ML20009C452

Proposed Findings of Fact & Conclusions of Law on Ucs Contention 12 Re Ability of safety-related Equipment to Function Under Conditions of Reactor Bldg Flooding

ML20009C452
Person / Time
Site:	Three Mile Island
Issue date:	07/13/1981
From:	Adler R PENNSYLVANIA, COMMONWEALTH OF
To:
References
NUDOCS 8107210076
Download: ML20009C452 (12)

Text

. - _ - _

4 PA 7/13/81 s =

UNITED STATES OF AMERICA g, / N NUCLEAR REGUIAIDRY C0tt4ISSION Y DocxGro BEFORE THE A10!HC SAFE 1Y AND LICENSING BOARD . ~g usyng 4

'l  ;'UU171981 > 2 g- Otti ~

In the Matter of De$e y J METROPOLITAN EDISON CQ4PANY, ) k

) Docket No. 50-289 o>

(Three Mile Island Nuclear ) (Restart) g - 4 Station, Unit No. 1) ) V' s C0ttOMEALTH OF PENNSYINANIA'S PROPCdED FINDIIES JU[ g 7981 A (.2 0F FACT AND CONCLUSIONS OF IAW ON UCS 12 (F100D LEVEL) u a ,,MIsIEuroer (c

tbte: Since the following proposed findings of fact and conclusfops of a\#)

lat relate to only one aspect of UCS 12, the Conulonwealth proposes thig D' rG/

these findings be inserted at the appropriate point in the Board's decision on UG 12.

Flood Level and Instrument Relocation

1. Tae Bo.,rd is particularly concerned with the ability of safety-related equipent at 1MI-l to function under conditions of reactor building flooo!.ng similar to those that occurred during the 341-2 '

accident. The water level in the reactor building that rcsulted from the TMI-2 accident has been approximately eight to nine feet above the containment floor. Keaten, et al., ff. Tr. 7558, at 6.

2. Licensee's witness testified that "the only environmental stress to which certain safety-related equipment at 1MI-2 was exposed that was beyond conditions for thich it was qualified was submergence due to flooding." It was then claimed that adequate protective measures will be completed prior to restart to prevent the recurrence of this problan. Braulke, ff. Tr. 6802, at 1-2. See also Cronenberger, ff.

Tr.16, 252, at 2; Keata , et al. , ff. Tr. 7558, at 4-5. 7he Staff's witness agreed with this testimony. Rosztoczy, ff. Tr. 21, 867, at 1.

3

~

4 8107210076 810713 PDR ADOCK 05000289 G PDR

3. According to the Licensee, all of the instr m ents on Table 1 of Keaten, et al., ff. Tr. 7558; with the exception of tlm NI-2 source range nuclear instrument indication are located below the 1MI-2 flood level. Therefore, Licensee presmes that these instrments failed due to submergence. Similarly, Licensee believes that all but six instruments from Table 2 of this testimcny (items 1, 2, 3, 5, 6, and 7) failed due to subnergence. Tr. 6884 (Braulke). These six instr m ents are above the IMI-2 flood level. Id_.

~4. Licensee testified that the instrunentation that presumably failed due to. submergence duriro 'he 3MI-2 accident was not designed for use under. water, and that the seals would eventually fail under submerged conditions. Tr. 6885 (Braulke).

5. The protective measure utilized by the Licensee to prevent a recurrence of instrment failure due to reactor building flooding is the relocation of safety-related instrunentation to above the maxinun ,

calculated flood level from design basis accidents. For this purpose, Licensee originally calculated a flood level of 5.94 feet above the contalment rloor. Tr. 6886 (Braulke). Later, this level was recalculated to 5A6 feet. Cronenberger, ff. Tr.16, 252, at 2-3.*

6. In explaining why the recalculated flood level was 5.66 feet rather than 5.94 feet, Mr. Cronenberger explained that certain conservatisms in the original analysis were unwarranted. For example, it was projected that the entire volume of the borated water storage tank would not be injected into the contalment, since a certain volune of water in the tank is below the level of suction in the tank. Tr. 16,

• . It is not cicar why the flood level was recalculated. The Poard notes, lx7dever, that some important safety equipment was relocated to a level below the original 5.94 feet but above the recalculated 5.66 feet.

See 17 infra.

. o 254 (Cronenberger). This does not explain, however, why the calculated flood level is lower than that experienc<u Juring the IMI-2 accident.

7. The steam generator and pressurizer level instrumentation have been relocated to an elevation of over 5.75 feet above the reacte building floor. This is the only equipment that is actually being relocated by Licensee. Cronenberger, ff. ~ Tr.16, 252, at 3. Licensee asserts that all other instrunentation f om Tables 1 and 2 which had a safety function were already located above the Unit 1 calculated flood level. Tr.16, 256 (Cronenberger) .

8. Thus, even the instrumentation that was relocated uas not raised to a level above the actual IMI-2 flood 1cuel or the original flood level calculation of 5.94 feet. The lowst portion of the housind of the relocated pressurizer and steam generator level histrumentation is at 5.8 feet. Tr. 16, 259 (Cronenberger).

9. The Board believes that the level of certainty inherent in .

Licensee's flood level calculation is questionable. The Staff's witness was questioned on this uncertainty as follows:

Q Do you think that three significant fir,ures is a reasonabl way to specify flood level?

Do you think you can detennine flood level that accurately?

A (W111ESS LaGRANGE) Well, I assime there are probably a lot of conservatisms in the calculation of that flood level, but I would find it hard to believe that you can get it right down to three digits.

Tr. 22, 002 (DiGrange)

10. The uncertainty inherent in flood level calculations is also dmonstrated by Licensee's teetinony on the failure of non-safety-related instn1ments. According to Licensee's witness, "a couple of than had tenninations that are very close to the flood level so we are not sure

e whether that actual termination point on the terminal block is actually covered with water or not." Tr. 6919 (Bruulke).

11. In light of this uncertainty and the significant difference between Licensee's calculated flood level and the level actually experienced at 'I12-2, the Board believes that a careful Staff review of Licensee's approach to this problem is warranted.

12. The NRC Staff did not, however, conduct a significant review of Licensee's calculations. The Staff witness testified that he did not review the adequacy of Licensee's calculation of the flood level in contairinent following a small break loss of coolant accident. Tr. 22, 000-01 (LaGrange) . Instead, the Staff simply assumed that the flood level calculated by Licensee was correct. Id. at 22, 001.* tbreover, the Staff did not consider the margin between the stated flood level and instrunent location. Id_.

13. The Staff's other witness agreed later, however, that an y extrenely conservative approach to flood level calculations is appropriate.

The witness testified that "the uncertainties of the calculation should be added to the txxninal valve, the calculated nominal wlue, and then the flood level established. But the established flood level should be something that the actual water level will never exceed." Tr. 22, 102-03 (Rosztoczy).

14. Apparently, the only information received by the Staff regarding the basis for Licensee's flood level calculation is stated in Licensee's October 31, 1980 response to IE Bulletin 79-01B. The letter states, in pertinent part:

Flood 1cvel in the Reactor Building was recalculated for 11R-1 following the 11E-2

• This was true both for the original (5.94 feet) and recalculated (5.66 feet) flood levels. Tr. 22, 091 (LaGrange). See also UCS Ex. 40, at 4.

accident. The flood level was based on the available tankage involved in accident mitigation and the fluid losses out of the break. The flood level was conservatively calculated to be 5.94 feet (Elevation 286.94 ft.).

~

UCS Ex. 34, at 2. This infonnation is not sufficiently detailed to determine what systens are assumed by Licensee to contribute to accident flood levels.

15. Ibrmal leakage fran other contaminated and uncontaminated water systans, in fact, contributed to flooding at 'IML-?. This leakage is calculated to be about 6,000 gallons per month, resulting in a monthly increase in water level of about one-tenth of a foot. Stoddard, ff. Tr. 10, 159, at 12.

16. In addition, it is apparent that the IMI-1 reactor building may accurtulate more water that at 'IMI-2, since there is "less likelihood that substantial quantities of liquid radwuste will be inadvertently transferred from the reactor building to tankage in the auxiliary building." Fuhrer & McGoey, ff. Tr.10, 020, at 6.

17. The Staff's witness was not even sufficiently familiar with the systems at THI-l to discuss which systems would contribute to reactor building flooding. The witness would, however, consider "any ... source of water which could [by) reasonable means be entering the containment or the appropriate building," including nonnal system leakage but excluding secondary breaks in addition to the initial accident. Tr. 22, 095-97. The witness did not know whether all of these sources were included in Licensee's flood level calculation, although he testified  !

that such NRC review would be desirable. Id.

18. The Board is at a loss to explain the lack of Staff review of this issue. The Staff's testinony claimed that "the staff performed

their own analyses and calculations to assess the adequacy of the licensee's specified environmental conditions." Rosztoczy, ff. Tr. 21, 867, at 3. Yet the Staff performed no analysis or calculation of tbe only envirorrnental condition that, according to both the Staff's and the Licensee's testimony, resulted in instrument failures at T41-2.

On cross-exanination, the Staff's witness provided no reasonable explanation for this approach. Tr. 22, 093-94 (Rosztoczy) .

19. Apparently, t.he Staff did not recalculate the flood level for 141-1 as part of a conscious decision merely to accept the flood level calculations of all licensees. Bis decision was based primarily on time considerations. In this respect,141-1 ms treated the same as all other reactors, despite the fact that the flood level of the TMI-2 accident exceeded the calculated 111-1 flood level. Tr. 22, 098-99 (iiosztoczy) .

20. The Staff's witness did agree, however, that the flood level .,

of the T4I-2 accident " raises some questions" regarding the TII-1 flood level. The witness responded to this question in tw ways. First, the witness claimed that the TMI-2 flood level ms high due to the pumping of 'IMI-l storage water into the T41-2 reactor building. The source of this information as conversations between the witness and Met Ed cmployees. The witness did not know the reason for this transfer or the patlway that was used. Tr. 22, 099; 22, 101-02 (Rosztoczy).

There is no assurance that this type of transfer, therefore, would not be repeated. Moreover, the Board questions the reliability of this evidence due to the heresay nature of the testimony, and in light of the witness' inability to answer any related questions.

21. The Board can only presume that T41-1 water was transferred

~_

through the M-2 borated water storage tank to serve a reactor coolant function, as suggested by the Conmonwealth's cross-examination, and not directly into the reactor building. See Tr. 22, 102 (Dornsife).

If this were the case, additional borated water could be used in this manner during an accident at M-1 as well, as long as a source of borated water was available. Alternatively, Licensee coulu continue to cool the reactor by entering the recirculation mode, as discussed infra.

22. The witness' second response was his " understanding".that the TMI-l operating procedures would not permit a repeat of the 911-2 flood level. Tr. 22, 099 (Rosztoczy). Yet the witness had sought 60 informition from other NRC branches to determine whether such operational limitations are appropriate or possible. The Staff intends to conduct this analyse s sometime in the future. Tr. 22,100 (Rosztoczy).

23. Essentially, the Staff's position is that it is safe to restart M -1 provided that the flood level is maintained below the specified -

value. The Staff's witness testified that, in order to ensure that the flood level does not exceed the specified value, it would be appropriate to impose a license condition or "something of that sort" on the IMI-l energency procedures. Tr. 20, 002-03 (Rosztoczy). This witness also testified, however, that the operational limitations necessary to achieve this result were beyond his responsibility and knowledge. Tr. 22, 004 (Rosztoczy).

24. Although the Staff's approach might appear appropriate on its face, it is not supported by.the record. Licensee relies on the reactor building in order to store liquid wastes in the event of an accident at IML-1. Licensee claims that "[t]he Unit 1 reactor building can safely contain the maxinum volune of water available from post-accident Unit 1

= . __

sourcee during an accident." The same testimny states that instrument relocation will "decreasr 'the likelihood that such'instru,entation will be flooded in the event it is necessary to ccotain large volumes of liquids in the Unit 1 reactor building." The building can contain approxhnately 456,000 gallons before instruments will be adversely affected. _ Fuhrer & McGoey, ff. Tr.10, 020, at 5-6. The Board notes .

the uncertainty in which this language is couched.

25. Licensee's witness on plant separation alco testified orally that the TMI-1 reactor building has a capacity of approximitely 456,000 gallons before important plant instrunentatica may be affected.

Apparently, Licensee's calculated flood level is based on this 456,000 gallon figure. Tr.10, 038-39 (Fuhrer) . The borated water storage tank alone has approximately 400,000 gallons, nost of which my contribute to the reactor building flood level. Therefore, this witness believed

, that an accident at 1MI-l could very much push the 456,000 gallon limit, ,

when considering contributions from other portions of the primary system. Id. at 10, 043-44.

26. The level of water in the containnent from a loss of coolant accident with heat renoval through th2 steam generators or through feed and bleed would result in the same maxinum water level in the containment building. This would result fran emptying the entire borated water storage tank and the majority of the primary system inventory. Tr. 4736 (Jones).

a

27. On cross-examination, Dr. Jordan asked Licensee's witness, Mr. Braulke, why the calculated flood level (then 5.94 feet) was lower than the 8-9 feet experienced at 1ML-2. Mr. Braulke responded that:

There may be some differences with the containment structure. I think as far as the 1MI-2 accident was concerned, a level similar

_8-

to what we have calculated for ntI-t was reached after, I think, about three days or four days.

And at that time there was the--at least it is my underrtanding--the conscious and deliberate decision not to go into the recirculation mode but to continue the mode they were in, which would have continued to add water to t' e building, and I think that increased a little over a half an inch a day up through the first month.

Tr. 6916 (Braulke).

28. herefore, in order to prevent recurrence of the HIE-1 flood level, Licensee would have to enter the recirculation mode. 'Ihis requires operation of "a few motor-operated valves." Tr. 6917 (Braulke) . In addition, it might be necessary to use the decay heat system coolers

. to cool the water prior to recirculation. Id.

29. In fact, the recirculation mode has not been used in this manner. %e decision during the 'IMI-2 accident was not to enter the recirculation mode. Tr. 6918 (Braulke).

30. Licensee's witness was not able to give an authoritative -

answer as to why Licensee elected not to enter the recirculation mode.

Le witness knew that one reason was that Licensee wanted to hold the containment water in reserve. On cross-examination, it was suggested that additional reasons involved the high-level of radiation in the containment building water that would have to be carried outside of containment, and the high leakage rate of the decay heat removal system that might have to be used. h witness merely answered that "there might have been some concern over (those] points." Tr. 6924 (Braulke).

Le Board is not able to find, based on this record, that Licensee clearly will be able to enter the recirculation mode safely under all circunstances.

31. W e margin of error for equipment relocation above the potential flood lu. .1 also affects the desirability of using the feed I

and bleed mde of heat removal frm the primary systs, since feed and bleed results in the flooding of the contaiment building. Tr.

4729-31 (Jones /Keaten). In fact, the B&W conputer analyses used to test the adequacy of feed and bleed heat rmoval do not account for the potential for equipment flooding. Tr. 4730 (Jones).

32. ' he Board is extremely displeased that the Staff produced no knowledgeable witnesses on flood level calculations, if such a review was conducted at all by the Staff. This record deficiericy occurred despite the Conmonwealth's specific request for a witness on-this topic. Tr. 22, 104; 22, 157-58 (Cutchin, R. Ad)cr). See Irtter from Robert W. Adler, Attorney for the Commonwealth, to Administrative Judges Smith, Jordan, and Little (June 23, 1981). The NRC Regulations provide that:

In a proceeding in which the NRC is a party, the NRC Staff will make available one or more witnesses designated by the Executive Director ,

for Operations, for oral examination at the hearing or on deposition regarding any matter, not privileged, which is relevant to the issues in the proceeding.

10 C.F.R. $2.720(h)(2) (mphasis added) . This requirement includes.the

" obligation to lay all relevant materials before the Board to enable it adequately to dispose of the issues before it." gnsolidatedEdisonCo.

of N.Y. , Inc. (Indian Point Station, Units 1, 2, & 3), CLI-77-2, 5 NRC 13, 15 (1977).

33. The approach of the Division of Inspection and Enforement appears no better. The Staff conducted an onsite inspection of only a "very limited nmber of components." UCS Ex. 40, at 5. 11mrefore, the Staff does not know what other instrments are potentially subject to failure due to submergence. Consistent with the Staff's approach moreover, the inspection would have focused only on equipnent potentially bel m Licensee's calculated flood level. There is even less basis for knaving the susceptibility to submergence of equipnent located above Licensee's calculated flood level but below the TMI-2 flood level.

3 ', The Board has already expressed some reservations regarding Licensee's approach to flood level calculations and instrunent relocation.

~

Tr. 6918 (Jordan). After further evaluation of the record, the Board concludes that the following requirements are necessary to provide reasonable assurance that IMI-l can-be operated without endangering the

' health and safety- of the public:

6 (1) The following short-teun actions must be completed and certified to the Conmission prior to restart:

(a) The Staff is directed to conduct a complete review of Licensee's flood level calculations. This review shall ensure that Licensee's calculations are based on the appropriate assunptions. In particular, the Staff shall ensure that all systems, including nonml systan leakagh, that may contribute to reactor building flooding, were included in Licensee's analysis. The Staff shall ensure that Licensee's analysis was

. performed with an appropriate degree of conservatism, as set forth in 113, supra. If Licensee's calculations were not based on the appropriate conservatisms, the actions set forth in (2), below, shall be required; (b) The Staff shall conduct a complete review of the operational limitations that must be imposed on Licensee to ensure that the reactor building flood level does not exceed Licensee's calculated maxinun flood level. This analysis i

O shall ensure, for the design basis accidents expected to result in reactor building flooding, that it is possible and appropriate from an operational standpoint to maintain the flood level within the calculated maximum flood level.

In particular, the Staff shall review the ability of Licensee to enter the recirculation mode under all postulated accident circumstances where the recirculation mode m uld be necessary to maintain flood levels within Licensee's calculation. 'Ihe Staff shall review all energency procedures for these accidents to ensure that these operational limitations are properly incorporated into the procedures. If the necessary operational limitations are not possible or appropriate for some postulated events, the Licensee shall be required to demonstrate prior to restart why the operat. ion of 'IME-1 under these circumstances will not endanger the health and safety of the public. ,

(2) If it is determined pursuant to (1)(a), above, that Licensee's calculation of nuxinun expected flood level did not cng .oy the appropriate degree of conservatism, a new flood level shall be detennined using the correct assunptions. Licensee shall then be required to relocate all equipment inportant to the safe operation of the plant above the newly calculated flood level. Tnis relocation shall be required by June 30, 1982.*

Respectfull submitted, e

ROBERT W. ADLER Attorney for the Conmorw.'ealth

-

• June 30, 1982, is the compliance deadline for environmental qualification of safety equipment set forth in Petition for Emergency and Remedial Action, CLI-80-21, 11 N.R.C. 707 (1980).