ML20209E134
| ML20209E134 | |
| Person / Time | |
|---|---|
| Issue date: | 09/03/1986 |
| From: | Bernero R Office of Nuclear Reactor Regulation |
| To: | Russell W Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20209E138 | List: |
| References | |
| FOIA-87-10 NUDOCS 8609100056 | |
| Download: ML20209E134 (11) | |
Text
_.
h#
,\\,.
UNITED STATES
/
f g
NUCLEAR REGULATORY COMMISSION g
- j WASHINGTON, D. C. 20555 k.....,/
September 3,1986 MEMORANDUM FOR:
William T. Russell, Director Division of Human Factors Te hnology i
Themis P. Speis, Director Division of Safety Review and Oversight i
Frank J. Miraglia, Director Division of PWR Licensing-B Thomas M. Novak, Acting Director Division of PWR Licensing-A Denwood F. Ross, Director Office of Nuclear Regulatory Research James G. Parlow, Director Division of Inspection Programs, IE Edward L. Jordan, Director Division of Emergency Preparedness
.and Engineering Response, IE FROM:
Robert M. Bernero, Director Division of BWR Licensing
SUBJECT:
GENERIC REQUIREMENTS FOR BWR CONTAINMENT
~.
RESPONSE TO SEVERE ACCIDENTS Enclosed for your comments is a Draft Generic Letter for Proposed BWR Contain-ment. Requirements which will enhance containment performance in severe accidents.
4 The requirements specified in tha Draft Generic Letter have been derived by
[
regulatory analysis comparing the behavior of BWR containments in severe i
accident environments, and the benefits derived from proposed containment enhancements. The regulatory analysis is also enclosed for your comments.
Our approach to containment enhancements will be discussed with the BWR Owners Group and IDCOR at meetings scheduled for September 11 and 23, 1986. The enhancements are also being discussed as part of the initiatives being under-taken by the Vennont Yankee and Pilgrim licensees. The material is scheduled to be presented to ACRS on October 9, 1986 and for ONRR review by November 12, 1986. A CRGR review is planned for November 19, 1986. The Generic Letter is scheduled to be issued for industry and public comments by the middle of December 1986, and as final by April, 1987. A chronology of actions related
(
to this BWR containment initiative is enclosed for your infonnation. Please provide your comments in time to support the schedu e.
/
)014-87~Ws obert M. Bernero, Director Division of BWR Licensing tyg,
Enclosures and cc:
See next page
{g
e e
2-
Enclosures:
1.
Draft Generic Letter on Proposed BWR Severe Accident Containment-Requirements 2.
Regulatory Analysis
- cc w/ enclosures:
H. Denton R. Vollmer W. Houston G. Lainas H. Thompson G. Holahan J. Funches W. Houston DBL PD's DBL BC's
~
CONTACT:
i
. M. C. Thadani x-28649 i
M e
l 1
f f
__z
7 CHRONOLOGY e
JUNE 16, 1986:
MEETING WITH BWROG/IDCOR PROPOSED A GENERIC LETTER, PRESCRIPTIVE SOLUTION, BY BACKFIT e
JUNE 30, 1986:
VERMONT YANKEE COMMITS TO GOV. K'UNIN TO DO A SPECIAL 60-DAY CONTAINMENT STUDY e
JULY 25, 1986:
BOSTON EDISON COMPANY BOARD DECIDES TO FIX PILGRIM CONTAINMENT e
AUGUST 19, 1986:
BWROG EXECUTIVES VOTE TO FUND AND CONTINUE DIALOGUE ON THIS WITH NRC, CONTACT NUMARC ABOUT BWR VS PWR e
SEPTEMBER 11, 1986:
MEETING WITH BWROG TO COMPARE BACKFIT NOTES AND STRAWMAN GENERIC REQUIREMENTS e
SEPTEMBER 11, 1986:
MEETING WITH VERMONT YANKEE TO REVIEW CONTAINMENT STUDY e
SEPTEMBER ~23, 1986:
NRC/IDCOR MEETING ON BWR/ MARK I ANALYSES e
SEPTEMBER 23, 1986:
ACRS SUBCOMMITTEE ON CONTAINMENT PERFORMANCE TO DISCUSS HARPERS FERRY WORKSHOP RESULTS AND BWR CONTAINMENT GENERIC APPROACH e
SEPTEMBER 24, 1986:
ACRS SUBCOMMITTEE ON CLASS 9 ACCIDENTS TO
-DISCUSS BWR/ MARK I ANALYSES AND SEVERE ACCIDENT PROGRAM l;
e NOVEMBER 19, 1986:
CRGR REVIEW 0F DRAFT GENERIC LETTER ON BWR CONTAINMENT REQUIREMENTS (T0 BE PUBLISHED FOR COMMENT) e DECEMBER 17, 1986:
ISSUE DRAFT GENERIC LETTER ON BWR CONTAINMENT REQUIREMENTS FOR PUBLIC COMMENT
~
e APRIL 1987:
ISSUE FINAL GENERIC LETTER-0N BWR. CONTAINMENT REQUIREMENTS
~
i e
i
- Y
'8 UNITED STATES
'[
g
= NUCLEAR REGULATORY COMMISSION
- j WASHINGTON, D. C. 20555
\\...../
i TO ALL BOILING WATER REACTOR (BWR) LICENSEES AND APPLICANTS FOR BOILING WATER REACTOR LICENSES Gentlemen:
SUBJECT:
. PROPOSED BWR SEVERE ACCIDENT CONTAINMENT REQUIREMENTS (GENERIC LETTER 86- )
J Severe accidents dominate the risk to the public associated with nuclear power plant accidents. A fundamental objective of the Commission's Severe Accident Policy is to take all reasonable steps to reduce the chances of occurrence of a severe accident and to mitigate the consequences of such an accident should one occur. The Reactor Safety Study report issued in 1975 found that for BWRs the probabilities of accidents resulting in core melt were low, but the containment performance following a severe accident was poor and tended to offset the benefits of low BWR core melt probabilities. Subsequent actions resulting from the TMI Action Plan have led to several plant modifications and required improvements in plant procedures to further reduce the likelihood 1
of severe accidents.
In December 1980, an industry initiative on severe accidents resulted in the formation of the Industry Degraded Core Rulemaking (IDCOR) group to address the concerns related to core damaging accidents. The IDCOR effort has led to industry methodology for Individual Plant Evaluations
-(IPEs) to search for the risk outliers and to address system reliability and containment performance on a plant specified basis. The staff has concluded, however, that for BWR corttainments, a set of generic requirements has been ii identified that moots the-need to await plant specific analyses of containmint performance and will lead to speedier implementation than would be possible via 4
the IPEs. Severe accident analyses have indicated several areas for improvement in BWR containments which should be promptly pursued as follows:
1.
Provisions should be made for symptomatic and reliable actuation of containment wetwell purge.and vent valves to open and close diverse purge.
And vent paths at set pressure conditions as'a means to ' assure that the -
containment pressure from beyond-design-basis events does not lead to i
overpressure failure and uncontrolled releases, and to provide a path for releases ~which will maximize the use of the suppression pool as a
[
condensing and filtering medium.
2.
Provisions should be made for reliable operation of drywell or wetwell l
containment sprays for a broad spectrum of accident sequences, including blackout sequences. The reliability of containment sprays should be i
enhanced by providing independent water and power sources. Backup water i
sources and pumps, hose connection and use of fire mains should be i
l considered. The provisions to be implemented should minimize occupational i
exposures that could result from manual actions and should be explicitly developed and expeditiously implemented as part of the BWR Owners Group i
development of the Emergency Procedures Guidelines. Such provisions l
should minimize the risk of implosion and hydrogen combustion, and assure that the equipment would be preserved in unflooded condition.
~ -.... -
. 3.
Combustible gas control provisions should provide substantial assurance that containment failure due to hydrogen combustion is not likely in the more likely severe accident sequences, including blackout sequences. The periods of operations while Mark I and II centainments are deinerted while at power, particularly during potential preshutdown conditions, should be minimized by reducing the present Technical Specifications permitted value of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.
4.
Paths for core debris travel should be evaluated for conditions representa-tive of a large scale core melt. Where the expected path of debris travel indicates a substantial likelihood of loss of the suppression pool as a release filtering or debris quenching medium in Mark I containments, the torus room under the suppression pool should contain a 3 foot high barrier to trap water and core debris.
5.
Emergency procedures and training should be reviewed and modified as necessary to ensure that operators are able to recognize severe accident conditions and use plant equipment to best advantage under such conditions. Revision 4 of the BWR Emergency Procedures guidelines should be implemented promptly following the staff's review and approval.
In addition, improved automatic depressurization system operation for station blackout response (battery backs), and drywell debris barriers in front of suppression pool vents for Mark I containments are highly desirable. The staff believes that such barriers are justified if the cost does not greatly exceed about one million dollars.
This proposed generic letter has identified areas for improvement in BWR containments which the staff believes to be effective in reducing risk and which can be implemented at a reasonable cost. We welcome comments on the proposed actions and other suggestions on the subject matter...The goal is to significantly reduce the likelihood of containment failure given a core melt.
Sincerely, Robert M. Bernero, Director Division of BWR Licensing Office of Nuclear Reactor Regulation
BWR CONTAINMENT PERFORMANCE j
DURING SEVERE ACCIDENTS
1.0 BACKGROUND
In its Severe Accident Policy Statement of August 8, 1985, the Commission indicated its objective is to take all reasonable steps to reduce the chances of occurrence of a severe accident and to mitigate the consequences of such an accident should one accur. Further, the Commission stated that the examination of individual reactor risks will " include specific attention to containment performance". The examination of severe accident risks at BWR plants with Mark I, II and III containments has lead the staff to conclude that a number of worthwhile and significant improvements in containment performance (in both design and operation) can be made at low to moderate cost.
Containments are required to protect the public from the consequences of accidents. The design and sizing of containment systems are based on the pressure and temperature conditions which result from release of the reactor coolant in the event of a design basis loss-of-coolant accident (LOCA). The containment design basis includes the effects of stored energy in the reactor coolant system, decay energy, and energy from metal-water reactions including the recombination of hydrogen and oxygen. The containment system is not required to be a complete and independent safeguard against a design basis LOCA by itself, but functions to contain any fission products released while the emergency core cooling system cools the reactor core.
Although a postulated design basis LOCA is not expected to produce more than a few percent fuel failures,' an accident radiological " source term" used in calculating offsite dose consequences is representative of a substantial core melt accident (10 CFR 100). Even for this source term, containments are designed such that calculated offsite doses are unlikely to result in an early or major latent health hazard if the containment were to maintain its low leakage capability *. What is at issue is the capability of containments to perfonn a mitigating safety function as long as practicable during very low probability, beyond design severe accidents when the consequences of containment failure may be very significant because of the physical forces imposed on them by core melt. This significance is important for the primary BWR containment types where, in spite of the positive pressure suppression feature, they are either relatively small (Mark I & II), or have very low designpressure(MarkIII),andtheirlikelihoodoffailureinasevere accident is perceived to be higher than it need be.
- Part 100 specifies "the expected demonstrable leak rate from the containment", a value which is made part of each licensee's Technical Specifications.
'q 2-e i
Overall plant core melt probabilities for Mark I, II and III BWRs have been i
estimated to range from one in a thousand per reactor year to less than six in a million per reactor year for eight BWR designs. Many of these estimates have not fully assessed the benefits of post-TMI backfits or operator responses, nor i
the increases in core melt probabilities arising from factors not considered in plant specific analyses such as earthquakes, floods and fires. Contemporary analyses break down such probabilities into classes and subclasses of accidents.
The sum of the core melt probabilities for all classes and subclasses of accidents is considered to be the overall core melt probability.
IDCOR has i
proposed five classes of events for BWR core melt accidents, depending on the I
initiating event and containment response, that are useful references as follows:
l loss of core cooling with containment at low pressure and failure after i
core melt; loss of core cooling with containment failure before core melt; loss of core cooling with containment failure soon after core melt due to i
high containment pressure at the time of core melt; loss of core cooling with containment failure before core melt due to failure to depressurize; and containment bypass.
i Our review of the core melt probability estimates to date generally indicates i
they are low. The BWR cor'e melt frequencies of past evaluations are sumarized in Table I.
Given a core melt, the estimates of likelihood of Mark I, II i
and III containment failures have been high relative to other containment types *.
In all of these past evaluations, little or no credit has been given i
to features which can be used with relatively modest upgrading to prevent or i
mitigate accidents.
(
- The Reactor Safety Study (WASH-1400, NUREG-75/14) indicates a conditional containment failure probability for the BWR Mark I containment reference plant That is, given a core melt (Peach Bottom) of about 90% (Table 5-3, page 81)).
in a BWR with a Mark I containment (Peach Bottom there is about a 90% chance I
of containment failure. In the November, 1984, IDCOR Technical Summary Report, Nuclear Power Plant Response to Severe Accidents, the estimate for Peach Bottom was about 20% (Table 10-1, page 10-6). For a BWR Mark II containment (Limerick), Brookhaven National Laboratory (BNL) estimated almost a 100%
i likelihood of containment failure given a core melt (BNL 33835; April,1984).
The staff evaluation of the GESSAR II standard plant design (NUREG-0979, Supplement 4 Tables 15.1, 15.2, and 15.12) also, indicates a conditional containment failure probability of close to one for a Mark III type containment.
In only the IDCOR and GESSAR II evaluations was containment i
venting considered. A 50% chance of Mark I containment failure, leading to a l
"large" release,following a severe accident has been used in this analysis.
i
,r
--e-www,-
-rwg 7g-.
.r.-,
,my.,we.,wgy.
.-w,,,-ww.,
..u,9cw9ww-y,,,we-mpq w,q y--w,-a,
-,.*-ww,gg-+-mp.,.9 q.---g-.,.y-
.i
...I a
=
4 l
TABLE-I DOMINANT BWR CORE-MELT ACCIDENT SEQUENCES IN PRAs Core-Melt Frequency Estimated in
- Median, Warranted PRA (Plant Mean or
{*
Authorization / Containment Product Thennal Power modifications Events Point PRA Source Configuration Line (MWT) not included)
Considered Estimate
}
Internal /
Big Rock Point Consumer Dry BWR 1 158 1 x 10-3/RY External l
Mean 4-Events Power Company (fire and risk j
j analyses included) j' Millstone 1 NRC-IREP*
Mk. 1 BWR 3 1727 3 x 10~4/RY Internal Pt. Est.
Events
[ ['
1 l
Shoreham Long Island Mk. 2 BWR 4 2436 5 x 10-5/RY Internal /
Pt. Est.
External i
Lighting Co.
Events I
i Internal /
Peach Bottom 2 NRC-WASH-1400* Mk. 1 BWR 4 1600
_3.x_10-5/RY.... External Median _ _,_,_ '
i Events (fire and risk analyses included)
Browns Ferry 1 NRC-IREP Mk. 1 BWR 4 3293 2 x 10~4/RY Internal Pt. Est.
Events Internal /
)
Limerick 1 Philadelphia Mk. 2 BWR 4 3293 1 x 10-5/RY
. External Mean
?
Events Electric Co.
(fire and risk analyses included)
Grand Gulf 1 NRC-RSSMAP*
Mk. 3 BWR 6 3833 4 x 10-5/RY Internal Pt. Est.
Events i
Internal /
4 x 10~0/RY External Mean l
GESSAR GE Mk. 3 BWR 6 3579 Events I
s p a
--.a 4 i.
I
For most' accidents considered, the core is postulated to melt, interact with steam, water, and the structural features in the vessel and coolant system, melt through the vessel, and attack the concrete and structural features of the lower containment. Depending on the sequence of events, the containment has the potential to fail either before or after vessel melt through.
For the remainder of the accidents postulated, the containment would be bypassed, allowing radioactivity a direct path to )ortions of a plant not designed to contain the releases, but with some capa)ility to attenuate radioactivity.
BWR containments respond to heatup of the fuel in the vessel directly or indirectly. The direct transfer of energy is through pipe breaks, or through blowdown into the suppression pool.
Indirectly, radiant heat is transferred through the vessel and piping. The blowdown or depressurization process, and the use of the relatively large quantity of suppression pool water as a heat sink and fission product scrubbing device, act in combination with the structural capability of the containment (including penetrations) to mitigate the high temperatures, pressures and radioactivity released in a core melt.
Core melt scenarios have been identified, however, which can produce conditions in the containment that could lead to failure. There is strong evidence that BWR containments are capable of withstanding substantially higher stresses than those for which they have been explicitly designed and this evidence can be capitalized on to provide additional protection to the public at low to modest cost. The longer a containment can be expected to hold, the greater the likelihood that failure can be avoided.
If failure were to occur, however, reductions in the radioactivity released can be achieved. Actions that can be taken to prevent a catastrophic failure of containment before the fission products are adequately attenuated include such items as operator actions to vent the wetwell space above the suppression pool, and providing reliable spray capability.
In a core melt accident with temperatures in excess of 5000 degrees F, fission products are released from the fuel in three general groups. The noble gases and the more volatile species of fission products are released from the fuel relatively early in a core melt upon the occurrence of fuel melting in the vessel. Later, the less volatile species are released as the fuel melts down into the vessel and combine with the in-vessel structural materials. Finally, after melting through the vessel, refractory materials may be released during interactions of core debris with concrete on the floor of the containment.
The amount of radioactivity that could be released to the environment in core melt or degraded core accidents has been the subject of considerable analysis for a number of years. Present estimates (NUREG-0956) for MK I and III BWRs indicate that substantial quantities of important fission products can be released in a core melt accident and these analyses provide clues that suggest that releases can be reduced by a number of actions to enhance containment l,
performance.
Within the core of a contemporary BWR with MK I, II, or III containment at full power there are over five billion curies of radioactivity. Severe accident releases to the environment for a MK I, II or III have been estimated to exceed i
40% of such important fission products as iodine and cesium (releases of over 300 million i
.,~..-_,-.,.-_--y-
,om._.,__
~
curies of iodine and over seven million curies of cesium for a 3458 MWt reactor). The release of fission products can be reduced by a number of-containment enhancements evaluated in Section 2.
2.0 MITIGATION FEATURES AND BENEFITS The vulnerability of BWR containments to failure from severe accidents and the s
source term attenuation capability of the containments can be further enhanced by a few cost effective modifications to the containments and support systems.
Although some of the proposed changes could provide additional core cooling capability and, thus, further reduce the low probability of core melt, the emphasis here is on mitigation of a severe accident should one occur. The goal is to obtain substantial assurance that, given a core melt, the likelihood that a BWR containment will be breached is substantially reduced, and the release of fission products to the environment minimized. Thefollowingmogifications have been evaluated using an estimated BWR measure of risk of 10 person-rem per reactor year *.
1.
Assure wetwell venting. The postulated severe accidents would vary in their venting requirements depending upon the phenomenon and the energies involved in specific sequences. Therefore all available venting paths should be considered and used symptomatically by matching the venting requirements with the venting capabilities of available pathways. For sequences which require venting capabilities beyond those of the existing pathways new piping, venting outside the reactor building, should be considered. Although containment venting has been recognized by BWR owners in the Emergency Procedure Guidelines as an effective pressure control measure, modifications may be required for ductwork, isolation interlocks, and valve accessibility during severe accidents. Remote manual valves should be provided to relieve pressures (and to subsequently reclose) in the wetwell space of all BWR containments at or near design pressures during severe accidents. Procedures to instruct the operators on the use of such valves should be included in the BWROG Emergency Procedures Guidelines, reviewed by the staff and implemented as soon.as practicable. Where confinement of releases downstream of such values cannot be assured, reactor building vents should
- 10 person-rem per reactor year is based on NUREG-0974 (Table 5.11h), the Limerick Final Environmental Statement. Because not all of the potential accident initiators were considered, and because of the uncertainities 4
involved in the Limerick risk estimates, we used 10 person-rem per reactor year and a cost factor of $1000 per person-rem per reactor year. The-probabilityofasevereagcidentleadingtoalargereleaseatLimerickwas estimated to be about 10- per reactor year (TaQE 5.11d of NUREG-0974).
If a conservative BWR core melt frequency of 10 per reactor year is used with 50% chance of containment failure, then the probability of large release is calcualted as 5x10-5 per reactor year.
.--r,--
-m
.--y._
4 b-
-5 be provided. To the extent practicable, existing valves and ductwork r
3 should be used. Where new facilities are required, seismic qualification would not appear to be necessary. Thg costs of such backfitting are expected to be no more than about $10 per reactor.
2.
Make drywell and wetwell containment sprays very reliable and provide sources of power and water independent of normal sources. This can be accomplished at most BWR plants by the addition of interties between the RHR system, other systems and the normal plant fire protection system.
These fire pumps are typically powered by dedicated diesels.and/or fire trucks, and have a variety of possible water sources available. Connec-tions for fire hoses on the exterior of containments and use of pumper trucks might also be useful. Modification of existing drywell spray capabilities to ensure adequate heat dissipation and fission product attenuation at flow rates capable of being delivered by fire pumps should be undertaken for all BWR containments. Reducing the dry well spray flow rate substantially (about 10% of present value) would also preserve the equipment in the unflooded geometry. Instructions for the use of such 4
equipmentshouldbeaccomplishedthroughtheBWROGEmergencyProcegure Guidelines. The costs of this backfit are expected to average $10 per reactor.
3.
Combustible-gas control differs between Mark I & II plants, which are inerted during operations and Mark III plants. Typically, Mark I & II plants are allowed by Technical Specifications to be deinerted 24-hours after startup and 24-hours prior to shutdown. Minimization of this deinerted time would reduce the vulnerability to hydrogen combustion, and for the 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> prior to shutdown, would tend to minimize risks.
Since deinerting typically takes 4-8 hours, and leakage inspection and minor repair takes 4-8 hours, a 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> period may be sufficient.
Suitable procedures to implement such provisions should be incorporated into Revision 4 of the BWROG Emergency Procedure Guidelines, reviewed by the staff, and implemented by January 1, 1988. Costs of such backfits is 5
not expected to exceed $10 per reactor.
4.
Consider the effects of debris travel in the presence of containment t
venting and spray enchancements in 1 and 2 above.
For Mark I containments, i
analyze and assure that torus water is retained in the torus room to
. quench debris in the event of debris attack on the torus.
If necessary consider partial blockage of the entrances to Mark I torus rooms to a height of 2-3 feet to highten the quenching poog. The costs of such partial blockages should be on the order of $10 per Mark I.
Alternatively, where feasible within the cost range of $10 per Mark I, consider debris l
barriers about 2-3 feet high in the drywell space below the reactor vessel i
to limit the possibility of core debris from entering the suppression pool l
through vent pipes.
^ 5.
Improve procedures and training.
Implementation of the most recent symptomatic Emergency Procedure Guidelines would remove uncertainty as to i
how to cope with severe accidents, and would greatly improve the probability of succesful mitigation. The procedures should be developed p.3
w
)
6-and implemented using the BWROG Emergency Procedures Guidelines, incorporating the NRC comments. Thecostsofguchprocedurechanges r? f]
and training are expected to be about $ 3 x 10 per reactor.
Each of the recommended backfits is expected to reduce the consequences and, therefore, the risks of severe accidents by more than 1000 person-rem per reactor year.- Such reductions are more than sufficient to justify the
, proposed backfitting.
O,
.:3 1 Y
t c'
t 48
- 4 lt r
e ane I
i
__