and this is one 5.
Comment: It is judged that there is no basis in reason why the IPE guidance called for proper NUREG-1560 to assert that the observed sensitivity studies to be conducted as pan of the variability in the IPE results is anything but Level 2 analysis. In some cases MAAP was dominated by plant differences in design, applied by the IPE analysts in a way that did not procedure, and training. (
Reference:
see Table follow industry recommended guidelines. NRC C.1,#15) noted "..the adequacy of the MAAP 3.0B code for use in the IPEs.. " but also stated that
Response
" licensees... bear the burden of proof that they have applied the code properly, and that they In discussing containment performance sneet the intent of the IPE generic letter."
perspectives, NUREG-1560 identifies the plant NUREG-1560 C-12
App C. Comments and Responses specific differences described in the IPE
Response
submittals which lead to some of the variability in the reported results. However, it is clear that There exists detailed discussion in the appropriate modeling assumptions also play an important role sections ef Chapters 4 and 12 of NUREG-1560 in the observed variability in containment on:
performance. Assumptions regardingthe amount and composition of core material exiting from the.
how venting was grouped to the different
- ^*""*"**
reactor vessel, the coolability of this debris, and the pressure and temperature rise in the how "early" and " late" was defined in the containment due to core debris dispersal are comparison of failure modes and reletses.
examples of modeling assumptions which had a significant influence on the assessment of how multiple containment failure modcs containment performance. Other assumptions were treated as they were reported in the IPE include the likelihood ofin-vessel recovery of the submittal (i.e. whichever failure.aode was accident, including the likelihood of retaining the considered dominant in the submittal base core debris in the reactor vessel via external case results was the one used in NUREG.
cooling of the vessel.
1560).
6.
Comment: It would seem pmdent to avoid The above comment on the treatment of misinterpretations by providing the specific NRC dynamic failure modes is not clear, and no assumptions used in extrapolating IPE submitted further clarification was provided at the words to the constmetion of the comparisons w rkshop; conseguently, no changes were made to NUREG-1560.
among plant results in NUREG-1560. These assumptions would include:
. C.5 Chapters 5 and 13: IPE What the relationship of containment vent Results Perspectives: Human treatment is to the.CCFP, th'e early releases, Performance j
and other measures of risk; 1.
Comment: It is stated in the report that in most
- what the W on is between each IPE cases there is little evidence that the human resuit for, srly and late releases and their reliability analysis (HRA) quantification method definition d "eas.y".md " late";
Per se has a major impact on the resu'.ts. This seems to imply that "the impact of HRA on PSA can best be describedas indeterminate"or "that how the assignment of multiple containment failure modes affects the assignment of the the HRA seems to have little efect on tk ~*sults
- I
- **8
- I *#*
allocation of failure modes in comparisons HRAs identified as impoitant shortcomings of the (e.g., shell melt-through followm.g wetwell IPEs and why is the quality of the HRAs a failure); and concem. (
Reference:
see Table C.1, #8,11,12, defining the treatment of dynamic failure modes and their associated failure locations
Response
as it relates to inferences about failure locations and timing. (
Reference:
see Table The interpretation that "the impact of HRA on C.1,#15)
PSA can best be describedas indeterminate"or C-13 NUREG-1560
- -. ~ -
App C. ' Comments and Responses that "the HM seems to have little efect on the HMs. " (F eference: see Table C.1, #1, 2, 8,11, results ofthe PM"is not what was meant. How 12,22)
. and how well the HRA method is applied and the resulting human error probabilities (HEPs) clearly
Response
' have significant impacts on the results of the PRA. Thus, it is for this reason that concern is Confusion arose regarding t:ie implication or raised in the NUREG about the " quality" of the meaning of the significant variability in HEPs that was identified for selected human actions
- HRAs performed by the di& rent licensees. The.
statement that "in most cases there is little across plants, particularly in terms of the quality of the HRAs. Figures displaying the UEPs for evidence that the HM quantification methodper several events (e.g., manual depressurization se has a major impact on the results, " was meant during transients) were presented in the report
- to imply that the HRA results from the di& rent and discussions of the reasons for the variability IF'is did not in general appear to vary directly as
- r* Provided. Many of the comments received a function of the particular or " nominal" HRA from licensees on this topic attempted to defend method used,' e.g., the Technique for Human.
the variability on the basis of the numerous Error Rate Prediction versus the Success reasonable factors that would lead to the Likelihood index Methodology versus the Human variability. That is, the values across plants may
' Cognitive Reliability model. The variability in have been developed on entirely di& rent bases.
results appearedto be more a function of how or For example, different plants have different how well the HRA methods were applied or the systent charactedsdes and may han Mere {
impact of plant specific characteristics, as Procedures.
Initiator and sequence-specific opposed to which nominal HRA method was factors and depender.cies will also lead to used. Due the confus. ion caused by the statement
- '" H Moreover, some plants only and the fact that the direct impact of the nominal used " screen.mg values" m modeling some of the method per se is difficult to evaluate given the
- exemined events. On the basis of these and other many other relevant factors, the statement was fu ors, the commentors ' indicated that such deleted from the final NUREG. ' Additional var ability would be expected.
clarification regarding the quality _of the HRAs performed for the IPEs is provided below in the This conclusion is, at least in part, one point the response to Comment #2.
staff was trying to make and which was stated in the report. That is, there are " reasonable
. 2. ' Comment: In spite of the assertion in the report explanationsfor much ofthe observedvariability that, "it appears that there re reasonable in HEPs across plants. la other words, the explanationsfor much ofthe variability in HEPs rather striking degree of variability, in at least
'and in the results of the HMs across the nominally similar human actions, is based to dsgerent IPEs," it is also assertM that because some extent on valid di&rences. From this "mmry of the licenseesfailed to perform high-perspective it can be argued that the licensees quality HMs, it is possible that the licensees attempted to consider relevant factors in obtained HEP values that are not appropriatefor obtaining the HEPs for operator actions and that their plants." These statements appear to be the results of the HRAs performed by the inconsistent. Moreover, others sections of the different licensees were generally consistent and report indicate that not all of the variability in therefore useful. In fact, the staff does not in
.HEPs could be, explained. ' Please provide general disagree with this conclusion.
clarification on what appears to be inconsistent
- statementsand address the assertion that "many of However, another conclusion reached by the staff x
'the licensees failed to perform high-quality and documented in this report was that not a.ll of NUREG-1560 C-14 4
f r
vr v
m
-yy
,u
.,.m-m
--4
--w rwr
App C. Comments and Responses the variability in the examined HEPs was easily consistency exists, it is not necessarily the case explained. Tht is, after" acceptable" reasons for that all the HEPs calculated by a particular plant varia%. ere considered, there still appeared to were realistic and valid for that plant. As noted H. some degree of unexplained variation the in Chapters 5 and 13, reasonable consistency can HEPs (see Chapter 13). While some ot this be obtained in HRA without necessarily
~
variation would be expected due to the lack of producing valid HEPs. An HEP is only valid te preeision in existing HRA methods, it is also the extent that a correct and thorough application possible that some of the variation was due to of HRA principles has occurred. For example,if factors such as analyst biases, invalid HRA a licensee simply assumed (without adequate assumptions made by analysts performing the analysis) that their plant is " average"in terms of HRAs, or superficial HRA analyses that failed to many of the relevant PSFs for a given event, but adequately examine and model the potential for then does appropriately consider the time human error (e.g., through careful consideration available for the event in a given context, the of plant-specific performance shaping factors value obtained may be.<imilar to those obtained (P3Fs), consideration of dependencies, use of for other plants with.imilar time frames for the simulator exercises, etc). Due to the limited event. Yet, the roulting value may be optimistic information provided in many submittals on the or pessimistic relative to the value that would derivation of particular HEPs, it is difficult to have been otained if the licensee had conducted determine the extent to which inappropriate a detail < J examination of the relevant plant-factors actually influenced the derived HEPs.
specif5 factors. Thus, while the degree of However, examinations of the submittals during consstency obtained by the licensees is the project indicated that not all licensees encouraging regarding the ability to compare the performed quality HRAs.
That is, not all results of the IPEs, and while many licensees licensees applied the existing HRA methods as performed excellent HRAs, the fact that some well as they could have. For example, tney aid licensees did not perform as thorough HRAs as not always consider dependencies, appropriately possible given the state-of-the-art in HRA at the assess the impact of time availability, or carefully time, means that the results are not as good as consider plant specific PSFs. Some failed to they might have been. It does not mean that model pre-initiator actions and others did not individual licensees and the industry in general conduct simulator exercises or perform did not obtain important information from walkdowns and timing of operatc'r actions to be performing the IPEs.
conducted outside the control room, etc. The conclusion that not all licensees conducted high-3.
Comment: By questioning the quality of the quality HRAs is further documented in some of HRAs performed for the IPEs, NUREG-1560 the staff evaluation reports (SEks) that have been seems to imply that the license.s shci have issued on the submittals. Some submittals attempted to extend the state-of-the-anin HRA in indicated as having met the intent of Generic order to ob ain quality results. (
Reference:
see Letter 88-20 were found to have various Table C.1, #8,11, 21) weaknesses that could have influenced the HEPs obtained for particular events.
Response
While the degree of consistency in '
's The staff believes that the state-of-the-art in HRA obtained for similar human actions in sim.m at the time of the IPEs was adequate for the contexts suggests that in general the HRA results intent of Generic Letter 88-20. The shortcomings from the IPEs were useful in terms of meeting related to the HRAs performed for the IPEs were the intent of Generic Letter 88-20, it should be in how the existing methods were applied, rather further noted that even when reasonable than the methods themselves. Of course, this C-15 NUREG-1560
i 1
App C. Comments and Responses position does not imply that improvements are C.6 Chapters 6 and 14: IPE not needed in HRA, but rather that useful results Models and Methods j
can be obtained with thoughtful and thorough Perspect.ives applications of existing methods.
Several comments were received expressing technical 4.
Comment: The NRC needs to initiate a number disagreement with some of the information provided of policy and research activities to address in these chapters. The text was revised where shortcomings both in the NRC's attitudes and appropriate in addition, several general comments strategies for ensuring that the licensees maintain were provided on the content of these chapters.
j These comment and associated responses are provided safe plants and in the development and use of J
below.
PRA and HRA methods and techniques. These activities (summarized) include establishing a 1.
Comment: Numerous comments were received regulatory attitude that encourages the licensees on the description of a
- quality" PRA.m to be pro-active i.aner than reactive (to the NRC)
Chapters 6 and 14 and on the comparison of the in ensuring plant safety, encouraging more IPEs to a quality PRA in Chapters 6 and 15 of thorough and realistic HRAs, supponing the the draft NUREG. Several commentors felt that development of multiple new approachesto HRA these chapters were inappropriate for NUREG-1560 and that they should be deleted from the (which include morc effective use of simulators),
fimal report. This recommendation was largely reevaluation of the real contribution of common driven by the assumption that the attributes of a cause to risk, reevaluating the use of Bayesian
" quality" PRA were intended to be standards or updating during " period of rapid changes in requirements and that all the attributes had to be maintenance," and.mvestigating the impact of met prior to using PRAs in future risk-informed management and organizational factors on plant regulatory activities.
Given that some safety. (
Reference:
see Table C.1, #4) commentors felt that the PRA attributes were too demanding, overly prescriptive and beyond the current state-of-the-art, it follows that if they
Response
were assumed to be requirements then they could be interpreted as a significant burden on the The author (of the comments summarized above) industry. Severalcomments emphasizedthatthe acknowledged that the " comments a e notjust on scope and attributes of a PD A to be used for risk-the NUREG document etself but an also directed informed regulatory actsities should be towards some overall es. : cts q/ PMs and commensurate with the application. This impnes HMs." However, none of the comments appear that PRAs with significantly less attributes and of to address the NUREG itself. Nevenheless the more limited scope than the PRA described in NUREG-1560 would be acceptable for risk-NRC does currently have programs addressing inf rmed applications.
Other commentors each of the issues raised by the author, e.g.,
stressed that any applications of the PRA development of improved HRA methods and attributes in NUREG-1560 to the creation of an consideration of the impact of management and industry standard should be viewed as organizational factors on plant safety. Further, developmental in nature.
An industry-wide the NRC staff has reviewed the comments and standard for PRA quality should be based on a will consider them in future directions of broader and more deliberate development effort research.
that involves practitioners from various NUREG-1560 C-16
~- -
App C. Comments and Responses organizations. (
Reference:
see Table C.1, #1,2, C.7 Section 7.1 and Chapter 15:
8,9,15,16,20,22 and 26)
Safety Goal Implications i
l
Response
Several comments were rceived expressing technical j
disagreement with some of the information provided j
Chapters 6,14 and 15 of draft NUREG-1560 in these chapters. The text was revised where have been significantly revised for the final appropriate. In addition, several general comments report. Specifically, Chapters 14 and 15 have were provided on the content of these chapters.
]
These comments and associated responses are been replaced with a new Chapter 14, and pr vided below.
referencesto the use of the IPEs in risk-informed regulation have been removed. Chapters 6 and s & the results 1.
Comment: The conce 14 in the final report summarize f'RA reported in the original It 31s are not characteristics and state that they:
current and could be misleadmg wl.
- ompared to the Safety Goals. For example, several plants are not " standards" nor do they represent identified in Chapters 7 and 16 in Draft NUREG-
]
e regulatory guidance.
1560 (Chapter 15 in Final NUREG-1560) as potentially approaching the early fatality are included only as a benchmark in order to quantitative health objective (QHO) have subsequently updated their PRAs with significant draw perspectives on the models and
'*'E* **I '* ****
methods used in the IPEs.
frequency, LERF (including Browns Ferry, Beaver Valley and Palo Verde). (
Reference:
see do not define the needed quality or scope of Table C.1, #22,25 and 26) e the PRA clements needed for a particular,s,
j regulatory application.
~
Response
2.
Comment: Several comments were related to NUREG-1560 has been revised to clarify that the the following statement in" draft NUREG-1560, Perspectives on the safety goal are based on the rig nalIPEs/PRAswhichmayhavesubsequently
".and other utility personnel are excludedfrom changed.
However, the results quoted m the peer review team."
This statement was NUREG-1560 will not be revised.
New interpreted by some comm-tors as implying tha'.
information obtained by the staffwill b: included no employees of any utility can serve as a peer in NUREG-1560 (see Appendix B). In the case reviewer. (
Reference:
see Table C.1, #1, 8,12, of the safety goal comparisons if any of the 15,16,20 and 22) plants that were identified as approacWg the early fatality QHO submit revised results, this will be noted in Chapter 7 and 15 and the reader
Response
will be directed to the appendix.
l His interpretation was not intended.
The 2.
Comment: Inferences that a few plants may statement was included simply to indicate that it approach the early fatality health objective based would be inappropriate for tatility staff to be part on a comparison of the IPE and NUREG-il50 of the PRA peer review team for plants owned results may not be valid. Additional insights and operated by their utility. NUREG-1560 has gained from the containment performance been revised accordingly, evaluations and recent research in the area may C 17 NUREG-1560
. - ~...
App C. Comments and Responses i
lead to different conclusions than the NUREG-define absolute risk levels, but rather to identify 1150 analyses. (
Reference:
see Table C.1, #16) plant severe accident vulnerabilities.
Consequently, the safety goal computations performed by the staff (described in Chapters 7
Response
and 16 of draft NUREG-1560) are not an NUREG-1560 has been revised to clarify that adequate technical basis on which such a NUREG il50 containment results were not used conclusion can be drawn. In a related comment, to link the IPE results to the safety goals. For SECY-90-104 was quoted, " based on the early4htality risk, a two step process was used.
sigmficant additional resources that would be In the first step, the frequencies of early requiredto make a meaningfut comparison ofthe containment failure and bypass were obtained IPE results with the safety goal policy statement from the 1 pes and plants with low frequencies and the potentialproblems associated with using
(<10 /ry) were screened out from further the as-submittedIPE data, the stagrecommends 4
consideration. For the remaining plants, the that no direct comparisons be made unless the frequencies of source terms with relatively large IPEs are reviewd to a greater level of detail release fractions (>0.03 Cs, I, Te) were obtained.
than currently planned" As the commentor The source term frequencies were then adjusted believes that a review of greater detail did not for population and compared to the goal, occur, it was recommended that the direct comparison of IPEs to :he Safety Goals in 3.
Comment: There is an implication in the report Chapters 7 and 16 be removed from the f' mal that the only way a comparison can be made to NUREG. (
Reference:
see Table C.1, #19,22) the " Safety Goal" is to have a Level 3 PRA.
Such a PRA was never mandated, requested or
Response
suggested by the NRC and there are a number of ways to compare to 'he Safety Goal other than The final version of NUREG-1560 has been having a Level 3 ERA. The NUREG could revised to clearly describe the limitations of the address how the NRC and industry (there are approach used to compare the IPE results to the several EPRI documents and other papers, safety goals and subsidiary objectives. However, positions and reports) have defined or linked the the use of Level I and 2 indicator (CDF and NRC " Safety Goal" in ter.ns of Level 1 and 2 LERF) as surrogates for the safety goals is surrogate indicators. (
Reference:
see Table C.1, consister.t with recent industry positions (refer to
- 15, 21)
Comment #3 in Section C.7 above) and consistent with the guidance provided by the
~
NRC for use of PRAs in risk-informed regulatory
Response
applications (Ref. C.5).
The manner in Tne approach used by the staff in Chapters 7 and which the IPE results are compared to the safety 15 of NUREG-1560 was based u.. asing Level 1 goals is consistent with the " Integration Plan for and 2 surrogate indicators to link the IPE results Closure of Severe Accident Issues," SECY 88-to the safety goals. The wording in Section 6.4 147 and also consistent with the has, therefore, been changed to make it clear that recommendations of SECY-90-104, namely, a Level 3 PRA is not the only way to make a
" indirect comparison of the IPEs and other comparison to the safety Goals.
available PRAs with the Safety Goals, focusing on the insights gained and the adequacy of 4.
Comment: One comment stated that conclusions regulations, is planned." The SECY further based on using the IPE results for ccmparisons to recommends that the " stag evaluate the IPE
. the QHos of the safety goals must be carefully results as a whole and summarize any qualitled. The purpose of the IPEs was not to conclusions and recommendations for the NUREG-1560 C-18
1 1
J App C. Comments and Responses Commission at the completion of the JPE C.8 Section 7.2 and Chapter 16:
P' "" ~
Impact of Station Blackout 5.
Comment: Several verbal and written questions Rule on Core Damage
)
were received at the workshop related to the Frequencies appropriateness of the current safety goals and the manner in which comparisons were made to Several comment: were received expressing technical these goals. (
Reference:
see Table C.1, #23,26) disagreement with some of the information provided in these chapters. The text was revised wl.ere
Response
appropriate. In addition, several general commer.:s were provided on the content of these chapters.
l The appropriateness of the current safety goals is These comment and associated responses are provided below.
a policy issue and outside the scope of NUREG.
1560. The use of Level I and 2 indicators as I.
Comment: Evaluation of the SBO rule wousu surrogates for the safety goals is consistent with benefit from a review of the results by j
the staffs guidance provided in the recently A
ec g eer a n us macer type published regulatory guides (Ref. C.x).
(
Reference:
see Table C.1, #16) 6.
Comment: The definition of an early release' Response: As is discussed in the response to particularly a large early release, and the time similar comments on Chapters 3 and 1I (and the available for effective evacuation after declaration report in general), early in the IPE Insights of a general emergency appears to be arbitrary.
Program the plants were grouped by I
Consideration of the accident timing, the site, and architect /engineerand the IPE CDFs within and the impacts on evacuation (such as an SBO) need among these groups were compared. No strong to be considered. (
Reference:
see Table C.I.
correlation with the architect / engineer was found
- 25) because there is considerable design variability even among plants designed by the same
Response
architect / engineer. A deci; ion was made to perform the analysis using plant groups based A unique definition of a large early release was upon the NSSS vendor to account for basic NSSS not provided :n NUREG 1560. A large early design differences. The BWRs were further tub-t release is defined in the staff's regulatory guides categori:'ed by vintage to account for differences (Ref. C.x) on the use of PRA m risk-mformed CCS design. The Westinghouse plan's were STWPed according to the number of loops since regulation.
Numerical objectives for the the ECCS and other general plant features for the frequency of a large early release are also plants m each of these groups are generally the provided in those documents. The frequencies of same (see Table 10.3). It is recognized that the early containment failure and bypass were used in balance of plant inc;uding support systems for NUREG 1560 to screen out plants with low piants in each of the designated groups can be frequencies. The frequency of source terms with different and skew any comparison of the results relatively large release fractions were then for a plant group. The NUREG consistently examined in more detail to estimate the potential identifies that these plant-specific features impact early health effects. The assumption was made the results and draws the appropriate conclusions that these releases occur prior to effective offsite on the resulting insights. Finally,it is recognized evacuation. This assumption could overestimate that further subcategorization of plants according the potential for early health effects.
to a selected parametercould be made. However, C-19 NUREG-1560
App C. Comments and Responses variability in other parameters would likely C.10 Chapter 8: Overall impact that comparison. Because of this fact and Conc!Usions and also due to resource limitations, funher Observations subcategorization was not pursued.
Some general comments conceming the content of C.9 Section 7.3 and Chapter 17:
Chapter 8 were received from several organizations and individuals. Responses to these comments are Contparison with NUREG-provided below.
1150 Perspectives 1.
Comment: Due to the nature of the IPE process requested in Generic Letter 88-20 (a search for Several comments were received expressing technical vulnerabilities, not characterization of absolute disagreementwith some of the information provided risk), the applicability of the IPE results for in these chapters. The text was revised where regulatory follow up activity should be limited.
appropriate. In addition, several general comments Section 8.2.4 states that the NRC staff plans were provided on the content of these chapters.
follow-up activities to determine if additional regulatory actions are warranted for plants with j
These comment and associated responses are provided relatively high CDFs or CCFPs. NUREG-1560 below.
does not consider revised CDF and CCFP values provided to the NRC, which in some cases, are 1.
Comment: Chapter 18 in Draft NUREG 1560 substantially different than the original IPE (Chapter 17 in Final NUREG-1560) presents a wbmittal values. Consequently, use of the IPEs comparison of NUREG-il50 results with IPE for comparison to safety goals, identification of results as a whole.
A more interesting
" outlier" plants, and for direction of inspection and follow-up activities should be minimized.
comparison would be between the individual Such actions have the potential to lead to NURFG ll50resultsandthecorrespondingIPEs.
ineffective use of NRC staff and utility resources This would provide a more dete!!cd information in pursuing areas which are known to be on specific modeling issues. (
Reference:
see outdated. The NRC staff should evaluate these Table C.1, #2) changes in the plant CDF and CCFP values before planning follow-up activities. (
Reference:
see Table C.1, #20,22)
Response
Response:
Section 7.3 indicates that the focus of NUREG.
1560 is on comparing global perspectives The IPE results and insights provide a useful discussed in NUREG-1150 with the overall source ofinformation for identifying areas where results of the IPEs. A plant-specific comparison follow-up activities might be warranted. The inf rmationcontainedinNUREG-1560,however, between NUREG-ll50 and the applicable IPE is merely a starting point and is by no means the analyses are provided in the individual SERs on
- '* ** 8 # #*E * '#
the five IPEs. Chapter 17 in the Final NUREG-plant-specific actions are taken, the best available 1560 has been revised to clarify the scope of the information will be considered, including any comparison in NUREG 1560 and to note that revisions to the original IPE submittals, plant-specific comparisons may be found in the recognizing that most of the newer information SERs.
has not yet received staff review. Further, any NUREG-1560 C-20 i
?
App C. Comments and Responses proposed regulatory actions are subject to the the NRC tends to be more concerned with Backfit Rule as described in 10CFR50.109.
eliminating vulnerabilities and reducing risks than with reducing burden.
However, the latter 2.
Comment: The NRC staff's approach in looking objective is desirable and the NRC encourages at CDF and CCFP as independent factors is the industry to submit requests for reduced incorrect. It assumes the existence of either a regulatory burdens in areas where they believe high CDF or high CCFP is evidence on its own that risks are low and substantial cost savings can of a potential concern. In reality, the two factors be achieved.
I, should be looked at together. They are each a
~
i
.part of the overall input to risk, which should be 4.
Comment: The discussion of the Maintenance the figure of merit (the CDF/CCFP criteria do Rule says it is acceptable to use the IPEs to
. not have any established technical connection '.o determine risk significant systems. However,this
[
the QHOs of the Safety Goal). (
Reference:
see is not compatible with the findings about the Table C.1, #8, 22) usefulness of the IPEs for risk-informed j.
regulation. Likewise, the NRC implies that for
Response
inspection purposes the IPEs are adequate for-t them to target areas for plant-specific inspections "Ihe major objectives of the IPE Insights Program but NUREG-1560 states that the PRAs are only
)
are outlined in both the Forward and Introduction adequate to identify dominant accident sequence of NUREG 1560. For at least one of those types and their relative importance. This seems objectives (i.e., providing perspectives on plant inconsistent. Furthermore, the NRC seems to be 3_
feature and assumptions that play a role in the attempting to use PRA information in a selective i
estimation of CDF, containment performance and manner, where it serves their purposes.
l
. human performance),it is useful to look at CDF
(
Reference:
see Table C.1, #22) and CCFP. separately.
The use of these parameters in NUREG-1560 does not imply'that'] [ '
Response
a high value for either parameter alone is a potential concern or will be, the basis for References to the use of the IPEs in risk-regulatory decisions. Instead, the use of these inforrned regulation have been removed from the parametersallows the staff 15 focus individually final version of NUREG 1560. Issues related to on the Level I and Level 2 analyses performed the quality and scope of PRAs needed for risk-for the IPEs, thereby accomplishing the informed regulation are discussed in the staff objectives noted ab:ve.
regulatory guides, and standard review plans.
The role of the IPEs in risk-informed regulation 3.
Comment: Conceming any follow-up iegulatory will be determined in the context of these activities,it's suggested that the investigation and documents, not NUREG-1560.
regulatory considerations not be limited just to the high CDF or CCFP issues. Areas where the 5.
Comment:
The report implies that until l
risk impact is small and the safety benefit is not
" quality" PRA requirements are fully met, PRAs l
appreciable should also be investigated for cannot be used for any regulatory purposes. If' 1
reducedregulatory burden. (
Reference:
see Table that is the case,"as is" PRAs are inappropriate to C.1,#6,16) support such areas as the Maintenance Rule and j
Technical Specification changes.
Such an
Response
interpretation is counterproductive and is not supportive of the PRA Policy that looks to l
The primary focus of the NRC is to assure the enhance use of PRA in regulation commensurate t
{
safety of the public. Therefore,it is natural that with the state-of-the-art technology. Recognized 1
C-21 NUREG-1560 L-
l App C. Comments and Responses weaknesses, and tools to deal with those submittals to the NRC, it is important for the weaknesses delineated in the Standard Review NRC staff to know if the credited improvements Plan makes the "as is" PRA applicable for a wide have been made.
j i
variety of applications while " quality" PRA 7.
Comment: The use of NUREG-1560 for a requirements are phased in. Waiting until perfect
" quality" of PRA is achieved before utilizing the variety of issues is discussed in Chapter 8.
results is impractical.
It is expected that However, most of the discussions are actually related to the use of the IPEs to address these
" quality" and " standardization " will evolve, not issues. NUREG-1560 should not be the source througtra priori definition, but through frequent, repeated application and peer review of PRAs.
of information for applications as discussed in
(
Reference:
see Table C.1, #15,17)
Chapter 8.
The IPEs/PRAs are the primary source and should be used. (
Reference:
see Table C.I,#8)
Response
The comment is similar to comments received on
Response
Chapters 6 and 14 (refer to Comment #1 in Section C.6). These chapters and Chapter 8 have NUREG-1560 summarizes a great deal of been significantly revised for the final report. It important safety information and provides a was not intended to imply that all the attributes starting point for identifying and addressing a in draft NUREG-1560 have to be met before a number ofimportant safety issues. As such,it is j
PRA can be use to support risk-informed an important document and staff resource.
However, the staff recognizes that some of the regulatory applications.
information is out of date and that the individual 6.
Comment: The NUREG states the NRC staff submittals contain more information. For any plans to conduct follow-up activities to monitor particular issue, the staff will use the best implementation of the potential plant available information, including any new improvements identified by the IPEs. The submittals, recognizing that some of the new improvements were identified as " potential information may require additional review.
NUREG-1560 also provides comparisons among improvements" which in most cases were identified as areas for further review. The NRC the IPEs on selected issues, and this information seems to be taking them as having been is useful to the staff when evaluating the commitments. These improvements should not treatment of an issue by a panicular plant.
be treated as commitments unless the utility clearly identified them as commitments.
(
Reference:
see Table C.1, #8)
C.11 Chapter 10: Background for Obtaining IPE Perspectives
Response
The NRC recognizes that the potential Several comments were received concerning the accuracy of the information provided in this chapter.
improvements are not commitments in a regulatory sense. However, in many cases the Corrections were made to the text where appropriate.
improvements were credited in the IPE.
No general comments were made concerning this
'Therefore, if the licensee uses the IPE in future chapter.
i NUREG 1560 C-22
-.~.._._
App C. Comments and Responses REFERENCES FOR APPENDIX C i
C.1
. Federal Register,
- Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance, Summary Report, Draft." Vol. 61, No. 221, Page 58429, November 14, 1996.
Federal Register, Individual Plant Examination Program: Perspectives on Reactor Safety and Plant
- Performance, Volume 2, Parts 2-5, Draft," Vol. 61, No. 239, Page 65248, December 11,1996.
C.2
. Federal Register, " Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance, Volume 2, Parts 2-5, Draft," Vol. 61 No. 239, Page 65248, December 11,1996.
C.3 NRC Memorandum (From Mary Drouin to M. Wayne Hodges), " Draft NUREG 1560 Public Workshop Summary Report," October 3,1997.
. C.4
- ACRS meetings on IPE insights:
November 18,1993 January 26,1996 (s abcommittee) i December 10,1993 February 8,19%
September 27,1994 May 23,1996 October 7,1994 June 11,1996 (subcommittee) i December 7,1995 June 12,1996 C.5 USNRC, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Current Licensing Basis," Draft Regulatory Guide DG-1061, June 1997.
USNRC, "An Approach for Plant-Specific, Risk-Informed Decisionmaking: Inservice Testing," Draft Regulatory Guide DG-1062, June 1997.-
USNRC, "An Approach for Plant Specific, Risk-Informed Decisionmaking: Graded Quality Assurance,"
Draft Regulatory Guide DG-1064, June 1997.
USNRC,"An Approach for Plant-Specific, Risk-informed Decisionmaking:TechnicalSpecifications," Draft Regulatory Guide DG-1065, June 1997.
1 C-23 NUREG 1560 4
)
y Summary of Results of IPEEE Reviews On June 28,1991, the NRC issued Supplement No. 4 to Generic Letter (GL) 88-20, which described the objectives and overalllogistics of the Individual Plant Examination of Extemal Events (IPEEE) program, for the evaluation of extemal events including seismic events, intemal fires, and high winds, floods, and other (HFO) extemal initiators. The primary goal of the IPEEE program has been for licensees to " identify plant-specific vulnerabilities to severe accidents that could be fixed with low-cost improvements."
In addition to the principal intent of GL 88-20, the four supporting IPEEE objectives have been for each licensee to:
1.
develop an appreciation of severe accident behavior; 2.
understand the most likely severe accident sequences that could occur at the licensee's plant under full power operating conditions; r
3.
gain a qualitative understanding of the overalllikelihood of core damage and fission t
product releases; 4.
reduce, if necessary, the overall likelihood of core damage and radioactive material releases by modifying, where appropriate, hardware and procedures that would help
' prevent or mitigate severe accidents.
The recommended guidelines of NUREG-1407, " Procedural and Submittal Guidance for the IPEEE for Severe Accident Vulnerabilities," and Supplement 5 to GL 88-20 have been developed by the NRC to help ensure that each of these objectives is met.
l Preliminary perspectives have been obtained from ongoing technical reviews of 24 IPEEE studies that have been submitted by licenses. These results primarily include: (a) summaries of findings reported in IPEEEs, and (b) perspectives on strengths and weaknesses of licensee submittals in achieving the IPEEE objectives. The IPEEE program appears to have been generally successful in meeting the overall intent of GL 88-20. However, the degree and consistency of such success have varied considerably from study to study, and have been strongly dependent on the level of detail, and methods and assumptions employed in the IPEEE analyses.
Based on the continuing review of the first 24 IPEEE submittals, it appears that the IPEEE program has led to an increase in overall licensee appreciation of severe accident behavior for 4
~
extemal events. As requested in NUREG-1407, each IPEEE has involved a seismic evaluation, an analysis of intemal fires, and an assessment for HFO events. These evaluations have
_ assessed the potential for extemally initiated severe accidents, and have evaluated plant-
- L specific behavior in responding to potential severe accidents.
+
A7i
i Consistent with the guidance of NUREG-1407, the emphasis in conducting IPEEEs has been on obtaining a qualitative, as opposed to quantitative, understanding. As expected, therefore, the IPEEEs do not generally convey a definitive ranking of the risk-significance of severe accident sequences or of the dominant risk contributors. Rather, by means of systems modeling and screening analysis, licensees have obtained a greater awareness of severe accident sequences and an improved sense as to the most important among those sequences.
By means of IPEEEs, licensees have been able to generally ascertain whether the risk of core damage associated with each extemal initiator is comparatively negligible (i.e., falling below the 104 screening threshold), low, moderate, or high. In some cases, this understanding occurred through direct quantification of core damage frequency (CDF), whereas in other cases, this understanding resulted from having knowledge of the hazard in conjunction with an assessment r
of the plant's ability to withstand the hazard.
it is important to note that, although in many cases licensees have reported numerical risk estimates (for CDF or frequency of significant radiological releases), the accuracy of such estimates is frequently limited due to simplifying assumptions and approximate procedures employed in the analyses. Hence, the results serve only as general indicators of risk level, and a comparison of CDF results between plants is not particularly meaningful.
Based on the first 24 IPEEE submittals, a majority of licensees have implemented or proposed plant modifications that have a beneficial effect on plant safety with respect to extemal events.
Such plant modifications include hardware changes, procedural changes, and implementation of severe accident management guidelines. Consistent with the qualitative nature of the IPEEE program, it is not usually possible to deduce the numerical risk reductions achieved by these modifications. However, some licensees have employed PRA in their IPEEEs as a means for determining whether or not plant modifications are warranted based on cost-benefit rationale.
Licensees have in most cases followed the guidance in NUREG-1407 in performing their IPEEE assessments. The guidance permits altemative methodologies. For example, there are various approaches for the seismic evaluation, including; seismic margin assessment (SMA) i i
using the NRC methodology, SMA using the Electric Power Research Institute (EPRI) methodology, or seismic PRA methodology. Out of the 24 IPEEE submittals reviewed to date, thirteen have been based on seismic PRA methodology, whereas seven have been performed using the EPRI SMA methodology; one has adopted both the seismic PRA and EPRI SMA approaches; one has been based on the NRC SMA methodology; and two have been performed using a site-specific seismic evaluation approach, in consideration of the " Optional Methodologies" provision of NUREG-1407.
NUREG-1407 has also allowed for the implementation of attemative approaches for the evaluation of intemal fires and HFO events. For fire IPEEE evaluations, licensees have implemented EPRI's fire-induced vulnerability evaluation (FIVE) methodology, fire PRA methodology, or a combination of these approaches. For evaluation of HFO events, the licensees either demonstrated that criteria of the NRC's 1975 Standard Review Plan (SRP) were met, or conducted at least one of the following forms of analysis: screening assessment, bounding analysis, or PRA methodology.
A7ii
4 These observations highlight an important fundamental difference between the IPE process and the IPEEE process. In the IPE process, comparatively detailed PRA investigation has been i
invariably implemented, whereas in the IPEEE process, a mix of deterministic methods and detailed PRA investigations has been applied, as well as a mix of screening analyses, simplified hazard-based analyses, and/or bounding PRA-based approaches.
Licensees have per#ormed or proposed various IPEEE-related improvements for their plants in seismic, fire, and/or HFO events. In the seismic area, hardware fixes have included items such as: anchoring equipment, bolting cabinets together, improving existing anchorage or supports, installing missing fastener and bolts, installing spacers on battery racks, eliminating potential interaction concems, and replacing vulnerable relays. Maintenance actions have included the removal of corrosion on equipment anchorages and application of corrosion protection.
Maintenance procedural enhancements have included provisions for proper storage of ladders, tools, gas cylinders, etc., and for proper parking of cranes and chain hoists.
Improvements to fire protection systems include hardware modifications and enhancements to, or development of, fire-response procedures. Additionally, improvements have often taken the form of severe acudent management guidelines that address specific accident scenarios related to intemal fires, potential effects of wind-induced missiles, and external flooding.
Implementation of some of the severe accident management guidelines has led to the acquisition or access of temporary or portable equipment (pumps, diesel oil tanker trucks, etc.).
One HFO IPEEE reported the strengthening of the stacks of two adjacent fossil-fuel units to reduce the high-wind risk, and refurbishment of a flood wall to reduce flood risk. The IPEEEs have also, in some cases, referenced plant improvements that had been made (or proposed)
. prior to the IPEEE, since those improvements resulted in a beneficial effect on plant safety in the face of seismic, fire, and/or HFO events. For example, at one plant, the addition of diesel generators was identified as a plant improvement in the IPE, and was correspondingly reported in the IPEEE since it reduces the risk of station blackout for seismic, fire, and HFO events.
A number of important perspectives and insights have been derived from the NRC's overall review activities pertaining to IPEEE submittals. Some of these key observations are described in the following paragraphs for the seismic, fire and HFO aspects of the IPEEE program. It was stated previously that the intent of GL 88-20 appears to be broadly met by the IPEEE submittals; however, the quality of the submittals has varied considerably from plant to plant.
Some of the weakness in the submittals are discussed below. When these weakness have been observed during the review of the submittals, the staff has sent requests for additional information (RAI) to the licensees to complete its assessment of the submittal.
Seismic Events i
Key observations obtained from a review of seismic IPEEEs include the following:
e A seismic walkdown was performed for each plant, and in most cases, the walkdown identified conditions pertaining to anchorage, interactions, maintenance, and/or housekeeping that required further investigation. As a result, a number of plant-specific fixes have been implemented at many plants.
j l
A7-iii
in seismic PRA studies, different hazard curves have been used (i.e.,1993 Lawrence
^
j e
Livermore National Laboratories [LLNL),1989 LLNL,1989 EPRI, and individual licensee-sponsored contractor results) from plant to plant. Hence, it is difficult to achieve a meaningful comparison of seismic CDFs across plants. However, the ranking
)
of dominant contributors has consistently been reported in seismic IPEEEs as being insensitive to use of the EPRI or LLNL seismic hazard curves.
Simplifications in systems analyses, unsubstantiated assumptions regarding human e
i error rates, and use of simplified screening fragilities have, in some cases, obscured findings pnJning to dominant seismic risk contributors and produced unrealistic (high or low) CLP estimates, Although the analytical methods vary and there are large uncertainties, it appears that e
the CDF contribution from seismic events can, in some cases, approach that from internal events.
Fire Events Key observations obtained from a review of fire IPEEEs include the following:
No fire vulnerabilities have been reported in the first 24 IPEEE submittals; however, fire-e initiated accidents have been found to be an important component of the external events CDF contribution.
While no plants have identified fire vulnerabilities, about half (of the 24 reviewed) have reported some fire-related plant improvement as a result of the IPEEE effort. These improvements include changes to existing procedures, development of new procedures, or plant modifications.
Overall licensees have expended a considerable level of effort in conducting fire o
IPEEEs. A few submittals clearly demonstrated the proper application of fire risk methodologies and data. However several weaknesses have been noted in applying the selected methods and data in some of the fire analyses which affect the robustness and completeness of the submittals. Some of these weaknesses are as follows:
Operator actions in response to the effects of fire on systems have rarely been modeled in detail.
Several submittals have used questionable methods, procedures, or data for fire l
damage modeling.
]
i
. Several submittals have used the Nuclear Safety Analysis Center (NSAC)/181 and/or the EPRI Fire PRA Implementation Guide documents for which some optimistic guidelines and data have been identified.
A7 iv
The possibility of active barrier failure, which may have a significant probability of occurrence, has not been included in most analyses. The significance of active fire barriers is a function of plant layout and separation of redundant trains. Also, the potential for barrier failure associated with large quantities of combustible materials concentrated in one area has not been considered in most of the submittals.
Although the analytical methods vary and there are large uncertainties, it appears that e
the CDF contribution from fire events can, in some cases, approach that from intemal events.
HFO Events (i e. hiah winds. floods. transoortation accidents and nearbv facility)
Key observations from HFO IPEEE submittals include:
Transpor%on and nearby facility accidents have been screened out in all of the 24
- lPEEEs that have been reviewed.
The HFO IPEEE program has had some impact on plant safety. For some plants a e
greater appreciation of the potential risk impact of high winds /tomadoes and extemal flooding / dam breaks has resulted from the IPEEE program. Some licensees have implemented or proposed plant improvements with respect to procedural enhancements, severe accident management guidelines, and hardware ir.stallation.
Procedural enhancements include sandbagging, closing doors, welding doors, hooking up pumps, and creating new circuits to reduce the risk from flooding. In two submittals, development of severe accident management guidance to reduce the risk of high winds is being considered. Hardware improvements include, for example, modifications that enhance flood protection.
Potential failures of upstream dams, leading to flooding at the site, were considered and screened out in many of the first 24 submittals. However, generic dam failure data has been employed in all cases without considering site-specific information such as dam type and vintage, e
in general, the CDF contribution from PCO is lower than that from intemal events.
Generic issues The IPEdE program has addressed a number of generic issues (Gis) and unresolved safety issues (USis) including USl A-45 (" Decay Heat Removal Requirements"), GI-131 (" Potential Interaction involving the in-Core Flux Mapping System at Westinghouse Plants"), GI-57
(" Effects of Fire Protection System Actuation on Safety Related Equipment"), Sandia fire ri:k A7v 4
-n
.n e
m>
scoping study issues, and GI-103 (" Probable Maximum Precipitation [PMP]"). Some key observations from the review of the first 24 IPEEE submittals include the following:
In general the seismic and fire evaluations of the IPEEE are capable of addressing USI e
A-45, without any special additional considerations. Also, no HFO evaluation reported any open issue pertaining to USI A-45.
For most applicable plants, GI-131 had been addressed through earlier upgrades and e
analyses. Some IPEEEs evaluated the capability of the in-core flux mapping system for beyond design-basis seismic loads consistent with the IPEEE review level earthquake (RLE).
Almost all licensees have followed the guidance in FIVE pertaining to the evaluation of the fire risk scoping study issues and GI-57. - In a few cases, seismic-fire outliers have been noted.~ No submittals have reported risk-significant findings associated with either the fire risk scoping study issues or GI-57.
Most submittals addressed the effects of increased PMP criteria with respect to roo' l
ponding and flooding due to intense local precipitation. In all such cases, the impacts of GI-103 were found to be accommodated by the existing plant design.
The IPEEE submittals also provide information relevant to some other generic safety e
issues (GSis) even though the submittals were not explicitly requested to treat, and the IPEEE program was not originally intended to resolve, such issues. These issues include: GSI-147 (" Fire-induced Altemate Shutdown / Control Panel interaction"); GSI-148 (Smoke Control and Manual Fire-Fighting Effectiveness"); GSI-156 (" Systematic
)
Evaluation Program (SEP]"); and GSI-172 (" Multiple System Responses Program
[MSRP]"). The IPEEE review process has identified the extent to which the submittals provide information relevant to these GSIs, and how these issues can be considered to be resolved.
QUAD Cities Fire IPEEE j
Although not part of the first 24 IPEEE submittals reviewed, the Quad Cities fire IPEEE submittal review has revealed some particularly significt.nt perspectives related to fire risk. A brief summary of the licensee's fire IPEEE process anc findings is provided below.
The licensee's fire assessment employed EPRI's Fire-:nduced Vulnerability Evaluation (FIVE) method for initial screening and EPRI's Fire PRA Im%ementation Guide for detailed evaluations for screened-in fire areas. These evaluations include: assessment of individual sources that can damage safety targets (i.e., safety shutdown equipment); identification of fire scenarios, taking into account fire features, such as detect lon and suppression; determination of conditional core damage probability for the specific fire targets; and calculation of a scenario-specific core damage frequency (CDF) value. Additionally, multi-compartment fire scenarios were considered in the event that the fire barriers credited in the single compartment analyses are unable to prevent fire propagation in adjacent compartments. Walkdowns were also A7 vi e
conducted by Quad Cities plant engineers together with supporting contractors in order to:
verify the compartment data; assess the seismic / fire interactions; identify the potential fire sources, safety targets, and locations of fire detection and suppression systems; and inspect the fire barriers.
The licensee estimated a total fire CDF at the Quad Cities to be about 5x10-8 per reactor year (RY). The licensee reported that the top five accident sequences, which contributed about 40% of the total fire CDF, were all in the turbine building involving postulated oil fires. The licensee stated that, even though the plant is in compliance with the NRC regulations, the lack of separation of certain cables in the turbine building and the complicated procedures needed for recovery actions were responsible for the high CDF nurrber. The licensee used Nuclear Energy Institute's (NEl's) severe accident vulnerability criteria (e.g., CDF exceeds 1x10" per reactor year) and identified fire at the plant as a potential severe accident vulnerability. The licensee has implemented an interim alternate shutdown method involving the use of an independent back-up power supply for both units at Quad Cities to reduce the fire CDF from 5x10-8 per reactor year to 7x10" per reactor year. Curre..tly, the licensee is evaluating long-term measures to further reduce the fire CDFs ar.d is keeping the staff informed cbout its progress.
e j
A7 vii
..