ML20040B634
| ML20040B634 | |
| Person / Time | |
|---|---|
| Site: | Zion File:ZionSolutions icon.png |
| Issue date: | 01/18/1982 |
| From: | Lentine F COMMONWEALTH EDISON CO. |
| To: | Wigginton D Office of Nuclear Reactor Regulation |
| References | |
| RTR-NUREG-CR-1988 NUDOCS 8201260269 | |
| Download: ML20040B634 (5) | |
Text
.
Sb-J96 1
4 C:mmonw::cith Edison G
One First National Plata. Chscago, Illinois O
address Reply to: Post Olfice Box 767 O
Chicago, Illinois 60690
,g REC 7d*!CD c3 JAN,? 5 i992:> 5 i
t jfN?
January 18, 1982 g
Mr. D. A. Wigginton Operating Reactors Branch No. 1 Division of Licensing U. S. Nuclear Regulatory Commission Washington, D. C.
20555
Subject:
SANDIA Report:
NUREG/CR-1988 " Analysis of a Hypothetical Core Meltdown Accident Initiated by Loss of Off site Power for the Zion 1 Pressurized Water Reactor"
Dear Mr. Wigginton:
Commonweatlh Edison has reviewed the subject report and offers the attached comments regarding the report.
Please contact us if you have any ouestions on this material.
Very truly yours, i
Y Y
F. G. Lentine Nuclear Licensing Administrator sd Attachment i
3203N I
8201260269 820118 Ok PDR NUREC 0
CR/1988 C PDR b\\ \\
i
Commonwealth Edison Company Comments on NUREG/CR-1988 General 1.
NUREG/CR-1988 covers the core-containment progression of a postulated TML8' event at Zion Station.
The model employed by SANDIA in developing this assessment differs in many details from the best estimate model employed in the Zion Probabilistic Safety Study.
H ow e ve. c the results in both s
efforts are similar.
2.
We have noted in this NUREG and in NUREG/CR-0850 a tendency to employ a " corrected" MARCH cooe analysis followed by qualitative discussions of the problems inherent in even the " corrected" MARCH work.
We urge the NRC and its consultants to expand their efforts relative to these remaining problems by the application of simple, first principle analyses as used in the Zion Study.
3.
The SANDI A work points out the possibility of containment failure due to hydrogen combustion late in the sequence if AC power is restored and if containment safeguards are activated without proper care.
(SANDIA is careful to note that this is not the best estimate case).
Such a situation i s, how e ver, entirely proper as a postulate.
Unfortunately, NUREG/CR-0850 has lifted this out of the context of the SANDI A work and concluded that hydrogen is a problem at Zion.
This view was reinforced in a 12/17/81 presentation to the ACRS subcommittee on class 9 events delivered by J. Meyer of the NRC staff.
A perspective on this event is therefore warranted.
Such a hydrogen induced containment f ailure would add to the f requency of release category 2-R in the Zion Study.
Given the existing f requency of category 2-R, we can examine the potential increase in t.e frequency of 2-R by examining the frequency with which power will be restored af ter 6 to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> (but w ithin 12 to 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br />), the f requency with which containment safeguards will be loaded back onto the ESF buses, and the f requency with which the operation of such safeguards will be allowed to reduce containment pressure below the valve associated
. with steam inerting of the containment while large hydrogen inventories exist.
The f requency of power restoration can be judged to be about 0.1 for a mean value.
We judge that one of the first operator actions, upon restoration of at least one bus, would be to load containment ESF onto the bus.
Theref ore, we assign a mean f reauency of 1.0 to this action.
Since the timing into the event has now permitted the operation of technical support facilities to be established, and since the operators are now past the point of initial reaction to the event, we judge that the issue of hydrogen control is not likely to be ignored.
- Indeed, we would expect extensive sampling to be employed to ascertain the hydrogen concentration in the containment.
As a result, we judge the f requency of ESF operation being a l l ow c d tg de-inert the containment to be no greater than about 10-J.
(Recombiner operation would be employed to ultimately remove hydrogen.)
Theref ore, the mean f requency w ith w hich restoration of A.C. power after 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> could lead to containment failure is judged to be 10-4 times the frequency or a TMLB' event.
Such a f requency makes essentially no contribution to the frequency of release category 2-R and therefore, offers no significant increase in risk.
Specific 1.
Pages 4-1 and 4-2:
The discussion related to outcome "c"
omits consideration of one very significant success path.
If power is available, a heat sink can be considered to exist by means of the charging pumps drawing water f rom the RWST and relief from the primary system thru the PORVs.
The f an coolers will condense the resulting steam and the sump w ater will be available for high pressure recirculation when the RWST level drops.
2.
Page 4-4, Para. 3:
Long term RHR cooling would not require a full cavity.
Steam from the cavity, condensed by the sprays or f an coolers, would fill the recirculation sump and be recycled.
3.
Page 6-21, Para. 3:
PORV availability at Zion is enhanced by accumulators supplying each such valve.
4.
Pages 6-22 & 6-24:
The discussion of hydrogen evaluation, culminating in a conservative use of 100% Zr-H 0 hydrogen 2
is reasonable as a bounding case.
How e ve r, it should be noted that the NRC has taken even this conservative case, added postulated hydrogen from other sources, and raised the result as an issue.
Clearly, more care should have been taken to spell out what is realistic versus What is conservative.
. i 5.
Page 6-24 and Figure 6-15:
SANDIA's continued reliance on the MARCH vessel failure modes has significantly distorted the results of this work.
The Zion Study clearly shows that the failure modes employed by MARCH are extremely unlikely.
The f ailure of the welds around the incore guide tubes is by far the most likely mode.
The detailed results of the subsequent transient vary as a result of these considerations.
6.
Section 7:
The use of "mean" and " peak" consecuences to depict results, coupled with graphical representations as a function of distance from the reactor, obscure the results of this section.
At the very least, SANDIA should integrate the consequence curves over the population distribution and present CCDF's in frequency - magnitude terms.
7.
Page 7-6, section 7.2:
The containment penetrations at Zion do not constitute a weak link in terms of structural response.
Also, due to the air receivers, the pressurization system is not expected to be lost even if AC power is lost.
8.
Page 8-5, section 8.2:
Slumping of the core barrel is, by the design of vessel internals and built-in stops, severely i
limited.
The scenario envisioned leading to early immersion is not realistic.
9.
Page A-5:
The evaluation of transient progression given a dry cavity is, perhaps, an interesting bounding exercise.
H ow e ve r, SANDI A would have done better to have clearly emphasized the nature of such evaluations since it appears that the NRC is viewing this as a realistic scenario.
Given the Zion design it is extremely unlikely, if not impossible, to have a dry cavity.
- 10. Page A-35, para. 2:
Simple calculations will show that the molten material in the lower head and the residual gases in the vessel will exhaust to the cavity before accumulator water can reach the lower head.
The surging effects discussed are therefore not only speculative but also unrealistic.
- 11. Page A-40:
The discussion of steam explosion potential is distorted by the vessel failure mechanism presumed by MARCH.
In reality, a very limited amount of material could be involved in a steam explosion and the cavity water would, along with the core debris, be pushed out of the cavity by the discharge of steam and gases from the vessel.
.___=-_ -.-..
~
l 1
1 j
. i.
]
- 12. Page A-47/A-48, para. 1:
The logic of this paragraph is i
unclear.
If containment safeguards are not operational, the hydrogen is not combustible in the overall volume.
Local concentrations would tend to be even less flammable due to the nature of vessel failure and hydrogen release.
If fan coolers and/or sprays are operational, mixing of a i
very extensive nature is assured even ignoring the dynamics
)
associated with the vessel f ailure.
i 1
I i
i 1
l i
i i
i 3181N i
j i
l i
3 J
1 i
j i
I i
i i
i
..