ML20127M203
| ML20127M203 | |
| Person / Time | |
|---|---|
| Issue date: | 03/01/1985 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Boehnert P Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1800, NUDOCS 8505230208 | |
| Download: ML20127M203 (2) | |
Text
D
'PoR osim-8}C /66-.
/
p-I Raady to send. Open down1'oad file.
Enter <CR> to start?
TOs PAUL BOEHNERT-FROM:
IVAN CATTON g
SUBJECT:
ECCS SUBCOMMITTEE MEETING H Street 21 February 1985 g
M:rch 1, 1985
-9 Of the four topics covered at the meeting, Appendix K change 10 the most important.
The industry and NRC must go to Best Ectimate methods even though the transition will not be easy.
A g
crcat deal of thought as to how it will be done is needed and a
plcn laid out that makes sense.
Emergency sump plugging by fiberglass, although important, seems to be only a part of the g
question about impact of high energy water release on its acrger curroundings.
Removal of UH1 from an ice condensor plant d; serves a great deal more study before it is allowed to happen.
g Appendix K. RES activity directed towards changing Appendix K are long overdue.
Clear thinking will be needed at the outset j]
or the desired goals will never be achieved.
Convincing oneself that there is conservatism in a calculation is f ar easier than trying to qu,antif y it.
Efforts to quantify conservatism in the
- g started when the advanced code effort began.
NUREGs have codes b en written on the subject.
Papers have been published and RES hns sponsored research on the subject.
No su,ccessful method for r3 quantifying conservatism has been found.
Statistical approaches
!d cuch as used by GE are unsatisfactory as well.
One can rs ctatistically analy:e data and argue that one has an upper or
. 'l lower bound of a
measured parameter like the heat transfer coefficient.
One cannot statistically evaluate the impact of the overly s2rplified flow regime map found in TRAC.
Marginal r"3 rGpresentation of physics is not a statistical problem and is not ccalable without a great deal of judgement.
oe Rh The advanced codes that will be used to obtain Best Estimate 3
cciculations (BE) have their own problems.
The codes are not y
unique in that results depend on nodali=ation and time steps.
7" Skilled users such as Ken Williams have learned through practice e
g how to use the codes and as a result one can depend on their E
f,!
g a results.
Calculated results'usually show good comparison with p
g integral facility data because there is a great deal of post-h diction before pre-diction.
Predicted results f or a f ull scale h@ y a
plant will be a strong function of the abilities of the analyst f3 o
om using the codos.
This is where the greatest uncertainty will 70 $
e S
"E lie.
8W8 Even with the gloom enpressed above, I believe BE is the N m n -
$$d only reasonable approach for a calculational tool.
We
- must, gg Lo however, be reasonable in our enpectations and not cloud the a
icsue with unnecessary exercises in statistics that are
[
irrelevant.
Conservatism should be added to our BE calculations j (3
by imposing requirements on decay heat and stored heat (the j
'd forcing functions).
Peaking factors and plant tech specs could
Sump Plugging By Fibcrglace-Tha impcct cf a higs.
Emergcncy tha becck appccro to bo moro covero than tho improcciona J
?
en rgy left me with.
I fcit that th9 cicplo equipment
' HDR testing it approach was inadequate and I am now convinced
- 3. Isurvivabilityentire area needs much more attent' ion.
Plugging the ic.
This cump appears to be minor compared to other problems that my o
)
rcruit from a high energy line break.
an excellent UHI and McQuire-The UHI licensing effort is
)
of why we need Best Estimate calculations.
There were cxcmple tiscs when UH1 looked detrimental to safety as a
result of App:ndix K EM calculations.
There were many discussions that in " darn it 1 know more water in the core will help."
The g
cndcd down ico condenser containment impacted on the LBLOCA by slowing core reflood rate.
To compensate, Westinghouse developed g
the through a similar exerci e and decided to stay KWU went UHI.
cwny f rom ice condenser containment but kept the UPI.
0 73,
,rgument, giy n ey ouu, po,,,
one w,,tingnou,e
,,c rcmoval of the UHI seem to be based on operational-aspects.
.I believe you remove a system that adds to safety because g, don't maintenance staff makes mistakes.
- Further, the concerns your Semiscale UTSG studies demonstrated cbout nitrogen are overdone.
""d
- "" "it" " '""** ""*' " " "
'"*" ~ "**
licensed.
This 0
to be the case when the UHI system was thought thould be revisited before the UHI is removed.
The sensitivity of the level in the UHI accumulator tank is The accumulator tank was laid on its side peculiar to McDuire.
roculting in a number of other concerns as well as the needed g
measurement sensitivity.
Laying the tank on its side was a lovel cictake and should not yield rationalizations for others to ccnsider removal of planned UHI.
Maybe Duke Power should g
compensate for the tank being on its side by installing pumps and rccoving the nitrogen pressure system.
p D
$)
-