ML19281A865

Comments on Questions Raised at Nov 1978 Subcommittee Meeting & Applicant'S Response.Discusses Fuel Bundle Lift Potential,Suppression Pool temp-time History,Containment Load Calculations & Downcomer Loads on Drywell Floor

ML19281A865
Person / Time
Site:	Zimmer
Issue date:	03/06/1979
From:	Catton I CALIFORNIA, UNIV. OF, LOS ANGELES, CA
To:	Savio R Advisory Committee on Reactor Safeguards
References
ACRS-CT-1094, NUDOCS 7904160040
Download: ML19281A865 (4)

Text

.,

Boelter Hall 2567

• University of California Los Angeles, California 90024 March 6, 1979 C ~T. / O 9 4/

Cl  ?

ADY150RY COMMITT[E ON To: Dr. R. Savio /' stAcron sartcuAans u.s. N.ar

/

From: Ivan Catton g,tg g 9979

Subject:

Zimmer Subcommittec Meeting, 27 February 1979 M1 FM l2 l 7891hilttl2:3pi;5;G Copies To: 11. Bender, M. S. Plesset and A. Bates {

Several questions raised at the flovember Zimmer subcommittee meeting hue been addressed by the applicant by written response. Others have been answered by directing proper documents to rqy attention. The answers have been only in part satisfactory. The Zimmer plant, however, seems to be sufficiently over designed that many e.uestions, although relevant to future plants, are not important.

The following paragraphs contain some comments about the content of the written response received from the applicant as well as the meeting. The written responses are discussed first.

APPLICANTS WRITTEN RESPONSE T0 QUESTI0t4S RAISED DURING THE NOVEMBER SUBCOMMITTEE MEETING

1. Fuel Bundle Lift Potential. My interest in the pcssibility of fuel bundle lift came from questions raised by Mr. Ebersole. Following up on his questions, one finds on page 4 .~3 of NEDE-21354-P the statement,

" Impairment of normal CRD operation will occur when the friction between the control blade and the channels has increased to the point where the control rod no lenger settles into a notch as

' described in ....".

The no-settle condition is used as the basis for establishing a channel 790416CoVO i

defle'ction limit. The Triction associated with the no-settle condition is then used in calculating whether or not the bundle will be lifted by the control rods during a LOCA. My question has to do with how the calculations were made. In response to my question, the written response of the applicant simply refered to NEDE-21175-2-P.

I contacted Mr. Smith of GE and obtained NEDE-21175-2-P. Some arguments were given as to why it is conservative to neglect control rod friction in the fuel uplift analysis (the analysis was directed more towards seismic loads than LOCA). The arguments were too terse and I could not follow them.

Mr. Smith indicated that a letter would be prepared with force diagrams and proper references for further review. I have not, as yet, received the letter.

Further, the question of fuel bundle lift was noted to be a generic question.

2. Suppression Pool Temperature-time History. It has been noted by KWU that high pool temperatures lead to large lateral loads on the downcomers.

I raised the question about pool temperature when I found no reference to it in any of the applicants documents. Subsequent information obatined from Dr. Crawford of S&L and Cliff Anderson of the staff demonstrated that the Zimmer plant design uses the loads resulting from KWU steam chugging experi-ments at high pool temperatures. The loads are much higher than any measured with the 4T facility. In that, some future Mark II containment designs are not as sturdy as that of the Zimmer plant, extended 4T tests are planned by the f', ark II owners group. The question of pool temperature effects on down-comer lateral loads does not apply to the Zinner plant.

3. Containment Load Calculations. Questions about how pressure was calculated during a LOCA arose because the engineer from S&L did not know how it was done.

The method used was clarified in the written response by Dr. Crawford of S&L.

2

. . . . .. . . . . . . . . . , .-. ._.,, 3.,,...,,.,

.,i y .,

. . ,' -7 .,....M.=.~ ,

_.s.,.

, .2 . . *F - * .2 ~ .4 h. '? ' .-.......<u.--1

, . Containment pressura is calculated assuming adiabatic boundaries. The concrete temperature is calculated using heat transfer coefficients obtained from a standard heat transfer textbook. The source for the heat transfer coeffi-v . cient may not be the most current but I have been assured by Dr. Crawford t

, that there would have to be significant changes before there would be an '

17 effect on the integrity of the concrete. S.

4. Downcomer Loads on the Drywell Floor. Understanding how the floor loads are calculated for Zimmer hc., been a moving target. GE reports (NEDE 23627 and NEDC-21669-P) state that the lateral loads on the downcomers are random

in direction and magnitude. It is my belief that the statistical analysis

. ha . c been carried out sufficiently to yield such a conclusion. The Zimmer plant does not require an appeal to statistics. The Zimmer closure report -

A c states that floor loading is calculated by imposing 8,800 lbs to each down-f p

comer in directions yielding the maximum floor loading. I have not looked

• into how the " directions" are established. It is my opinion that the calcu-lations are sufficiently conservative to alleviate any concern about down-comer induced loading of the Zimmer plant drywell floor. J

, Some future plants may not be in a position to accommodate the bounding approach used in the Zimer design. It is my opinion that two physical phenomena govern the loads due to steam chugging and that the loads associated

[ with each must be analyzed separately. Data from KWU, 4T, UCLA and Creare

• $ all show a moderate amplitude pressure oscillation punctuated with occasional

+ i y.

f high amplitude spikes. Both the UCLA and Creare movies show that the  ;,

y gy

? moderate amplitude pressure oscillations are due to steam bubble growth and 14 y Y collapse near the vent exit. W The high amplitude spikes occur whc7 steam is g; s;

trapped against a solid boundary and a water hammer occurs. The two types 9 of phenomena must be assessed separately. If this is done, I believe it I

will be difficult to show randomness in amplitude and direction.

P 3 *

. . . - . . . . ..._.m. . , . . . . , ,

. -- ,, ,. y

, s.  ; g=  ; ; yym - ; -- y ; ; gn, g - -

.,~

e n * , * .

TOPICS DISCUSSED AT 27 FEBRUARY !C TING Reactor Cavity Suppor_t,. Pressure buildup in the annulus between the biological shield and the vessel during the initial blowdown phase of a LOCA lead to asymmetric loads on the reactor. The flow in the annulus is two-dimensional and transonic and is being analyzed with a one-dimensional code. The question is whether or not the code predictions are conservative.

Mr. Kudrick of the staff addressed this question. He indicated that it is a generaic question and that he does not have an answer at this time.

Mr. Gido at LASL is looking into the adequacy of the one-D calculations.

Reliance i Jeing placed on the BEACON code to eventually settle the question.

The question remains open at this time.

Ring Vortex Model. The ring vortex model developed by Dr. Chu and Dr. Grolmes, consultants to S&L, is used to assess loads on structures beneath the downcomers resulting during vent clearing. Use of the model leads to reduction of the loads found based on present methods described in the Mark II DFFR. It was not clear how much reduction was found. I have reviewed the document Dr. M. Grolmes based his analysis on and find that the Mark II downcomer transient jet falls outside the region of applicability.

I have not yet reviewed Dr. Chu's work. Discussions with Dr. Crawford, however, brought out that the Zimmer design does not need to appeal to the

" Ring Vortex Model".

Safety Relief Valves. The Zimmer plant will use T-Quenchers. Loads, however, are still based en use of the rams head design. This in my opinion is grossly conservative.

4