ML20213D755

Requests Assistance in Assuring That Facility Load Definition Program Proceeds on Schedule Consistent W/Licensing Process.Complete Description of Definition Methodologies Must Be Filed by 820812

ML20213D755
Person / Time
Site:	Columbia
Issue date:	07/02/1981
From:	Rubenstein L Office of Nuclear Reactor Regulation
To:	Tedesco R Office of Nuclear Reactor Regulation
References
CON-WNP-0369, CON-WNP-369 NUDOCS 8107200009
Download: ML20213D755 (2)

Text

--_ - _ -

~'

']

s h e /

1. x JUL 2 1981 f ~  %({\y1 yyh I/0y%

d D q0 p 10

- T$$' ,.}

MEMORANDUM FOR: R. Tedesco, Assistant Director for Licensino/DL .

\Qu (r, . ./x -

FROM: L. Rubenstein, Assistant Director for Core and Con,taTn '

ment Systems, DSI

SUBJECT:

REVIEW SCHEDULE FOR WPPSS-2 The present NRR work schedule calls for a Safety Evaluation Report (SER) on the WPPSS-2 application by February 12, 1982. If we are to meet this SER date with-out an excessive number of open items in the area of dynamic loads in the sup-pression pool, we will need your assistani:e in getting the applicant to acceler-ate his work program in the associated loads definition area. Specifically, we will need the applicant's final submittal on load definition by August 12, 1981.

The background information dealing with this concern is as follows. In January 1981 we had a meeting with the applicant to discuss his improved chugging and condensation oscillation load methodologies. Formal documentation of this in-formation was scheduled for early May 1981. During the January meeting, we in-fonned the applicant that due to the availability of staff approved generic load specifications, they will have to justify the need for a plant unique re-view to the Director, Division of Systems Integration (DSI) (see memorandum from F. Eltawila to W. Butler dated Febesary 9,1981 (Enclosure 1). We also indicated that if such a plant unique review was found to be necessary, our review could not begin before January 1982, due to the severe shortage in staff resources.

Since then, the applicant has neither requested a meeting with the DSI Director to justify the need for a plant unique review nor submitted the docu$entation for the new plant unique methodologies.

The accelerated licensing schedoles in NRR has brought our reviews for LaSalle, Zininer, Shoreham and Susquehanna to substantial completion. As a result we are in a position to acconsnodate a limited amount of plant unique review for WPPSS-2.

Ilowever, based on prior experience on this subject of load definition and based on our understanding of the plant unique approaches being used by the WPPSS-2 applicant, we expect to need at least six months to complete our review once the associated documents are filed.

Accordingly, we request your assistance in assuring that the WPPSS-2 load deft-nition program proceeds on a schedule that is consistent with our licensing schedule. This means a complete description of the WPPSS-2 load definition methodologies (SRV load methodology, chugging load methodology and condensation oscillation load methodology) needs to be filed no later than August 12, 1982.

Central files -

4

$bn ~ ~ ~ ~ ~ ~

/]

"1[7 Lester S. Rubenstein, Assistant Director for Core and Containment Systems k DOOKob000$7 se ,/cf , Division of Systems Integration As stated c"q . I  ! CSB:DSI .. ,- !CSB:DSI !CSB:DSI 'AD:CCS:DSI

~'

'"""k .

.cc: (See.page.2).. ' WBuder "I RdbensEein cm

Contact:

F. Eltawila, CSB:DSI ]l FEltawi12;jf

, / ')./81 7 7/ /81

! JXudrick I

7/ /81 7/ /81

_ _ _ . _ . _ - __. z 29418_. __

- J

~ ~ ~ "

m: 4 - m -

c ?,lc u t p.gc c K F c P T 7 1-

s

- q

~

4< -

! h,

~

JUL 2 1981 R. Tedesco -

2-cc: R. Mattson W. Butler A. Schwencer J. Kudrick R. Auluck F. Eltawila

• • • • es-. . , ._ ., , , . ,

4 yg

.f.

.sm, j  ; i l

'N I 4 i

r ,

' ' ' ' ' ~ . ~ . . . . . _ , _ _ , _ . ,, _

c.==>c u:..? = == .:.:.,. , ; 2.

_ f Mi

% ^

% UNITED STATES 8 7. NUCLEAR REGULATORY COMMISSION g ' : E WASHINGTON. D. C. 20555

\;-

• • ..+

/ TES 0 19 9 ;

MEMORANDUM FOR: W. Butler, Chief, Containment Systems Branch, DSI FROM: F. Eltawila, Containment Systems Branch, DSI

SUBJECT:

MEETING WITH WPPSS TO DISCUSS CONTAINMENT HYDRODYNAMIC LOADS FOR WNP-2 (JANUARY 2,1981, BETHESDA, MARYLAND)

In a letter dated December 19, 1980, the Washington Public Power Supply System (WPPSS) requested a meeting with the staff to discuss the chugging and conden-sation oscillation (CO) load definitions. Preliminary results from the ongoing plant unique chugging and C0 methodologies were presented to the staff during the meeting of January 21, 1981. An attendance list and a copy of the meeting handouts are enclosed.

Background

UPPSS initiated a program to develop an improved chugging load specification representing a reduction in the loads from the conservative lead plant chugging load specification set forth in NUREG-0487. A report titled, " Chugging Loads -

Improved Definitien and Application Methodology to Mark II Containments," was submitted to the staff in 1979 describing the WPPSS method. The staff reviewed this report and expressed concerns regarding the statistical averaging of the 4T data to arrive at a source specification at the vent exit.

During 1979 and early 1980, the Mark II Owners Group conducted additional tests on a modified 4T facility (4TC0) to better understand the LOCA related condensa-tion phenomena. Results of the 4TC0 tests led to the raodification of the NUREG-0487 specified lead plant CO and chugging pool boundary loads. The staff. is currently preparing Supplement 2 to NUREG-0487 containing an evaluation of the proposed alternatives to the lead plant CO and chugging load specifications.

The staff requested WPPSS to factor in the recently developed data from the 4TC0 l tests into the present improved chugging load definition, recognizing that its concerns regarding the statistical averaging of the 4T data, which is utilized by WPPSS, remain unanswered.

In addition to the plant unique chugging load definition, WPPSS advised the staff, in a letter dated October 1,1980, of preliminary indications that the generic CO l load definition may not be tolerable for WNP-2 and plant unique review may also be necessary. This item was not previously identified by the applicant as requir-ing a plant unique review.

WPPSS Presentation .

I. Improved Chugging load Methodology Dr. M. Ettouney of Burns & Roe provided an overview of the new improved chugging load methodology for application to WNP-2 containment. The load L a a c ni Q f.

_na w

m t TE3 0 m7 W. Butler 2 valuation methodology reflects the data obtained from the 4T tests that was utilized in the original chugging load report and the new 4TCO test data.

Engineering evaluation of the 4TCO data base indicated that the chugging data can be grouped into two distinct types. Each excites different modes of the 4T system and, therefore, it 's not appropriate to consider them together in the statistical evaluation to arrive at source specifications.

The two groups are:

1) Normal chugs, similar'to those observed in the 4T data, influenced by vent response; and

2) Super chugs, with characteristics (pressure amplitude and frequency content) differer.t from those of the normal chugs, influenced by both the vent as well as the rest of the 4T system.

The improved load definition methodology for the normal chugs is the same as presented in the present chugging load report with the necessary amplf-tude changes. This report was previously reviewed by the staff and it was concluded that sounder arguments than those presented should be given to justify the statistical analyses of the 4T data to arrive at a conservative source specification.

It was also concluded that available multi-vent test data should be studied to verify the statistical averaging of the 4T data.

A new improved load definition methodology for the superchug (a total of 7 chugs have been identified as superchugs) has been developed. This new method accounts for the 4TCO facility components whose signature, such as wave form (acceleration vs pressure), water and steam properties, is pres-ent in the superchug boundary pressure traces. It was determined that the normal chug will be best represented by a pressure source and the super-chug will be represented by acceleration source.

The source terms (for normal and superchugs) thus developed will be used in an improved chugging load definition to be submitted to the staff in draft form in April 1981. These source tems w'ill also be used to gener- -

ate JAERI wall loads for comparison with the available JAERI data.'

!!. Improved CO Load Definition The applicant has not proposed CO load specifications for WNP-2; however, engineering evaluation of the 4TCO data indicate that fluid structure in-teraction are significant and when accounted for in the 4T system, signifi-cant reduction to the CO load will result.

III. SRV 1mproved load Definition The purpose of the meeting was to review the applicant's response to a number of questions related to the WPPSS-2 plant unique SRV discharge load methodology (see memorandum from E. Eltawila to W. R. Butler, dated

~

m _

FE 3 1931 W. Butler '

September 18, 1980 for a summary ' description of the methodology). These questions were fonnally trancmitted to the LPM and informally to the ap-plicant by the Contaitsent Systeins Branch on December 19. 1980.

With one exception, the appl $r. ant's response: to our questf or.s were satis-factory. The information which was supplied ~in the fenu of handouts should suffice to clarify most of the issues and concerns raised by the staff and its coasultants, Additional revike of the riew informaticn is required, however, to decide shetner any of the questions can be eliminpted.

The exception mentioned above refers to the . issue of multiple valve actua-tion (MVA) effects. The applicant f adicated an unawareness that the Caorso inplant test data exhibited higher boundary loads during NVA tests than first actuation single valve actuations. This stems from their excluslye use of Sensor P19 to serve as an indicator of toundary load magnitudes. We suggested that it was more appropriate,particularly for MVA cases, to use the highest recorded values for test purposes, The applicant agreed with this position and stated its intent to re-examine the data base Jr1 this light.

1 One additional action item which the applicant intends to pursue is related to the vertical distribution of pressure amplitudes. (The applicaqt's ap-proach differs significantly from the existing DFTR methodology.) We ,

pointed out that the sensors from which the Caorso pressure data are obtained to demonstrate the adequacy of their proposed distribution are characterized by the General Electric Company (in the Caor50 test Report NEDE-24757-P) as being of poor accuracy. In view of this, the applicant intends to reanalyze these data and establish whether the reported poor accuracy will impact on their design specification.

Staff Comments The staff indicated that it did not envision a major review regarding the type of source to be used with the improved chugging Icad definition methodology. It appears from the applicant's presentation that a significant review of the type of source is needed. In addition, substantial comparison of results obtained using the developed source in the improved model with data from the multivent test facility (JAERI) and with data from another single vent test facility (GKM-!!M) will be necessary and will require additional reviev efforts.

The staff also indicated that, since the Mark II Owners Group is c.grrently in the process of establishing a generic chugging load specification, WPPSS should justify the need for plant unique review to the Director, Division of Gystems Integration. We have also indicated that if such plant unique review ms fcond to be necessary, our review will not begin before January 1982 due to the Severe manpower shortage. In addition, the staff and its consultant expressed concerns in the following:

m -

~

FEE ? 1931 W. Butler ,

1) Revised Chugging Load Definition The need .to introduce a new type of chugging source, i.e., an acceler-ation pulse instead of a pressure pulse, to account for the large am-plitude chugs in the 4TCO tests is troublesome for several reasons:

a) it seems very difficult to accept on physical grounds, and if ac-cepted, would open many questions about the applicability of types of sources in general.

b) the stark difference in frequency response obtained from pressure versus acceleration pulses in the WPPSS 4T model is contrary to usual structural response experience, and suggests that the model-ing and entire methodology must be carefully examined to preclude any sort of built-in bias, c) verification o'f the methodology by comparison with JAERI data could be very difficult with the present axisymmetric model, i

2) Revised CO Load Definition The arguments presented were sketchy and based on very simplified analyses. It is difficult to believe the major effect attributed to FSI in the presentation since:

a) the arguments were neither complete nor convincing.

b) the particular 4TCO run chosen to present the case was not shown to be representative of C0 observations in general.

The contradictions between the conclusion drawn by WPPSS regarding the importance of FSI in CO and those stated in the generic Mark II C0 methodology need resolution by comparing data bases from facilities of differing stiffness such as GKM-II-M.

Wk Farouk Eltawila Containment Systems Branch Division of Systems Integration ec: D. Ross L. Rubenstein R. Tedesco ,

B. J. Youngblood D. Lynch R. Auluck J. Kudrick J. Lehner, BNL C. Economus, BNL

.