ML20197K236

Comments on Idcor, Matl Access Authorization Program Uncertainty Analysis. Submittal Gives Incomplete Description of Uncertainties in Severe Accident Analysis

ML20197K236
Person / Time
Issue date:	09/03/1985
From:	Barrett R Office of Nuclear Reactor Regulation
To:	Rosztoczy Z Office of Nuclear Reactor Regulation
Shared Package
ML20197K240	List: ML20197K236 ML20198B860 ML20199D807
References
FOIA-85-781 NUDOCS 8509060310
Download: ML20197K236 (3)

Text

.

s

,' s' . .

SEP 0 31985 ,

MEMORANDUM FOR: Zoltan R. Rosztoczy, Chief Research & Standards Coordination Branch Division of Safety Technology FROM: Richard J. Barrett Reactor Systems. Branch Division of Systems Integration

SUBJECT:

IDCOR UNCERTAINTY ANALYSIS The IDCOR submittal on "MAAP Uncertainty Analysis" is an incomplete description of the uncertainties in severe accident analysis. Most of the problems I have with this submittal have been brought to IDCOR's attention at previous technical exchange meetings. Specifically, the submittal has the following shortcomings-

1. It deals only with the uncertainty in containment response and fission product behavior. It does not deel with core melt frequency uncertainties or with uncertainties in the calculation of consequences (given a source term)

2. The containment and source term uncertainties do not include possible variations in physical models. For instance, there is no mention of direct heating failure or hydrogen burns following condensation deinerting in Zion. For Sequoyah TMLB, there is no mention of disabling the MAAP model that continuously burns off hydrogen in the reactor cavity. In Grand Gulf, there is no discussion of the possibility that the Morowitz plugging may not really stop pool bypass.

3. Uncertainties due to variations in sequence progressi.on are largely ignored. For instance, there is no mention of the possibility that the Zion V-sequence could be a large pipe break instead of a pump seal LOCA, or of the effect this would have on source terms. In Peach Bottom and Grand Gulf LOCA sequences, the CRD pump is always assumed operable. Furthermore, there is no discussion of the TC sequence, where venting and other sequence variables can have an enormous effect on risk. There is also no allowance for the possible effect on in-vessel hydrogen production of unpredictable automatic and operator actions such as repeated refloods, bumping RCP's, cycling SRV's or a failed attempt at feed and bleed.

The impact of operator actions on overall uncertainty is discussed in 7 p{ -

gp one page of text (page 5-1), and only insofar as the op r misu g A reduce risk. ggg [f g i i -

e n,ea u .. .. . . . . . . . .. . .. . . . . . . . .

.p.y . ..

su.~e > . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . .

CMt> . . . . . . . . . . .. .. . . . . . . ... . . . . . . * * *

• re soau m oo so, ucu ono OFFICIAL RECORD COPY v5cm e-ne

j',*,

. EP O 31985 f-n 3 , _ , . _ _ _ _ a _. . ._

4. The only uncertainty dealt with in any detail is that due to variations in parameters of the MAAP code. Even in this respect, the analysis is incomplete. Only single parameter variations were performed; thus no insight is given on the synergistic effect of simultaneously varying multiple parameters._ An' example which has been discussed with IDCOR is the effect on in-vessel hydrogen production of simultaneously turning off the channel blockage option (FBLOCK) and selecting a high value of the oxidation cutoff temperature (TZ00FF).

In some areas (and I have not done an exhaustive search for these), the parameter variation has been narrow. For the Peach Bottom TQVW 2

sequegce,thesizeof'theholeincontainmentwasvariedfrom1ft to 10 ft . InthgZion-TMLBsequence,anincreaseincor.tainmentfailure size to 0.5 ft led to significant increase in the cesium and iodine releasgs. I think it is easy to contemplate failures of greater than 0.5 ft For the TMLB sequence in Sequoyah, the full range.of calculated containment failure times was from 18.85 hr to 19.74 hr. The calculated in-vessel hydrogen production varied from 459 lhe to 492 lbs. Surely, this narrow band of values does not span the entire range of uncertainty. -

5. The. discussion of source term uncertainties is not sufficiently detailed to be reviewed. There is no discussion of Lanthanide group elements, and for Peach Bottom, only Cs! and Cs0H are discussed.

The uncertainty due to variation in the degree of core dispersal for Peach' Bottom is discussed, but it appears that the impact of a confined core on the release of non-volatiles was not examined. The report is not clear in this respect.

To be understandable, the report should describe the impact of each.

major uncertainty on the distribution of fission products throughout the plant and in the environment. Then the reviewer could understand whether the effect of one variation was being masked by some other

, phenomenon. For instance, does core confinement have little impact because the increased non-volatiles are being scrubbed by the

!- suppression pool? .

. In sunnary, the IDCOR submittal does not constitute a complete uncertainty l analysis. The assessment of containment performance and fission product i uncertainties should be expanded to include uncertainties in severe accident

_ phenomenology and accident sequence progression, including operator

, actions. Parameter variations should span a wider range in many instances,

and should include simultaneous variation of related variables.

1 l

l-

omcep .................. .................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

j sua==s> .................... ......./..............

.. .. A + 0.. ..

l onep ....................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Mooau na no soi mcw ono OFFICIAL RECORD COPY owo mi-m ea

w SEP 0 31995 8

The source term analysis should present more detail about the effect of uncertainty on the distribution of fission products in the plant.

The analysis should be expanded to include uncertainties in core melt frequency. Furthermore, the effect of uncertainties in calculations of offsite consequences should be documented.

Richard J.-Barrett Reactor Systems Branch Division of Systems Integration cc: R. Houston Distribution T. Speis Central File B. Sheron RSCB Rdg J. Rosenthal R. Barrett F. Coffman A. Buslik 9

1Y or.ca p .RS.B:DSI

. . . . . . . . . . . . . . . . . .7.. . 7

. . . . . . . . . . .. .. . . . . . . . . . . . . . . . . . .. .. . ~ ~ . . ~ . . . -

A suounme > .RBARRETT

. . . . . . . . . . . . . . . . . . . . . dm ...:.......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... .. .. . - u.$ o .. ..

can> T/J/.85 ............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . .

'*:'oaa sie no eos sacu os.o OFFICIAL RECORD COPY usc o i,. .m w I