ML20235E018
| ML20235E018 | |
| Person / Time | |
|---|---|
| Issue date: | 09/19/1986 |
| From: | Morris B NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Speis T Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20235D912 | List: |
| References | |
| FOIA-87-10 NUDOCS 8707100385 | |
| Download: ML20235E018 (6) | |
Text
r
,, - ' *)
h(Id162%g g '
\ / s&y SEP 19 1995 MEMORANDUM FOR: Themis P. Speis. Director Division of Safety Review and Oversight, NRR .
FROM: Bill P. Morris,. Acting Director Division of Reactor System Safety, RES
SUBJECT:
GUIDELINES AND CRITERIA FOR MARK-1 FWRs We have reviewed the proposed guidelines and criteria for MARK-1 EWP cs requested by your memorandum of August 15. The document is a good first attempt at developing regulatory requirements that are directly connected to the major issues associated with severe accidents. We present below several comments on the policy implications and technical content of the report which we believe should help to improve the document.
- 1. It is not clear when the criterie established in the document would be required on operating plants, nor how these reovirements interface with the existing backfit rule. This relationship should be clearly established in a policy statement to accompany the contractor document and in the Guidelines presented in Chapter 3.
This should include a discussion of how the goals for the severe accident program (as stated in Section 2.2) rela.te to the Commission's Safety Goals.
- 2. The screening criteria in Chapter 3 imply that the criteria are to be applied to any sequence which represents either (a) 5 percent o' the total core snelt frequency or greater, or (bg are estirrated to have a frequency of occurrence greater than 10~ per reettor year.
In addition, the criteria would be applied to any sequence involving suppression pool bypass he'ving an estimated frecuency j
greater than 10 per reactor year. If this is the intent, it dres !
not appear to be in concert with the Safety Goel Policy Staternent ;
singeamajorearlyreleaseetanedimatedfrequencyofabout 1 10~ per reactor year would be rouchly comparable to the ear'y fatality safety goal.
?. We believe the results of the Severe Accident Risk Rec'uction Procrty (SARRP) have been misrepresented by focusing or. the central es tibatt results from the containment event trees. These do not repr(tent a best estimate end they will not be presented in NUREG-1150. Petter, they tend to be close to the lower bound of the uncertairity rence ir many instances. We suggest information f rorn the SARPP uncertairi -
analyses shewins. the plausible range in which the meer,salut tu should be used ir rince of centrr.1 estimates, insights fror ve uncer u inty erelyses sometires differ sigr.ificar.tly frnr tbnr gained if the central estirrate is considered a "best estimaic.
~
8707100385 870701 PDR FOIA TYE87-10 PDR N3 /
/
.0- .
Themis P. Speis 2 gip U t
j
- 4. We request that references to early drafts of the ASEP report (p. 1-5 of the draft Guidelines) and to the draft position peper on direct containment heating (p. 19-20) be removed. Draft material which has not been finalized and is expected to be modified is not a valid reference.
- 5. The intertance of support systerrs can be better shown by examples such as AC dependencies of turbine-driven AFWS trains, or the component cooling water system importance at Indian Point.
- 6. The NRC does not yet have an agreed on policy on coherent core slump as it applies to direct containment heating, use of preliminary staff efforts should awatt joint concurrence by the Offices involved.
- 7. The criteria presented with regard to Reector Pressure Vessel Injection (Guideline 4.A) are intended to complement and supplerrert the proposed Station Blackout Rule. If,the proposed rule is inadequate, it should be modified prior to issuance, rather than issuing new requirements at a later time. This would permit better utilization of resources by licensees.
B. The criteria presented in Chapter 4 are, in general, well founded.
However, if they are to be used for regulatory purposes, they tieed to be carefully edited to eliminate ambiguities and to clarify requirements more precisely. Examples of the types of modifications needed are included in our rnere detailed coments which are appended.
- 9. While we agree that containment ven' ting is desirible under certain circumstances and can prevent catastrophic releases by ensuring the availability of the suppression pool scrubbing in certain sequences, we caution that the detailed engineet;ing will be difficult and note that the discussion of containment venting criteria in Chapter 4 doet not appear to be consistent with,the goals presented in Chapter
- 3. In sequences where core melt prec6 des containment failure, wetwell venting will not occur until.after the pressure vessel hes failed (assuming venting at 75 percent of design pressure as suggested by criterion 1.B.1.1.a). Venting at this time will rektse a significant amount of noble geses to the environment.
This could approach the entire noble gas invertory, depending on the duration of venting. Our calculations indicate that a large neble ges release alone could cause life-threateners doses in artes neer the site bcundary under certain weather conddtions un'ess residents were evacueted. It is difficult tc ccr. sider the i centairrent function as having been maintain (r (p. 3-1) Uncer these circumstances. Certainly, such a releese wculd exceed the Part 100 crittria proposed in the recert memorantum to the Cor.rission rer r ding application of revised source terrs. It ray be advisabb to iru stigate venting earlier, before t, reel c' the reactor vessc'.
tc c(tert..ine if such a practice coult' c'elay m tainet nt f ailure.
-- __- a
.Themis P. Speis 3
- 10. We suggest the criteria be examined carefully to identify potential areas where the modifications imposed by the criteria might lead to adverse effects in other areas. For example requiring that the ADS be " capable of being defeated reliably" obviously can impact the reliability of its functioning during small LOCAs. . Similarly, the a ATWS procedures could impair the ability of the operator to respond-to morc frequent transients and LOCAs. -
- 11. Appendix A provides much useful information.. ilowever, it is written to an audience of PRA practitioners, and will not be easily i
understood by the average engineer or scientist. We suggest it be I expanded to explain the PRA " jargon" in conventional terms,
.providing more infn,rmation on the engineering and plant-design reasons for differences in results.
- 12. The guidance for Criterion 1.A suggests using curbing to block pedestal doorways. We note this would , require curbing at least 4 j feet high and would subject the pedestal walls to attack by the I corium. Also, drywell sprays cannet contact the molten material if it is confined in the pedestal regior,.
1 Bill M. Merris, Acting Director Division of Reactor System Safety Office of , Nuclear Regulatory Research l
DIST: .
Subj.
Circ -
Chron ,
Morris Murphy Ross Beach Director's files bec: R. Bernero g
M f
Wdpv gc: , p, t*Y
- i.
f 9/)/S6 I i
l l
Detailed Coments on Guidelines and Criteria for P.ARV-1 BWPs
- 1. Examples of items which need either more information ard or improved editing in the criteria intiudt the following: -
(a) delete " primary system and substitute
when discussing BWRs.
(b) "Ultimete desig,n pressure" is not a conventional term; substitute either design pressure" or " ultimate capacity" depending on intent.
(c) Define " active filtering' (1.F..l.6)
(d) Provide the basis for requiring vent functionability after opening for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> (1.B.I.7).
(e) Define "re'iable" (1.B.I.10, 4.A.5). We note that well-engineered single train systems such as HPCI and RCIC will usually have an unavailability greater thar 10 '/ demand.
(4) Provide the basis for selecting 1 percent of decay heat for establishing drywell spray flor rates and 0.5 percent decay heat for alternate ECCS flow rate. ,
(g) Define periodically (2.A.1)
(h) Provide -he basis for a PC hr. wa,ter supply for ECC injection.
- 2. In generel. we recomend probabilities and frequencies be presented to one sdtrificant figure.
- 3. Thc rany re'erences to "St.FP" should be changed to "SARRP" to avoid confusier w ith the Severe Acci' dent EeSearch Program and properly reference the Severe Accident Rist Peduction Program at Sandia Natic r C Liboran ries .
4 Criterier ..!.1.1 stec'd reauin ut*dne X hours after loss of all AC. pcu - tr'.ess AC po,;er is et: rtc tcJt ie inverters. It is diff4.' . r ru d'c1 -- ect u ' : Fc+ ire wier batteries vill deplete sirc< -t cre ret u r. ' ' o r- > u tiu .
F. Criter ac'. I . A." r4 cc't. b< tter c.ehr.itior, cf the f ailure modes to cci.5 4ce r tir 4r g relief sci'.es. A prcbabilistic cutoff may be we<-
- c. , ,... ; r.- r -, . .,
. u s irertsd during power c;.c c i r : t. 15ct. qn
, (;tir.- g e rr it sone power ascent eu ;r, d w ebt operr. V - p '+ ra - i r.e r t e c . This limi
- i. . . lit., of Contairrett
- t .t , l '. I , tc 2 .cxirur. of 10~js or the b r.i s ( . .
1 c.. : + .
t -
f s
7.- The basis for the suggestion for placement of ceramic bricks to prevent drywell attack (p. 3-7) should be referenced.
- 8. The statement that the suppression pool is the most important system
.might be overstated. Accident prevention systems and the containm(rt are equally important. -
- 9. The RHR unavailability presen'ted in the Browns Ferty IREP was i' calculated, not assumed.
- 10. The probability of steam explosions on p. A-21 should be unitiess.
- 11. The comments on drhell spray on pp. A-21--A-21 rrPy be overstated.
l The spray flowrate required by the criteria (200-250 gpm) is low compared to PWR spray systems. Its ability tc remove sensible heat to quench the raelt_is limited, as is its fission product removal effectiveness, in our analysis of Surry, we did not credit the sprays in preventing the direct containment heating pressure spike because of the short time involved.
- 12. The factor of 10 claimed'as minimut for suppression pool DF is not universally true. We calculate of DF of 2 for cerditions existing l.
during core-concrete interactions for certain secuences. ,
1
- 13. The comparison of the RSS, IDCOR and ASEP for TW secuences should note that the Emergency Procedure Guidelines did not exist at the time of the Reactor Safety Study and thus venting was properly not I considered as a option.
14 The discussion of the timing of releases to the reactor building on
- p. A-25 should recognize that the rate of releese can have a significant effect on reactor building DF.
l
- 15. The discussion of revaporization on p. A-27 sFcuid acknowledge the J consideration of revaporization in thb SARFP u* certainty analysis and delete reference to the CEI central waiktt-ough.
- 16. Sigr.ificat ; irsights on BWR perforraerce in se.Ue accicents beve been obtained by the SASA program. These studies should be r.oted. )
In particuler SASA has found that flow contrc' rey be easier to implement than level control in ATWS secuences.
- 18. In implementing the criteria regarding drywell sprays, ccrsideration should be given to rapid condensation of steem leadirp +o vacuurr corditions arc', if a recirci'etine sprey systen is ervisior4 c te the deleterious effects of c'"culatino fission-ru act lader u ter threert the RHE syster to the rr ector j buildinc. i i
l 1
1 l I.
- 19, SASA analyses indicate the debris bed is relatively cool at the time of vessel breach in a BWR. Thus, the potential for direct containment heating may be significantly reduced.
t l
8 e'
1
- - - - _ _ _ _ - _ - _ _ _