ML20147G980
| ML20147G980 | |
| Person / Time | |
|---|---|
| Issue date: | 12/08/1988 |
| From: | Stello V NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO) |
| To: | |
| Shared Package | |
| ML20147G663 | List: |
| References | |
| TASK-PII, TASK-SE SECY-87-297, NUDOCS 8803080354 | |
| Download: ML20147G980 (18) | |
Text
3mm -amou=.wh_
m_.. ;.,,
s +
- At 4..... W.$1.p ',f. ; f.,L,.~.S.t kMx. 2 FL.
'an POLICY ISSUE December 8, 1987 (Infom18 tion)
.SECY-87-297 For:
The Comissioners i
From:
Victor Stello, Jr.
Executive Director for Operations
Subject:
MARK I CONTAINMENT PERFORMANCE PROGRAM ~ PLAN
Purpose:
To present staff plans to resolve issues relating to the performance of MARK I containments during severe accidents.
~
Sumary:
In this paper the staff proposes a plan to effect closure of generic PJdtX I containment performance issues. The plan.
stems from a staff' judgment that MARK I containments can have an improved level of mitigation under severe accident conditions. For those issues for which sufficient informa-tion exists, closure is to be effected by interim recomendations to the Comission in A?ril 1966, and final recomendations in August 1988. For those issues for which sufficient information coes not exist to effect closure, the severe accident research program will be used to provide bases for potential future recomendations.
Assessments of accident sequences indicate that the're is i
substantial safety margin in the ability of NARK I contain-ments to attenuate accidentally released fission products.
There are, nevertheless, some low probability severe accident sequences for which the integrity of the con-tainment function can be seriously challenged.
The key issue is reasonable assurance of the capability of I
containment systems to mitigate the consequences of core melt accidents for moderate to low probability sequences.
This issue should be viewed in terms of defense-irf-depth in i
that it involves striking a balance between accident prevention and mitigation. The plan is intended to achieve l
l
Contact:
J. Hulman, RES 492-6016, 443-7622 P
l
j l
2 regulatory resolution of contairment performance issues starting with Boiling Water Reactor (BWR) MARK I contain-ments in 1988. The plan includes development and i
application of criteria for judging of containment perfor-mance.
Discussion:
The staff briefed the Conunission on a plan for closure of severe accident issues, including matters relating to SWR MARK I containmerts, on July 15, 1987. At this briefing, the staff indicated its intent to pursue an integrated approach to the resolution of severe accident issues.
Included were the Individual. Plant Examination (IPE) program, a containment performance program for each of the various containment types, a program to improve plant operations, and a program to' provide guidance on severe accident management strategies.
In addition, severe accident / source term and risk reassessment research programs support the integrated program.
The closure plan for MARK I containments calls for a two step process; 1) an NRC staff, researcher and industry ) a identification and narrowing of technical issues, and 2 staff evaluation process. The issues would include those associated with core melt phenomena, contairunent failure modes, and those associated with the efficacy of potential improvements, Many analyses of MARK I containment perfor-mance have been done by the staff, staff contractors and industry analysts. This work will fonn the primary bases for issue identification. To aid in narrowing and focusing issues, additional in-vessel and ex-vessel core melt pro-gression calculations are to be made and related experimental data are to be assessed. The staff evaluation would serve to eliminate some generic issues as not sufficiently important to consider' further, to undertake research to provide suffi-cient information to resolve other issues, or to recommend It is the identification and nar-regulatory initiatives.
rowing of issues, focusing related research, and assessing whether improvements are justified that the staff will pursue l
t in this program.
An interim report will be provided to the Consnission in The report will discuss the major areas of April 1988.
agreement and disagreement between analysts and researchers on the important issues, and will indicate whether analyses at that time justify reconenendations for near-tenn improvements to MARK I containments. A fiaal report for MARK I containments is scheduled for August 1988.
The containment performance effort is being carried forward The philosophy for an by RES in coordination with NRR.
approach to the evaluation of MARK I issues is descrihd in a memo from T. E. Murley to V. Stello, Jr., dated June 29, 1987 (Enclosure 1). This philosophy was used as the basis for
3 the staff severe accident discussion with the Comission on July 15, 1987. Enclosure 2 is the staff's plan for resolu-tion of MRK I issues. A descr'ption of staff plans for integration and closure of all severe accident issues is scheduled to reach the Comission in April. Enclosure 3 illustrates the relationships between the primary MRK I tasks, and indicates several important milestones. This effort will be coordinated with the anticipated utility and 5
staff IPE efforts, and will use such information as can )e provided from utilities.
While recognizing the importance of individual plant variations, the staff has also recognized that there are potential severe accident vulnerabilities that have a common character within a class of plant containment systems. This recognition has evolved from severe accident research and plant evaluation programs both here and abroad, including findings from numerous probabilistic risk analyses starting with the Reactor Safety Study and including the recent draft HUREG-1150. For BWR MRX I containments in particular, these vulnerabilities are reflected in rela-tively high estimates of the probability of containment failure, given a core melt (also referred to as the condi-tional containment, failure probability, CCFP). Staff spon-sored research' presented in draft NUREG-1150 indicates that this conditional probability is highly uncertain, but could l
be quite high for MARK I plants.
Industry sponsored research, l
on the other hand, has provided estimates of CCFP for two l
reactors (Peach Botton and Vermont Yankee) at less than 10 l
percent. The staff's judgment is that these discrepant views are unlikely to be fully reconciled soon. In view of the Comission's defense-in-depth philosophy, therefore, the staff believes it is prudent to examine ways to improve the capability of MRX I containments to mitigate the poten-tially large fission product releases that could result from outlier accident sequences.
Our examination of these differences in expected MRL I containment performance has resulted in two cone'lusions.
First, many technical differences may be narrowed by further discussions among staff, researchers, industry representatives and interested members of the public. The discussions and any regulatory decisionmaking can be fac'litated significantly by short term analyses and assessments of existing experimental activities related to BWR MRX I core melt phenomena and containment response.
Second, sone residual differences are likely to remain that only answers from a relatively long term research progran can provide.
In an August 11, 1987, meno from S. J. Chilk to V.
Stello, Jr., the Commission requested (H870715A) an
4 assessaent of whether or not additional resources for this activity could be used effectively. Resources were budgeted for FY 88 and subsequent years for related activities in the recent RES budget submittal. No additional resources are considered necessary for FY 88.
In sumary, the approach identified in this paper is ex-pected to result in both improvements in our understanding of the performance of MARK I containments during severe accidents, and in the identification of potential design and operational improvements.
BWR MARK I containments are to be assessed by the end of FY 88, assessments of the other containment types are to be completed by the end of FY 89. A report to the Commission with interim MARK I reconnendations is scheduled for April 1988. An integrated plan for effecting closure of severe accident issues for all plants is also scheduled for submission to the Commission in April 1988. A final report on MARK I containments is scheduled for August 1988.
/
.N Y ctor Stello, xecutive Director fo Operations Enclosures 1.
June 29, 1987 memo from T. E. Murley to Y. Stello, Jr.
2.
Program Plan 3.
MARK I Key Activities & Milestones
Contact:
J. Hulman, RES 492-8016, 443-7622 J'
i l
ENCLOSURE 1
~[*
UNITED sT ATE s
,'k NUCLE AR REGUL ATORY COMMISSION i*
wassiuc ton. D. C. 20$$$
e...e JUN 2 91987 l
MEMORANDUM fJR:
Victor Stello. Jr.
Executive Director for Op.erations FROM:
Thomas E. Hurley, Director Office of Nuclear Reactor Regulation
SUBJECT:
PROPOSED COURSE OF ACTION ON BWR MARK 1 CONTAlHMENT Your memorandum of April 20, 1987 directed NRR to detennine a recomended czurse of action with regard to earlier proposals for an initiative to enhance BWR containment perfonnance in the event of.a severe core damage accident.
The staff has for some time recognized the potential vulnerability of BWR Mark I containments under certain severe accident conditions (see e.g.. NUREGs 1079 and 1150) and, as a result, has studied means for reducing the Mark I containment failure probability. Last year the staff developed a set of proposed generic improvements with the general intention of reducing the conditional probability of Mark I containment failure during severe accidents. It was thought that, if these improvements were implemented, it would be unnecessary for these BWR plants to have containment perfonnance evaluatedaspartoftheIndividualPlantExaminations(IPE).
In the intervening time since the generic improvements were put forvard there have been several discussions among the staff, industry groups and the research corrtounity. The Reactor Risk Reference Document (NUREG 1150) was completed in February 1987 as well. The conclusion that seems to have emerged from these activities is that there is no clear consensus on whether t)e Mark I generic improvements are needed, whether the cost estimates are realistic, and whether the proposed improvements would be effective in significantly reducing risks. After reviewing these matters I have conclud2d that a more comprehensive approach to this issue should be taken.
The approach outlined below is net intended to delay clear safety improvements but rather to ensure we look at all reactor types and understand those areas where we are most likely to attain safety improvements.
e
V. Stello, Jr. M O IC in examining the broad question of how to reduce the risks of severe accidents, the following three areas must be considered, 1.
IMPROVED PLANT OPERATIONS Every safety stucy since WASH 1400 has sh:,wn the sensitivity of risk to human errors. Our own analysis of operating experience confirms the importance of reducing maintenance, surveillance, testing and control room errors. Thus, an overall approach to this issue must include a program to improve plant operations and should consider at least the tasks below:
(a) Continued improvement of the SALP program; (b) Regular reviews by senior NRC managers to evaluate those plants that may not be met ting hRC and industry standards of operational
- performance; (c) Diagnostic Team Inspections to probe further the performance of those plants above; (d) Regulatory actions to improve operational performance where it has fallen below expected standards; (e)
Improved Technical Specifications; (f) Continued improvement of Emergency Operating Procedures (EOPs); and
[
(g) Expanding E0Ps to include Severe Accident Procedures.
2.
COMPREHENSIVE SEARCH FOR SEVERE ACCIDENT YULNERABILITIES The Severe Accident Policy Statement contemplated a program of Individual Plant Examinations (IPEs) that would be a systematic approach to efamine all plants for possible significant risk contributors. The staff has been working with the IDCOP industry group to develop the IPE methodology and has reached conclusion on a proposed program. The IPE program will have to be integrated with the improved operations program and with the containment performance research program below.
l l
3.
_ CONTAINMENT PERFORMANCE RESEARCH The assessment of containment performance during severe accidents is a very difficult problem, and years of research have not yielded a consensus on what improvements are needed, if any.
We should anticipate there will be the need for a long-range, continuing research program to
YdterStello,Jr
-y-U*M I IE,
assess the challenges to containments, to evaluate potential improvements, and to continue improving our understanding of source terms. Within this long range program there should be near-term results where the weight of technological evidence supports recommendations for containment improvements. The BWR Mark I would be one area targeted for near-term results. Clearly, this r(search effort must be integrated closely with the IPE
~
program which will be examining accident vulnerabilities that could threaten containment integrity at specific plants.
The comprehensive program for reducing severe accident risks outlined above has not been fully developed. A schematic portrayal is shown in the attached figure. When developed and implemented I believe the program should lead to closure of the severe accident issue. Nonetheless, elements of the containment performance research program and the improved plant operations program will no doubt extend well into the future as we gain more research knowledge and more operating experience.
x In keeping with the intent of the reorganization that RES develop resolutions for generic safety matters, I suggest that RES develop, with NRR guidance and support, the overall program outlined above.
I further suggest that RES develop an interim response to the Comnission's request for an options paper (February 9,1987, memo from Chilk to Stello). This interim response would provide an outline of the program discussed above and would provide schedules for implementing key parts of the program such as IPE and containnent performance evaluations.
Finally, because of the importance of this issue, I will continue to work closely with the Director, RES, to coordinate the overall guidance for these activities. Similarly, the NRR and RES staffs will work closely on this program.
original sigsed by N+as I. Earley Thomas E. Hurley, Director Office of Nuclear Reactor Regulation
Enclosure:
As stated l'
cc w/ encl.
E. Beckjord, RES
)ISTRIBUTION
- entral Files
.RStarostecki
'Hurley AThadani ISniezek
'Miraglia
- DF :NRR
- T Murley:kb :
- 6/29/87 OFFICIAL RECORD COPY
SCHEMATIC SEVERE ACCIDENT PROGRAM IMPROVED PLANT OPERATIONS CONTINUING EVALUATIONS SALP MANAGEMENT REVIEWS DIAGNOSTIC INSPEC110NS v
y IMPROVED SEVERE ACCIDENT TECH SPECS PROCEDURES s.
1P INDIVIDUAL PLANT IDENTIFY PLANT SEVERE ACCIDENT EXAMINATIONS VULNERABILITIES CLOSURE A
y y
POTENTIAL MARK I OTilER CONTAINMENT IMPROVEMENTS IMPROVEMENIS A
ss k
CONTAINMENT PERFORMANCE CONTINUING RESEARCH CONTAINMENT s
RESEARCH e
ENCLOSURE 2 9
i MARX 1 CONTAINMENT PERFORMANCE PROGRAM PLAN INTRODUCTION - The ability to mitigate.the consequences of accidents is a function of the containment systems that are provided at all U. S. light water reactors.
One class of containments is referred to as MARK Is, which have been used with 24 licensed BWR reactors. Although all U. S. light water reactors have containments designed to safely attenuate the energy that would be released in a loss-of-coolant accident in which safety systems would function to supply cooling water, MARK I containments have among the smallest internal volumes. This relatively small volume is offset, for some accidents, by a pressure suppression water pool. Such volumes and suppression systems are important if safety systems do not function properly and a large pressure rise ensues from releases of gases such as hydrogen and concrete ablation products.
In addition, MARK I containments have no deep concrete slabs or water pools directly beneath their reactors. As a result, for many severe accidents MARK I containments may be viewed as potentially more susceptible to containment failure than other containment types.
The designs of these containments consider external events (such as earthquakes and tornadoes), while the containment temperature and pressure design bases are determined by a postulated design basis loss of coolant accident (LOCA) in which operation of the emergency core cooling system (ECCS) would prevent a core melt from occurring. The peak MARX I containment pressure associated with such a postulated LOCA has been estimated as high as about 57 psig.
Despite this containment design basis which would not result in core melting, the radiological consequences of a substantial core melt are nevertheless postulated in accordance with the provisions of 10 CFR 100.11. This postulation is used to assure the adequacy of certain plant features such as containment leak tightness and fission product filter systems, as well as the adequacy of the reactor !,ite. The temperature and pressure conditions associated with a core melt accident are not part of the containment design bases. There is some assurance, however, that existing containments are capable of surviving the temperature and pressure conditions associated with some severe accidents, as well as for arrested core melt accidents. The THI accident is an example which represents a partial core melt accident that was arrested prior to reactor pressure vessel failure, and was one in which the containnent was not failed. Furthermore, there is an expectation Jhat some containment failure events may result in significant fission product attenuation in adjacent plant buildings.
Studies of exampics of various containment types under beyond design basis loading conditions (NUREG-1079) indicate survival at load levels of 2 to 3 times design basis LOCA loads and at elevated temperature conditions. Although only a few detailed structural analyses of MARK I containments have been attempted, inferences and extrepolations from design assessments and testing on scale models of containments and penetrations at Sandia National Laboratnry con-firm these higher failure pressure conclusions.
Such confirmation, however, assunct containment isolation devices (including seals) isolate and do not fail.
PRA asserscents to date of some MARK I plants indicate the initiator for the most risk significant accident may be a station blackout (500) event. One
i i
\\
.contenporary MARr, I risk assessment (draft NUREG-1150) indicates an SB0 is dominent, and t5e probability of core melt accident may be very low. Other s
assessments for.a 11nited number of MARK ! plants do not confirm the low prob-ebility conclusion.
It is important to note, therefore, that risk assessments of other MARX Is could identify other risk significant initiators and substan-tially different risk levels.
A generic issue, A-44, dealing with SB0 events is close to completion. The objective of a proposed rule change (51FR9829) is to provide assurance that the probabflity of core melt arising from station blackout will be at or about 10-5 per reactor year ~or less. However, the rule change nay not eliminate the event as a potentially dominant one at some MARK I plants, nor eliminate concern over the ability of such a containment design to mitigate such accidents.
Furthermore, anelyses of the characteristics of other severe accidents indicate the potential for generic concerns over the ability of the BWR MARX 1 containment type to mitigate the consequences of such events.
Stated another way, it is not clear that the balance between accident prevention and mitigation called for in the Severe Accident Policy Statement for the various combinations of reactor types and containments has been achieved.
~
6 CHALLENGES - There are a number of potentially important challenges to MARK I containrents. These are:
i 1)
Containment bypass (including failure to isolate containment on demand, suppression pool bypass, and interfacing system LOCAs);
2)
Early overpressure or overtemperature failures (including sequences involving melt quenching in-vessel, direct containment heating, and noncondensable gas generation and potential ignition);
t 3)
Missiles from rapid steam pressures; 4)
Core debris atteck on the steel containment liner resulting in liner melt through; 5)
Later overtemperature or overpressure failure; and 6)
Basemat penetration.
Recent MARK I assessments have identified early overpressure (2), core debris attack on the steel liner resulting in liner melt through (4), and later over-temperature and overpressure (5) failures as the primary challenges. The like-lihoods of early failure and liner melt through are areas of controversy anong some analysts. The core melt progression phenomena associated with accident sequences which could lead to such challenges, therefore, are also important issues reouiring better understanding.
CONTAlletENT FAILURE MODES - Containment response to challenges can have several
~
outcomes.
These can range from relatively little leakage to large scale failures.
Large scale failure modes can generally be of two types. One type is a slowly develeping breach of containment; e.g., a progressive failure of gasket naterial trcund a containment penetration such as an equipment hatch, a small structural failure of a suppression pool vent pipe exptasion bellows, or a structural tear in the steel liner of a *.oncrete containment.
The other type of failure involves a very rapid depressurization such as would be displayed by the entastrophic rupture of steel containment.* The locations of predicted pressure induced t
+Only two itARK ! containeerts are of reinforced concrete construction; Brunswick Unit 1 and Unit 2.
3 structural f ailures for the MARK I containment MUREG/CR-3653) are in the drywell at either the knuckle between the upper cylinder and lower spherical section, or at the drywell head. Failure in the wetwell air space or suppression pool have also been postulated.
(Leakage at the'drywell head prior to failure in the vatwell airspace has been identified.)
P0TENTIAL IMPROVEMENTS - A small number of relatively low-cost 1.n have been assessed by the staff, its contractors, two utilities (provements Vermont Yankee and Boston Edison), and IDCOR. These improvements may substantially mitigate potential offsite releases. Some of these potential improvements are; a)
Hydrogen centrol - In normal operation MARK Is are inerted by replacing much of the oxygen in the containment atmosphere with nitrogen. Both the time HARK I containments are allowed not to be inerted, and the ability to keep such containments inerted during long duration station blackout sequences, have been questioned.
Information provided by the Vertnont Yankee licensee indicates that the relatively brief periods of time required for inerting and deinerting during startup and shutdown periods may be acceptable.. An analysis of improvenents proposed for Pilgrim by Boston Edison indicates improved nitrogen supplies are potentially warranted for long term ar.:ident sequences.
b)
Containment spray - During a station blackout, power for pumping water to,
." ntainr'ent sprays or the vessel would not be available. One proposal has been
.o cross-tie an existing diesel powered fire pump to the water system for use in the vessel to prevent core n'elt, or for containment spraying if vessel failure has occurred. This proposal has been found technically feasible by t
IDCOR and the Vr'mont Yankee and Pilgrim licensras. An alternate (Hope Creek) h to provide an external valve on the Reactor Building which would allow water M be supplied via a fire truck. Providino spray water durir.g accidents such as a station blackout would serve several functions. Such water can help c14sipate heat, cool core debris, ard scrub fis< ion products from the containment atmosphere. Because th! pumping capability of the fire Sater i
system is a fraction of that of thr containment spray system, t N use of the
( Pe pump without other modifications would not produce an adet ste containment I
spray pattern. By relatively sim>1e modifications, the existing containment spray heads may be modified to etsure an adequate spray coverage for fission product scrubbing and heat removal for some scenarios.
Indeed Boston Rdison has proposed blocking 6 of 7 nozzles in each spray head. The impact of such modifications on other accidents still requires assessment.
c)
Ventino - Venting the containment can reduce core melt probabilities for some accidents, and prevent overpressure or overtemperature failurfs for others.
It may also b* :1*wed as a last ditch effort to prevent the containment from burstir2 (ovrq usure).
tf done before the core melts, little in the way of fission pr~% would tw cleased.
If done after core melting, substantial i
fission F :, a coula.
- W sed. These fission products are of two types; products. Filtering or scrubbing can be effective
'a noble gav,
2 in redue F i~wae4 fv ion products. However, only relatively lone period L.V6,,r 4 t d' a :an he effective in reducing their potential biological i s e3 m
'O. I suppression pool is an excellent potential port-accide.. <cr w - 5:
the other fission products. Therefore, any venting i
- Uspace to at least take advantage o' tu>pression shM
- Aa of the se n 4
m a sbbing of non-c.le gas fission products.
To the staff crowledge, m 4te filtered vents such as have been or are being installe in Europe t
4 I
i /.. Sweden France and Germany) have not been considered for a BWR MARK 1*
in the U. S.
- A filtered vented contenment has been proposed at one U. S. MARK 11 plant.
n.
---nr,w
--,---mnn-e
~,_n,v,,
,--.---,-----,.,-,-,-----,-----,--..e-_.
, - - ~
i 4
'It is noted that venting procedures using existing equipment have been incor-porated in the emergency procedures for some U. S. reactors.
To the staff's knowledge, all MARK 1 plants have such procedures for utilizing various sized penetrations.
Most existing BWR MARK I wetwell vent paths outside primary containment are in-capable of operating in, or withstanding the pressu es and temperatures asso-ciated with, severe accidents.
If such vents were used without modification, their failure could result in contamination and hydrogen ignition in vital spaces outside containment. By connecting wetwell airspaces to the existing 1
Standby Gas Treatment System (SGTS) stack with vhlves and piping capable of j
withstanding severe accident tenperatures and pressure, and providing for remote manual operation *, fission products would to discharged without con-taminating vital areas, and would gain benefit from d spersion at a high (up to about 100 neters) level, Criteria for emergency venting through relatively small containment penetrations has been approved for. licensed BWRs as part of the implementation of post THI improvements. However, the smaller vents are generally not capable of sufficient pressure relief during severe accidents to ensure containment integrity.
Un-r cessary and untimely venting could put the public near a reactor at some risk. The procedures for venting, and the control of decisionmaking, have bee.n raiset as issues that require further assess.nent. Therefore, a systematic assessment of the negative safety impacts of containment vents will be made.
d)
Core debris control - Proposals by the staff have been made to provide for core debris control on the drywell floor of the centainment in the form of guide walls, and in the torus room under the steel suppression pool liner in the form of an additional water / debris catcher.
Preliminary issessments of containnent guide walls indicate they are unlikely to be effective in directing core debris away from downcomers or the containment wall. Curbs in the torus
. room in the reactor building would be expected to form a dam if core debris penetrated the steel suppression pool liner, and would retain suppression pool water. The water would help quench core debris and scrub fission products.
c)
Enhanced Reactor Building Fission Product Attenuation - If core debris fails or bypasses the steel suppression pool linar in the torus room, a direct path for fission products through ventilated spaces in the reactor builcMng would exist. Attenuation of fission products (see draft NUREG-1150) could be enhanced significantly by the use of sprays from the plant fire system. The enhancement could be accomplished by either a significant design change with little procedural impact, or a small design change with a significant procedural impact. The forocr has been studied and found to be relatively ed'stly. Whtt Ls not been fully considered is an improvement such as use of fire hose no:zles to provide a low flow rate "fog" in important airspaces. The nozzles would be clamped to hand ratis, and iniciated prior to major reactor building contaminetten.
f)
Basemat isolation - The possibility e.;:ists that the basemat may be penetrated by core debris, and contaminate water supplies.
Methods for isolating such core debris have been evaluated in the U. S. (Research Letter 150), and were useri nt Chernobyl. Because of the torus design of MARX Is, this type of event is considered relatively unlikely. Because of this, such an undertaking cculd be done on an ad h0c basis with only references provided in plant energency procedures.
One possible mear.s of powering such valves in station blackout events msy bc by the use of small portable DC generators. Such generators could also be used to power ADS valves and reduce accident likelihoods and consequences.
6
'g )
Automatic De)ressurization - The automatic depressurization system (ADS) may not be ava11a He during 550 scenarios. System availability has been gener-ally recognized as important in preventing core damage.
However, the use of the system to mitigate the high pressures, temperatures and fission products in the vessel after core melting, but befort vessel failure, may be useful. The advantages of improvements to the system for mitigation purposes have not been fully examined.
h)
Procedures and Training - Existing emergency procedures and training at
-BWRs with MARK I containments have not been fully developed With respect to risk significant severe accident challenges.
Improvements in existing proce-dures and operator training should substantially improve the capability of operators to cope with severe accidents.
(It is noted that procedures and training are also related to other severe accident programs such as IPEs and improved licensee performance.)
i CONTAINMENT PERFORMANCE RESOLUTION PROCESS - Resolution of issues is to be achieved by a two stage process. The first stage will consist of issue characterization, parametric studies, experiment assessments and a critical focusing on each of the relevant technical issues. Both phenomenological issues ar.d potential improvenrent issues potentially important to the ritigation of MARK I severe accidents are to be identified. Examples of phenomenological.
i issues are the manner in which a core disassembles in-vessel, how debris in the bottom of the vessel attacks the lower head and may induce failure, ejection of debris, and core debris attack on the containment liner. Examples of potential imsrovement issues include the usefulness of venting, containment sprays, ADS en1ancements, hydrogen control improver:ents, core debris control, reactor building fission product attenuation, and basemat isolatien. Parametric studies will include assessments of related experiments, and analytical evaluations of the impacts of a range of core melt progression assumptions and potential in-i provements on containment performance. After initial issue characterization, a meeting will be held with representatives from RES contractors, industry, other experts and interested members of the public on each issue.
The second stage will be a sorting ud evaluation process performed by the staff where each issue will be cati 9#l zed as being either a resolved or unimportant, b) poten-tially resolvable by future research, or c)) candidates for regulatory initiatives.
The criteria to be used for judging if a regulatory initiative will be recomended (to effect closure) include the bat.kfit ru'e (needed for safety, or a justifiable safety enhancement), and the Safety Goal Policy and implementation plan.
The process and related target dates are sumarized below:
MARK I CONTAINMENT PERFORMANCE RESOLUTION PROCESS r
1.
Prepare Program Plan a.
Prepare Commission Information Paper responding to Nov. 1987 i
SRMs.
Include early identification of challenges, failure codes, potential improvements and primary oencric issues related to the containment type being considered.
Identify details of the process for narrowing and resolving issues. Cecrdinate with NRR.
- See Enclosure 3 for relationship of activities and schedule dates.
j 6
b.
Seek ACRS coment on plan and closure criteria.
Dec., 1987 c.
Revise plan based on ACRS and Comission coment.
Dec., 1907 2.
Prepare for and hold a meeting with representatives from National I. abs. Industry, other experts and interested members of the public to narrow and, to the cxtent possible, resolve phenomenological and improvement issues.
Identify details of i
the criteria the staff will use to judge whether or not an initiative is warranted. Consider the use of "success states" defined in terms of the magnitude and timing of releases from outlier accident sequances based on a definition of a large release. Use -available resources (NUREGS, etc.) to forou atz initial issue characterization, and request meeting invitees to coment on characterizations prior to the meeting. Issues are l
to be related to containment challenges, failure modes and potential improvements. Use contractors to help with is' sue characterizations, to facilitate a meeting,(resolve issues, and to prepare sumaries. Use reviews of PRAs i.e., Peach Bottom, Cooper, etc.), utility containment safety studies (Vermont Yankee and Pilgrim), IDCOR evaluations, and the anticipated NUMARC evaluation to characterize issues. Undertake parametric core melt and containment challenge calculations, where practicable, and 1
review experiments to aid in focustrig issues.
i a.
formulate initial issue characterization Dec., 1987 b.
issue meeting invitation Jan., 1988 i
c.
revise issue characterizations based on Feb., 1906-coments, develop preliminary bases for staff evaluation of issues in terms of the magnitude and timing of fission product releases for outlier sequences, and hold meeting.
d.
issue draft sumary for coment Mar., 1988 a.
issue final sumary Apr., 1988 l
3 Prepare interim report to Comission with possible Apr., 1900 recospendations for improvements.
Identify primary areas of agreement and disagreement am6ng parties. For those issues for which analyses indicate that safety improvements could be effective (e. g. a form of venting), recomend a reDulatory initiative.
I unimportant, b)) issues that may be best resolved by future 4.
Identify a issues that are resolved or are
/yune,1988 analytical and experimental research, and c) candidete issues for regulatory initiatives using the criteria based on the Backfit Rule and Safety Goals. Complete Backfit/S&fety Goal assessment.
S.
Prepare Commission paper and/or NURGi alth staff re:eemendatior:.
l a,
complete draft Comission Paper June, 1988 b.
ACRS discussion July, 1988 c.
CRGR meeting July, 1988 l
d.
complete final Comission Paper Aug., 1983 l
e.
undertake implementation of Comission To be approved intiatives (e.
g., rulemaking, ion) determined Generic letters and licensee implementat
4 1
.6.
Continue long term MARK I research.
To be determined 7.
If warranted, assess other containment types.
To be determined RESOURCES
' Existing RES resources related to severe accidents and KARK 1 containments of approxtrately $1H and 4FTEs are considered adequate for FY 83.
These resources include those earmarked for resolution of MARK I issues, and those allocated for longer tenn BWR severe accident / source term research.
Additional resources during this period would not be expected to accelerate completion because of the time required for assessing, narrowing and resolving issues.
MANAGEMENT _ - The organizational unit responsible is the Severe Accident Issues Branch in RES.
Project management for MARK I resolution is to be provided by the branch chief.
Inputs from other branches in RES and NRR are to be solicited.
Turthermore, a senior level management steering group composed of representatives from RES, NRR and AE00 is to provide oversight.
t
)*
t e
~
~
.7
(
ENCLOSURE 3 MARK I CONTAINMENT PERFORMANCE KEY ACTIVITIES & MILESTONES COORDINATION WITH/NRR COMMENTS ON PAPER PRELIMINARY DRAFT,COMISSION PAPER / ACRS REVISE STATEMENT OF
\\
afROGRAM PLAN MTG.
PLAN ISSUES 11/6 11/30 12/30 EXPERIMENTAL, ANALYTICAL &ENGINEERINGASkSSMENTS
/
ISSUE MTG. ASSIMILATI DRAFT FINAL CMARACTERIZE INVITATION /ComENT
& MEETING
SUMMARY
ISSUES W/ ISSUE CHAJt. HOLD MT
SUMMARY
~
I
~
~
A 12/ 0 1/30 2i?8 3/30 4/3 i
ISSUE APVANCEO TICL W MEETING DRAFT ACRS FINAL C0fellSSION MTG. &
ComISSIOY 12/15 APER & FINAL REPORT Co MENT PAPER & REPORT RESEA CHER 5/15 7/30 8/30-MODE 1 ADEQUACY pEETING(BASIS)
V INTERIM CRGR p0 MISSION REPORT MEETINGS 4/IS 8/IS PRELIM. PARAMETRIC CALCS &
FINAL ASSESS ASSESS OF EXPERIMENTAL OF CALCS & EXPER-BASIS MENTS i
j (1)
~
e l
SUPPORTING RESEARCH 1.
EXPERIMENTAL a.
ACRR FACILITY CORE MELT DF-4 EXPERIMENT (SNL) - This DF-4 BWR early melt. progression experiment with control blade, canister wall and fuel pins has been completed.
It re) resents an initial data base for modeling BWR early melting in tse ORNL assessment of Peach Bottom.
The documentation of the exper.iment is to be completed by about April, 1988.
b.
B C/ INTERACTIONS (SNL) - One intermediate mix test has been completed 4
~
l and analyzed. Other tests are planned and documentation is to be completed.
Information from the one completed test is expected in December, 1987.
c.
EUTECTICS IMPACTS ON CORE MELT PROGRESSION (ORNi.) - Some small scale tests (a few kg.) are expected to be completed by January,1988 with-documentation to follow. The results are to be used in the ORNL assessment of Peach Bottom.
d.
SIMULANT MELT SPREADING & CONTACT (BNL) - Benchtop tests have been completed and applied in analyse of melt spreading, including the presence of an overlying water pool. Documentation is to be completed and the results used by ORNL.
e.
MOLTEN CORE CONCRETE INTERACTION TEST - Information on molten material L
spreading was obtained from an intermediate scale test (187 kg.).
This. test is used to support analyses of melt spreading within contain-ments by ORNL.
f.
HIGH TEMPERATURE HYDROGEN COMBUSTION - The combustion behavior of hydrogen, oxygen, steam, carbon monoxide and carbon dioxide mixtures in the reactor building is being investigated using recently developed l
models. These models cannot be experimentally vedfied at temperatures above 150 degrees C, but an initial peer review will be completed in I
November. The results are to be used by ORNL.
l 2.
ANALYTICAL a.
MARK I MELT SPREADING (ORNL) - This is a "first efforti parametric analysis using results from the experiments identified'above to develop code models for the eutectics formed by zirconium and uranium oxides.
The. liquid / solid temperatures of constituents affect predictions of concrete ablation, outgassing and aerosol containment emissiens. Melt LINER HELT (BNL) y and liner erosion rate predictions are also affected spreading velocit b.
reading, liner contact erosion and melting.
of time based on ccre melt sp(0RNL) This represents an initial parametric BWR CORE MELT PHENOME!!0 LOGY c.
medeling effort. The DF-4 tests, the THI-2 examination, and other severe fuel damace experiments are consistant with the BWR models to be used for FARK I :nal u s.
Further review of the models within the context of eeergh;g research on overall BWR core melt progression is to be conducted.
(2)
d.
PARAMETRIC ASSESSMENTS (ORNL) 1)
REACTOR BLDG FISSION PRODUCT ATTENUATION FOR PEACH BOTTOM &
BROWNS FERRY (ORNL) 2)
ADVANTAGES OF ADS IMPROVEMENTS (BNL/0RNL) 3)
FIRE WATER SPRAY ADVANTAGES (ORNL) 4)
ADVANTAGES OF CURBS & WATER IN TORUS R0d4 (ORNL)
S)
ADVANTAGES OF VENTING IMPROVEMENTS (INEL/0RNL) 6)
UNCERTAINTY ASSESSMENT (0RNL) 3.
ENGINEERING ASSESSMENTS a)
ISSUE CHARACTERIZATION (BNL/ STAFF) b)
USE OF CURBS IN THE DRY WELL & TORUS ROOM (ORNL/ STAFF) c)
USE OF FIRE WATER / SPRAYS AND SAFETY IMPACTS (ORNL/ STAFF) d)
ADS IMPROVEMENT (ORNL/BNL/ STAFF) e)
YENTING IMPROVEMENTS & SAFETY IMPACTS (INEL/ORNL/ STAFF) f)
H CONTROL IMPROVEMENTS (STAfr) 2 g)
IMPROVEMENT COSTS AND BENEFITS,(STAFF /INEL) h)
REGULATORY ISSUE EVALUATION (STAFF)
/*
i e
(3)