ML19345B565
| ML19345B565 | |
| Person / Time | |
|---|---|
| Site: | Indian Point, Zion  |
| Issue date: | 09/17/1980 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-1764, NUDOCS 8012020068 | |
| Download: ML19345B565 (8) | |
Text
NA6 -l76 Y
,L l l If %
DRAFT COPY h
DATE ISSUED:
9/17/80 o
9 3/so MINUTES OF THE ACRS SUSCOMMITTEE MEETING ON CLASS 9 ACCIDENTS JULY 2,1980 INGLEWOOD, CA The ACRS Subcommittee on Class 9 Accidents met on July 2,1980 in Inglewood, California to continue examination of the role of Class 9 Accidents in the licensing process and to consider the question of possible mitigation features at Zion and Indian Point.
The Subcommittee also continued its review of the FY 81 and FY 82 NRC research budgets.
ACRS members 'present were W. Kerr, J. C. Mark, C. Siess, D. Okrent, M. Plesset, P. Shewmon, and H. Etherinoton.
NCRS consultants in attendance were J. Lee, R. Seale, W. Stratton, and S. Siegel.
Presentations to the Subcommittee were made by Thomas - NSAC, Cybulskis.- Battelle Columbus, Paddleford - Westinghouse, Peoples - Commonwealth Edison, Walker - Offshore Power Systems, Garrick - Pickard, Lowe & Garrick, Walser-Sargent and Lundy, Toland - United Engineers and Construc-tors, Meyer - NRC/NRR, Bernero - NRC/ PAS, and Kelber - NRC/RES.
A list of documents provided to the Subcommittee is attached.
Mitication of Small-Break LOCAs in PWRs (Summary of NSAC-2 Report)
Mr. G. Thomas, NSAC, gave a summary of the NSAC-2 report.
The presentation emphasized a TMI-2 scenario. He discussed ways to determine the occurrence of a small break LOCA and ways to mitigate the event. Protection from small break LOCAs is assured if the accident is correctly diagnosed and water can be delivered to the core. An accident in progress can be terminated i'f cooling is reestablished.
It was noted that no unambiguous method exists to detennine whether a small break LOCA has occurred.
It was also said that the source range neutron detectors may in the future serve as a means of detecting increasing core temperatures.
In current systems, however, they are not usable for this
)
purpose.
8012020 OGS
u,
_ Class 9 Accidents'nq - July 2,1980
' Reasonably small; quantities of water are sufficient for core-cooling using feed and-bleed.
For a 900 MWe plant..the quantities of subcooled (room tempera-ture) water required are:
340 gpm 15 minutes after shutdown; 170 gpm 21/2 hours af ter Shutdown; 100 gpm 1 day after shutdown; and 40 gpm 5 days after shutdown.
Mr. Thomas recommended that improvement of the man-machine interface have a high priority in order to reduce ambiguity of information for the operators.
Dr. Okrent noted the NSAC-2 report does not contain a good description of its assumptions, limitations, or omissions.
He indicated that lack of qualifiers on results is a common problem in reports,'and that it is a particularly important issue in risk analyses.
MARCH Computei Code Mr. Cybulskis, Battelle Columbus, discussed the MARCH (Meltdown Accident Response Characteristics) code. The code models containment response to cora melt accidents.
It treats boil-off of water from the core, and oxida-tion, melting and slumping of the core. The code also models hydrogen produc-tion, hydrogen combustion, steam production, and steam explosions. There are three meltdown options in the code, with varying amounts of conservatism.
The code includes containment spray, ccntainment coolers, and heat sinks. Fi ssion product transport is handled by coupling MARCH with CORRAL.
Calculations made with MARCH indicate-that with respect to hydrogen burn, the containment atmosphere will be more flammable with containment sprays on 1
rather than off, due to condensation of steam'.
Mr. Cybulskis indicated that MARCH should not be-used for design purposes. He recommended minimizing the chance-of early containment failure as the best was to reduce accident risk.
Code documentation should be comp 1ete'by ~ October 1980.
Coolability of TMI-2 Core for Various' Scenarios (Calculation Performed for the Kemeny Commission) kr. Paddleford, Westinghouse Nuclear Safety Department, described the results of calculations, relating to TMI-2, performed for the Kemeny Commission.
Analyses were performed on coolability under the following states of core degradation:
A e
1 Class 9 Accid 2nts Mtg ~
July 2,1980
.(a) The' core basically intact with no ECCS other than throttled (40 gpm) makeup,:and_the block valve open.
.(b) The core as a particle bed or molten pool.in the vessel with the vessel ' flooded externally with water, both with and without insulation on the vessel.
(c) -The core as a particle bed on the reactor cavity floor, flooded
~
with water.
The oxidation model used.was Cathcart/Pawel.
In case "a", approximately the top one foot of the clad was calculated to exceed 4900F, while the peak fuel (U0 ) temperature was ~4700F.
Considering heat-up and boil-off rates, 2
significant clad oxidation does not'begin until the water level is about 3 feet below the' top of the core, and then it proceeds very quickly.
In case "b", the particle bed was' found to be coolable if throttled makeup was supplied and the vessel was not insulated. Melt through of the vessel would not occur, though a portion of the vessel wall would melt. With insulation present, melt-through was predicted.
For case "c", the core as a particle bed in the reactor cavity was predicted to be coolable.
Short-Term Risk Study of-Zion and Indian Point Mr. Peoples',' Commonwealth Edison, gave an overview of Zion / Indian Point risk assessment efforts.- He indicated support for risk assessment and quantitative safety goals.
He believes the analysis of Class 9 accidents should be treated using best estimates, rather than using the usual licensing conservatisms.
Mr. Walker 0ffshore Power Systems, reported on the short term risk assessment of-Zion and Indian' Point-(Z/IP). The study followed WASH-1400, with a few changes to make.it' Z/IP specific. Dominant accident sequences fran WASH-1400 were selected, alor
'h containment failure modes and fission product release predictions.
.ccident sequences were then added and deleted, according to design differences between Z/IP and Surry. The short term study indicated Z/IP risk is no greater than Surry.-
l i
Class 9 Accidents Mtg July 2, 1980 Long-Term (a/8 month) Risk Study of Zion and Indian Point Mr. Garrick,-of Pickard, Lowe, and Garrick (PL&G), provided a summary of the Z/IP risk as'sessment being performed by PL&G.
The study is expected to:
(a) provide a. quantitative basis for evaluating the impact of various plant modi-fications on risk; (b) provide training for utilities on how to do' risk assess-ments, and (c) ' aid in determing the value of energency planning.
~
The assessment utilizes site specific consequence modeling, including effects of terrain, meteorology, and evacuation. Earthquakes are being treated in terms of probability vs magnitude for various size quakes. The effort is ex-pected to be'ccmpleted in September 1980.
Steam Exolosion and Core Coolability Analyses Mr. Henry, Fa'uske Associates, discussed potential for and consequences of steam explosions, relative to Z/IP risk assessment.
His study does not predict in-vessel steam explosions are for system pressures greater than 150 psia. Steam explosions are likely for lower pressures given the proper conditions. However, the event will not fail the pressure vessel nor generate a missile.
Rapid steam production (as opposed to steam explosion) from quenching an overheated core might overpressurize the primary system and fail steam generator tubes.
Fauske Associates studied this question, with a' preliminary conclusion that overpressurization does not appear to pose a problem.
An ex-vessel steam explosion in the reactor cavity was studied.
It was concluded that the pressure pulse generated does not challenge contaimnent.
Cooling of the core by reflux boiling'with heat removal through the~ steam genera-tors was considered.
It was shown to be feasible provided a sufficient amount of water'is present.
The question was investigated as to of whether the core would be coolable follcwing melthrough of the bottom of the pressure vessel.
He concluded that for all scenarios investigated, the _ core would be coolable. Depending on the melt through scenario
'^-
.s' +
- Class 9 Accidents Mtg July 2,
- 1980 either (a) much of the core material _would be swept out of the reactor cavity and dispersed throughout the containment floor or (b) the core material wou'.d re-main in' the reactor cavity, but:would be comprised of particle size and porosity amenable to cooling.'
Hydrogen Combustion in Containment' Mr. Lyparulo, Westinghouse, summarized some aspects of hydrogen burning. He.
indicated that (in dry air):
at 4 v/o hydrogen can burn upwards; at 6 v/o it can burn upwards and sideways;- at 81/2-v/o it can also burn downwards; up to 18 v/o the burn rate is reasonably slow; from 18 v/o to 59 v/o detonation occurs. The flame front temperature' required for burn propagation was indicated 0
to.be 710 C.
Below % 1/2%' concentration, only a few percent of the total hydrogen burns while above ~12%, almost all the hydrogen reacts.
About G.05 v/o water droplets of a'50 y a size would inert the containment atmosphere.
Strength of Zion Containment Mr. Walser, Sargent & Lundy, presented results of a study of the strength of Zion containments. Failure was defined as 1% tendon strain, which he calculates to correspond to a containment pressure of 134 psig.
The weakest portion of containment is the hoop tendons.
The containment liner is capable of 16% strain in biaxial tension before fracture occurs, whereas maximum liner strain was calculated to be 6% at 134 psig.
Strength of Indian Point Containment
, r. Toland, United Engineers and Constructors, gave results of a study of the -
. strength of Indian Point containments' Failure was defined as the yield point of the rebar. This was calculated to occur at a pressure of 126 psig.
ine factors of. safety that account for the difference between design pressure (47 psig) a'nd the calculated failure pressure are as follows (a) load factor of 1.5, (b) use of 0.9 times the minimum specified yield strength as opposed 4
f..
Class 9 Accidents:Mtg ~ July 2, 1980 to actual mechanical test results for the rebar; (c) excluding liner strength from the total. wall-strength;.and (d) excluding contribution of seismic rebar to wall strength.
'NRR Efforts on Z/IP, Risk Assessment, and Mitigation Features Mr. Meyer discussed the NRR study of mitigation features for Zion and Indian i
Point. The eventual objective is to produce an' order of magnitude reduction in risk. A report on this study is expected by the end of.1980.
In J21y 1980 NRC will, issue, for comment,'. criteria and requirements for mitigation systems.
There was discussion on the merits of issuing. criteria for-mitigating systems at this time,,as opposed to waiting until more is learned about the dominant accident sequences:and other aspects of risk.
NRR will clos'ely follow the: progress of the Pickard, Lowe, and Garrick risk study of Zion and Indian Point.
If the various risk studies of Z/IP show thEre is no undue risk,~ then these plants will not be singled out, but rather will be folded into Class 9 rulemaking.
Class 9 Accident research programs sponsored by NRR were described. These include (a)- steam production from molten material / water interaction, (b) i fragmentation and dispersion of molten material that is quenched, (c) molten core / concrete interaction, (d) molten core / refractory material interaction, (e) filtered vented containments, (f) feasibly of installing a core catcher in Zion and Indian Point; (g). hydrogen cortrol system;. and (h) containment failure modes.
RES Short Term Risk ~ Analysis of Indiaa Point Mr. Bernero, summarized a study performed by RES on Indian Point. The I"
study considered need.for power, analysis of_ comments by petitioners to shut-down Indian Point, and'a risk-assessment of the plant. The study considered the-specific site and plant design, as well.as effectiveness' of evacuation.
Four measures of risk were included: early fatalities, early injuries, latent cancers, and property damage. Early fatalities-(defined as >300 rem exposure) are generally confined to a 5 mi'le radius; early injuries (>50 rem exposure) are confined to a.50 mi.le radius; latent cancers extend to a 200 mile radius.
It m.
~
v y
y_
.i Class 9 Accidents Mtg July 2, 1980 was noted that evacuation is only effective in reducing exposures if it starts promptly (within one hour of diagnosis of trouble) and-proceeds quickly. The study indicated that despite the higher than average population density in the immediate vicinity of the site, Indian Point does not contribute inordinately to societal risk from nuclear power plants.. It was noted. however, that some risk contributors such as fire, sabotage, and earthquakes are not accounted for.
Little risk reduction would occur if the plants were derated to 50% power, it was concluded.
Class 9 Accident Research Program-Mr. Kelber, NRC/RES, discussed the research budget for Class 9 research. A four year program totaling $100-120 M was outlined.
He said that the in-formation needs are clearly defined and that the know-how exists to carry out the work. The generation of information by research should be such as to keep pace with Class 9 Accident rulemaking. The program includes research on miti-gation features, steam explosion, debris bed coolability, and core catchers.
Future Meetings The Subcommittee will meet in Washington, D.C. on August 28, 1980 to consider matters relating to-hydrogen in containment.
NOTE: For additional details, a complete transcript of the meeting it available in the NRC Public Document Rocm,1717 H St., NW, Washington, DC 20555 or from Alderson Reporting Company, Inc., 300 7th Street, S.W., _ Reporters Building, Washington, D.C. 20024 (202) 554-2345.
.:~
Class 9 Accidents Meeting 7/2/80 Documents Provided to the Subcommittee 1.
NSAC-2, March 1980
- 2. - ANL Report on Steam Explosions, February -1980 3.
Report of the Task Force on. Interim Operation of Indian Point (SECY-80-283) 4 June 12, 1980 4.
Mitigation of Small-Break LOCAs in PWRs - 11 Slides 5.
MARCH Computer Code - 15 Slides 6.
Coolability of TMI-2-Core for Various Scenarios - 19 Slides 7.
Short-Term Risk Study of Zion and Indian Point - 25 Slides 8.
Long-Tenn Risk Study of Zion and Indian Point - 29 Slides 9.
Stelm Explosion and Core Coolability Analyses - 34 Slides
- 10. Hydrogen Combustion in Containment - 27 Slides
- 11. Strength of Zion Containment.- 17 Slides
- 12. Strength of Indian Point Containment - 12~ Slides
- 13. NRR Efforts of Z/IP, Risk Assessment, and Mitigation Features - 10 Slides 14.
RES Short-Term Risk Analysis of Indian Point - 17 Slides i
- _ _