ML20127M191
| ML20127M191 | |
| Person / Time | |
|---|---|
| Issue date: | 05/16/1985 |
| From: | Catton I Advisory Committee on Reactor Safeguards |
| To: | Boehnert P Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-CT-1799, NUDOCS 8505230205 | |
| Download: ML20127M191 (3) | |
Text
,
$[* Nff e
s,.
7DR66\\38$
j\\p Paul Boehnert FROM:
" Ivan Catton
SUBJECT:
ACRS DHRS Subcommittee Meeting Albucuercue. 22 January 1985 I
Whether of not better DHR is cost beneficial has been addressed by noth NRC contractors and the industry.
Every study concludes that a
cecicateo DHR system will yield significant risk recuction.
Risk reduction numbers range from 10 to 250.
It seems to me tnat the question is one of relating DHR reliability to risk recuction and oeciding whether or not enhanced relianility
- 1) is needed or 2) can be achieved if needed in lieu of a cedicateo DHR system.
how reliable a DHR system needs to be, To make a cecision acout secarate other causes and uncertainties from core one must first melt so that DnR recuirements will be visible.
This was where the Cave /Kastencere wark came into play.
Although I disagreed witn now they solit the risk in assigning a value to the DHR l
aooroach was sound.
I was very disapoointed reliability, the I
wnen tnis effort was out in a bad light.
It is one of the real 1
meaninpf ul contributions of the A-45 effort.
l lei In early aiscussions with
- Cave, he indicated that a
-d unavailabilty was needed.
In tnat a 10 unavailability of a DHR system is unoelievable, this leads one to conclude two are neeced.
He subsecuently shifted the share of risk and the 1EI number got larger.
Nevertheless, the accroach was simole, direct The staf f snould be encouraged to re-visit the ano meaningful.
DLQiAND ORIGINAL 8505230205 850516 PDR ACRS CT-1799 PDR Certified By_ 3fG N
1 l
. -.. an. ~
p' Cave amoroacn.
Studies of two olants.
point Beach and Quad Cities, will be done by February 1985.
The analysis outlined to us has a great deal of uncertainty as a result of not being able to model various chenomena in enough detail.
It is my view that the resulting uncertainties will be so large that further olant studies may not be too meaningful.
Study of cifferent plant types will yield information about the sensitivity to DHR system reliability but plant to olant comoarisons will not be oossible.
Many such studies have alreacy done so.
I don' t believe more are needed.
I can sucoly references to such studies if you don't already have them.
The studies being done for A-45 should ston at DHR reliability and impact on core melt frecuency.
One could then use the TMI-2 event.
our only real datum point, total cost as measure for cost / benefit analysis.
Tne fire analysis cone by SNL is an examole of how uncertainty Tne modeling was crude and tne criteria for failure creeps in.
may be on the wrong side of conservative.
There are calculational tools available that are better than COMpBURN K. T.
Yang develooed by experts on fire such as Yogesh Jaluria, and otners.
I nave no way of telling how this impacts the bottom line that will be used for cost / benefit assessments.
Tnere are other examples brought out by members of the subcommittee that emphasize the point I am trying to make.
.... ~. _ _ _.. _
l
, g..
I t@uld nave thougnt we would have heard more about the heat flow path from the DHR system to the ultimate heat sink.
The service water system and how it interfaces with DHR
- systems, component water cooling.
oil coolers etc.
is an essential element of guaranteec DHR.
A plant I visited uses mobile fans plugged into wall outlets to cool tne service water cumps during the summer to evoid overheating.
This must imply something about their DHR system heat flow cath.
Otner cuestions one might raise deal with (a
oil cooler ano filter comminations feeding a number of pumps occassion).
cause of common moce failure of all SI oumps on one I bring these ascects out to try ano demonstrate that the A-45 in
-deoth analysis may lack substance in certain areas.
l l
i n