ML20207T271

Notifies That Hearing on Emergency Planning Issues for Plant Postponed by Will Be Held on 861118 in State of Ma.Ofc Will Forward Testimony & Answers to Subcommittee After Review by EDO & Staff Legal Advisors

ML20207T271
Person / Time
Site:	Seabrook
Issue date:	11/04/1986
From:	Kammerer C NRC OFFICE OF CONGRESSIONAL AFFAIRS (OCA)
To:	Asselstine, Roberts, Zech NRC COMMISSION (OCM)
Shared Package
ML19306D588	List: ML20197D222 ML20205L245 ML20207A140 ML20207F465 ML20207Q051 ML20207Q052 ML20207S785 ML20207T164 ML20207T271 ML20207T335 ML20207T387 ML20207T396 ML20207T407 ML20207T416 NUREG/CR-4540, Forwards BNL 861117 Request to Amend Page 60 of NUREG/CR-4540, Review of Seabrook Psa:Containment Modes & Radiological Source Terms, to Clarify Wording Re Conditional Failure Probability
References
NUDOCS 8703230378
Download: ML20207T271 (9)

Text

.. .. ...

/  %,,

UNITED STATES 8'

o NUCLEAR REGULATORY COMMISSION

{ msHwarou, o. c. 2oses

• +++'

/

November 4, 1986 MEMORANDUM FOR: Chairman Zech Commissioner Roberts Comissioner Asselstine Commissioner Bernthal Commissioner Carr FROM: Carlton Xamerer, Direc+

Office of Congressio Aff its

SUBJECT:

MARKEY FIEl.D HEARIN ON SE BR00X EMERGENCY PLANNING By letter dated November 3, 1986, Congressman Markey notified the Comission that the previously postponed hearing on emergency planning issues for the Seabrook Nuclear Power Station will be held in Massachusetts on November 18, 1986. This letter also forwarded additional pre-hearing questions.

The General Counsel has advised that, since matters being adjudicated in Seabrook may have to be reviewed by the Comissioners acting in an adjudicatory capacity, the testimony and answers to questions for this hearing should not be reviewed by the Comissioners, their offices, or adjudicatory advisors in the Office of the General Counsel. OCA plans to forward the testimony and answers to the Subcomittee after review by the EDO and the staff's legal advisors. Because of the adjudicatory

- considerations noted, the testimony and answers will not be provided to the Comissioners or their staffs for review.

c: SECY l OGC EDO 8703230378 870313 PDR COMMS NRCC l CORRESPONDENCE PDR i

..,,. msg,emm-n- o -ee - .- wm.-. e===.e. we -

-w ,e - -

- .-.~._ - ~.. _ .- _. . - _ _ _ _ _ . - . _ . .. . _._.m _ . . .

.. .. y S. LONG'S COMMENTS ON BNL DRAFT REPORT ON

^SEABROOK EMERGENCY PLANNING SENSITIVITY STUDY Deficiencies:

1. Section 2.4 on containment isolation failure is missing. . The NRC staff has decided to provide a memorandum addressing the probability of containment isolation failure. BNL should use the_ memorandum as a referenced appendix and provide an assessment of the associated risk under section 2.4, in section 7, and in the summary.

2. The section on preexisting containment leakage is missing. (This was originally designated section 2.3.) The NRC staff has decide to provide a memorandum addressing the probability of various rates of containment leakage being in existence prior to the beginning of a core melt accident. BNL shoul use the memorandum as a referenced appendin to provide an assessment of the associated risk as a new sub section in chapter 4, in section 7, and in the summary. ,

3. The PSNH response to RAI 025 (concerning the reasons for not modeling the remaining ISL pathways) should be incorporated into Section 2.1.2.

4. Section 2.1.4 addresses the probabilities of recovering from the Event V sequences by opera' tor actions.. Although the section addresses the reliability of the operators, it only mentions the associated hardware reliabilities. IN

.particular, the ability of the LPSI injection side MOVs to close against reactor system pressure should be addressed. If these valves cannot close against full system pressure, then the potential for premature closure attempts to disable these valves should be addressed and the associated risks estima,ted.

. 5. The degrading effects of of age are mentioned as important considerations in determining the strength of the RHR Heat Exchanger tubes, but the PSNH response to RAI 075 were not used in section 2.1.5.

6. Page 3-15 discusses the need for considering the cracks that develop in the containment under large defgrmation conditions when considering the concrete shear cones that are involved in resisting punch-through of small penetrations. The report states that SMA's response to a question on this issue is still being evaluated by BNL. THe results of BNL's analysis should be provided.

7. The totalled curves for thyroid dose probability versus distance curves were not provided.

8. The whole body dose probability versus distance curves should be provided for dose levels of 1,5 and 50 rems, as well as the 200 rem curve. (Figures S.7 and 6.12 or addtional figures, as necessary for clarity.)

Comments:

1. The preface misrepresents the nature of the normal NRC regulatory review

.  ;._ a . . - ... - . .

T' .

t o

-process. It is not NRC practice to " independently reassess, requantify and validate" calculations performed by the applicant. Rather, the NRC's regulations specify standards for quality control and assurance to which the applicant and its contractors must conform in performing its calculations.

NRC staff review performs an audit type check of the calculations an'd sometimes the applicant's adherence to the QA and OC regulations. In this

-case, the BNL effort is a part of the audit process, and it has checked the applicant's analyses in the areas that are considered to be most critical to the results. The preface should not refer to this as " unfortunate". The

. preface should be rewritten "to focus on what the ENL report is intended to accomplish, and refrain from language that appears to establish new goals for NRC reviews. -

2. Tne Summary is not as clear and readable as it should be, and it is not as complete as it should be.

3. The curves in figure S.4 marked " Impact of ..." are actually depictions of additional risks from two sources that were not included in the curve marked "PLG-0465". Therefore, the labels should indicate that these curves are additional to the "PLG-0465 curve. As presently labeled, it can be (mis)interpretted that BNL's calculations showed reduced risk at distances inside 2 miles.

4. The shaded area in figure S.7 should not extend below the curve marked "PLG-0465". In fact, the shaded area should stop perceptibly above the "PLS-0465" curve in the' probability range around 0.01. The same comment.

applies to figure 6.12.

5. The introduction needs to be expanded to provide more information about the various documents that update the SSPSA. It should state the specific conclusions that were presented in PLG-0465, the bases for these conclusions as put forth by PSNH, and the rational for the focus of the BNL review agenda.

6. The ciscussion of the contents of Section 6 that is presented on page 1-4 should state that Section 6 presents the overall effects of the results of the BNL investigations on the risk parameters.

7. Section 2 is very difficult to follow. It shows the effects of too many different authors working without a unifying editerial presence. There are many places where it is not readily apparent whether a conclusionary statement is providing the PSNH position, the BNL position, or a PSNH position with ,

which BNL concurs. A clear and concise introduction is needed to explain what topics are investigated and why. Following sections should be introduced with a reference back to the Section 2 introduction so that the reader begins the section wiht the proper context.

8. The first two paragraphs of the Secticn 2 introduction seem partially redundant and should be combined. The third paragraph has several faults: it mixes the " completeness" issue and the organi:ation of the the section in a confusing manner; it substitutes emotionally connotaive wording for a concise statement of the " completeness" issuel and it presents an unrealistic picture of PRA analysis. The organi:ation of the chapter should be the dominant theme of its introduction. The rational for including the various subjects in the BNL review derives mainly from the i criori knowledge of ,the sensitivity of 1

1

4 the conclusions to these factorsi these sensitivities should be mentioned in othe introduction section and_more fully explained in their respective subsections. The " completeness" issue, which always arises in PRAs, should be explained dispassionately. The report should note that a systematic completeness review was not conducted, although many NRC and BNL staff members did provide knowledgeable judgements concerning the inclusion of significant issues, as reflected by the presence of several BNL review topics that were not incl.uded in the PSNH documents.

9. In section 2.1.3.1, the relationship between the thoughts is not clear in the second and third sentences of item e on page 2-5.

10. In the next to the last paragraph on page 2-6, it is not clear if the BNL

~

tables contain more events than the PLG tables or if htey contain events that are more relevant, or both.

11. The first paragraph on page 2-7 refers to a "' clean' subset of check valves".

r The basis for choosing the valves in the subset should be specifically stated and the use of " clean" as the subset specifier should be avoided since it may be inferred by some readers that this is somehow'related to the valves exposure to boric acid.'

f

12. The last paragraph in section 2.1.3.1 implies but does not specifically state that BNL agrees that the PLG frequency calculation method is appropriate. BNL should specifica11'y state- one way or the other.-

13. The phrase "which may be too conservative for the increased statistic" appears in the first sentence on page 2-9. Does conservative mean large or small in this case?

I 4

14. Why is a range factor of 14 assumed for the median frequency of exceeding 1800

gpm leak rate? (p 2.9)

15. Perhaps the real answer to the " double valve" leakage frequency issue (raised by BNL for accumulator valve leakage data) lies in a backflow rate dependence .

for lamda-d which is neglected by this analysis. It is reasonable to suspect that a very low leak rate in an upstream valve has less probability of firmly seating a downstream check valve. Similarly, slowly developing upstream leakage may be less effective than sudden leakage for causing a check valve to

. seat tightly. A dangerous ISL would require rapid development of a large leak rate through both check valves, which means the disks must be opened or ruptured. This is'no the same failure' mode as the failure to seat tightly.

Thus, the mechanism for developing a large ISL may involve a different failure I

mode than is addressed by the one being used to empirically derive the frequencies.

16. At the end of the second paragraph on page 2-13, it is stated "The need was

- identified by a detailed review of the above four procedures." Was this the BNL review or the applicant's review.

17. In the top paragraph on page 2-14, it appears that HEP stands for Human Error Reliability.

O

a. . -2. - ~ . . - - - . . . . _ _ . - . , , . -. _ _-

}

4

-18. Section 2.1.4 neglects the hardware portion of the operator mitigation c.odel.

Is the reliability of the hardware properly represented by PLG's analyses?

Can the MOVs in the RHR discharge lines close against system pressure? If

+ not, can premature (prior to sufficient RCS depressuri:ation) operator actions

" to close these valves during an ISL cause the valves to fail in the open

- position, defeating oportor mitigation? If so, how does this affect reliability estimates for successful operator actions?

19. The second paragraph on page 2-19 mentions that onset of core uncovery can

/ begin in as little as 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> starting with a loss of DHR with the vessel j T '

partially drained and the decay heat load at 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> after shutdown. This

} f lf condition itself is rare. What is the time dependency of core uncovery (from j f the drained condition) as a functin of time from shutdown to loss of DHR?

Also, what is the. time to fuel damage? It should be significantly more

[p 4.

e

/ delayed ecmpared to the timing for accidents at power operation.

Ve

/ 20. On page 2-20, the video alarm system audible alarm for the Seabrook RHR S= tacti$n valves is mentioned. Is there a significant outage rate for this "i . alarm system during shutdown conditions?

21. On page 2-22, it is stated that Seabrook Tech Specs require either the PORVs or the RHR relief valves to be available, but not both. When the FORVs are not available, what is the status of the pressur's interlocks on the RHR suction line valves?

22. The last paragraph in section 2.2 (page 2-23) gives the NSAC-84 probability of core damage while shutdown for the Zion plant. What assumptions were made about the number of shutdowns or the amount of time spent in shutdown per year to allow this probability to be put in units of "per reactor year"? The cited dominant risk contributor sounds like a once-per-shutdown type activity, but other sequences are more likely continuous risk type situations. Are we sure that-the cited probability is actually intended to be the average contribution to the risk of one year of normal operation, as opposed to a full year at

- shutdown? ,

23. On figure 2.1, the line marked "BNL" is termed an " eye fit" to the data.

i Given the nature of the data, a computed linear fit would surely look the same, but it should be done, anyway, and the term " eye fit" should be deleted. *

24. On table 2.6, the value of 350 degrees F should be used for the initiation

temperature at Seabrook.

$ 25. On table 2.8, the risk of early fatalities increases when evacuation distance is increased form 1 mile to 2 miles for the optimistic analysis of SGTR. THis cannot be correct.

26. On page 3-1, the term " diameter" is used where the radial dimension of tne containment dome is given.

27. On page 3-10, the last sentence in the first paragraph says that "an underestimation of the, deformations would result in hicher predicted failure pressure for the penetrations." Several people have questioned whether it should say lgugr. ,Apparently, clearer wording is neeced. As I unc c stand the issue, BNL is concerned that SMA's calculations contain a nonconservatism that e

--y- y - - , - - ~ . - -w.-, ,m..-r --.r..

7% w + w v. -.w-, w p.,- r p,- m -, ww w - * ' r' w r'mr ==vv--"-wr'mww=reCw-+-h'=w- - - - - '

. .. .; .--:. ----..a.-----... - - . . . .- .- .~

F'.

1 BNL believes would cause SMA to predict. penetration failure pressures to be higher than BNL would predict them to be.

28. Are the values and units for the leak areas correct as apecified at the top of 1 page 3-137 - l

. l

29. The paragraph on page 3-15 concerning overtemperature failure of the i containment isolation valves should be included in the missing section on containment isolation-failure.

30. There seem to be several places where.the units " psi" were used in place of' "ksi". One is for the the median yield stress for rebar in table 3.1. The report should be reviewed for occurrence of this error elsewhere.

31.Ontable3.2,itisnotrealhclearifthedashesusedwiththenumbersinthe elevation heading are always indicative of negative elevations, or if some of them were intended to separate the ends of ranges.

32. In the second paragraph on page 4-1, there is a reference to early centainment failure, but I believe that it is really the combined probabilities of containment failure and containment bypass that is meant. This problem with

-terminology occurs throughout the section. " Containment failure", especially.

-" early containment failure", should mean failure of the containment due to -

stresses _ caused by the accident. Bypasses and preexisting leakages should.not be confused with c'ontainment structural failures caused by accidents. When an overall term is required for all plant damage states that result in early releases, they should be described in terms of "early releases". The second paragraph on page 4-4 contains another case of this confusion. The whole a

sectid*should be reviewed for the purpose of making this terminology clear and

.. consistent.

, 33. On page 4-4, what does "the composite failure of the two plants appear to be quite different" mean?

34. On page 4-4, does the term "this work" refer to NUREG-1150 or to this BNL Seabrook report?

i

~

35. At the bottom of the first paragraph of section 4.2 (page 4-5), it would be useful to state why the dry accident sequences resulted in higher containment failure probabilities.

36. The second paragraph on page 4-6 refers to the S6W release catagory as

" benign". This is not a proper choice of words, since it is a severe release

! catagory from the consequences perspective, and especially because of the word's connotations with respect to cancer. The same word is also used in the l

last paragraph of the same page and on figure 4.1. A better word should be

used throughout the report.

1

37. The third paragraph on ge 4-6 speaks of a shear failure mode at 140 psig, but page 3-5 speaks of y shear failure mode being above 157 psig.

38. I don't understand the fourth sentence in the second paragraph of section 4.3.

Isn't the probability of early containment failure a result of the more I

i

', -- _ _______.__.__.-_,_m_,._,....___ _ _ _ , _ _ _ . _ . _ . ~ . , , _ _ _

a. . . ..-.-.... - .~. . . - . - . - . . . . . . . . . , -

9 conservative treatment of containment performance? How can it be " combined with" in the sense of the consequence modeling?

. 39. At the end of sectio 5.2, it should be stated wtere the reader can find the results of combining the frequency and release catagory effects of the ISL

analyses that BNL has done. It might better be done at the end of section 5.3, but it should be done in at least one of these places.

t

40. On table 5.1, the Xenon release fraction correction has not been made for the 4

S2W source terms.

41. The first sentence of the third paragraph on page 6-1 describes the si:e of .

the EPZ as the most important guidance for emergency planners. It is probably the most definitive (and thus influential) guidance, but it is probably not the most important guidance.

f

42. On 5x10page and6.2, 6x10 the$,irst respectively. and third paragraphs Which is correct? give the core melt probability as

43. On page 6-8, the statement is made that the absolute risk numbers in table 6.2 "cannot be directly compared wit e information on individual risk provided in PLG-0465." More, accurately, cannot be compared directly to the NRC Eafety Goal,"but it can be compar to some of the information in PLG-0465.

44. Somewhere in this'section, dose probability vs-distance curves should be provided for the 50, 5 and i rem whole body dose levels ant 'or the 300, 25 and 5 rem thyroid dose levels. To do this properly, contributions from S3W and perhaps other low consequence source terms should be considered to see of they are significant contributors to the lower dose. level probabilities.

45. Tables such as 6.2 and 4.11 that present results for various distances of evacuation should be footnoted to indicate that no protective actions other than evacuation are included in the calculations.

46. The references should be properly completed as soon as feasible to provide for NRC staff review.

e

+ .+~-e 9 - _ c.m- ,--,_-

yw -,,n ,---,_,.,...._,_..,,+~.s., - - , - - , -, - ,,,-m., --,w

. . _ . _ _ _ . . _ _ _ -__ - , _ - - . . _ . .__ - -- - -- - = - - - - - - -

'6? ,

e i December 12, 1986 COMMENTS PERTINENT TO BNL REPORT " TECHNICAL EVALUATION OF THE EPZ-SENSITIVITY STUDY FOR SEABROOK," Draft Technical Report A-3852, December 5, 1986 Th8 overview message I get from this report is a limited conclusion that:

1. With no immediate protective actions, there is a possibility of exceedin NRC safety goals. .

2. With a one mile evacuation, none of the BNL sensitivity studies showed a j release which. exceeded the NRC safety goal. (I'm not clear on whether this is for an individual release or the conclusion can be applied to th'

-total picture; I suspect the former.)

3. None of the BNL work resulted in a conditional frequency of exceeding whole body doses as described in NUREG-0396, and at most distances, the

_ frequency is significantly less than in NUREG-0396.

4

4. Containment bypass sequences are important, and require further.

consideration. Particular attention should be paid to failure to isolat containment at power, accidents which originate under shutdown conditions, and steam generator tube rupture occurring as a result of an accident involving approach to core melt.

5. Investigation of containment response did not result in any significant differences from the licensee conclusions. BNL did find slightly lower containment failure pressures than estimated by the licensee, and occasionally estimated a larger containment opening at failure, but

~

.these findings would have no significant effect on containment reliability conclusions.

This is a draft report, and its quality and completeness clearly reflects tha- .

4 Otatus. A couple of weeks or so of concentrated work is needed to bring this up to a quality document. It is particularly weak in

/. 1 The summary, which should clearly reflect the findings and provide clear

. comparisons to such items as the NRC safety goal and should provide a better contrast to the licensee determinations. Containment response, 4

for example, isn't even in the summary.

2. The qualifications. The preface, for example, isn't even correct, aside L

1. from trying to relieve BNL of any responsibility for anything. (It is not a focused review of areas identified in PLG-0465 as most influential to risk, and one of the most important findings is evaluation of items that are not addressed or identified in PLG-0465.)

J 3. Readability. One must be up on the technical jargon of the field to understand this report. The nature of interest in Seabrook should be

[ relected in the writing so that a reasonably intelligent individual can read this report as a " stand alone" document.

i .- - - - - - , - . .--.-.,n m.. - - . ~ --~,w,'~.=,, -

.,-m,v,-,,-wm--. wmn, , sn-- - ---, .-.r---..,r-n,, -

c ', f . .

. r ..

What of the-future? First, we should. follow up on this report and its companion that I prepared on steam generator tube rupture, and we should cddress the outstanding (published) questions and responses. Minor follow up with the licensee and its contractors is indicated. I do not see this as a Icrge task, but its completion will take us into January, given the holiday season.

In toe longer range, I believe it will be necessary for us to be in a positici to address the Seabrook PRA and its follow up documents from an overview status. If I remember correctly, we stated we would do this with any PRA tha-was submitted, whether the PRA was requested or not, and this work was originated for Seabrook. It should be completed to a depth that we, and our contractor (s), can confidently state that we understand the work, and we support conclusions which we provide upon completion of our evaluation. This een be accomplished in a six month to one year effort involving an experience' contractor who is intimately familiar with Seabrook and the Seabrook PRA related information. About 3 contractor personnel will be required full time

_, plus contractor and NRC supervision and guidance.

Warren C. Lyon

& 6 9

1 F F ~+-?- , - - - - - - - - - - e + = w -- - . --- w- - ---w .www--v- v