ML20023A435
| ML20023A435 | |
| Person / Time | |
|---|---|
| Site: | 05000000, Grand Gulf |
| Issue date: | 03/24/1981 |
| From: | Hodges W Office of Nuclear Reactor Regulation |
| To: | Jeanette Curry NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| Shared Package | |
| ML20023A436 | List: |
| References | |
| FOIA-83-123 NUDOCS 8103300276 | |
| Download: ML20023A435 (3) | |
Text
fn
(
(
f UNIT ED STATES p ')-'
,'g
$.-}
, 'v.
NUCLEAR REGULATORY COMIMSSION e
o E
v.armotos, o c. 20sss t.t-
\\.,$. *j "AR 2 4 1981 MEMORAt4DUM FOR: Jim Curry, Systems Analysis Branch, Division /
of Systems and Reliability Research, RES THRU:
J. Mazetis, Section Leader, Section C, Reactor lo Systems Branch, DSI, HRR FROM:
Wayne liodges, Reactor Systems Branch, DSI,fiRR
SUBJECT:
COMMENTS ON DRAFT OF "THE REACTOR SAFETY STUDY METHODOLOGY APPLICATI0 tis PROGRAM RESULTS FOR THE GRAND GULF #1 BWR POWER PLANT" I have performed a cursory review of the RSSMAP report on Grand Gulf and find it to be an interesting comparison with the Reactor Safety Study review of Peach Bottom.
However, there are some omissions and simplifi-cations in the report which result in a severe overstatement of the risk of operating this plant.
Although you and the authors of the report are aware of these ommissions and simplifications and appreciate the conservatisms they introduce, the general public may not be willing or capabie of cor:sidering the magnitude or impact of those conservatisms.
In tr.e RSSMAP study, the dominent sequence lea $ing to core melt and containment failure is a loss of feedwater event with offsite power available, a stuck open relief valve and failure of the RHR system to givenforthissequenceis8x10golwithin20 hours.
remove heat from the suppression p The probability I think we all would agree that if this probability value is realistic, then the Grand Gulf plant and other plants of similar design would be of concern; however, fog the reaso n.
listed below, I consider the probability value of 8 x 10- not to be realistic, and thereby misleading as to the actual risk:
1.
No consideration is given to the main condenser as a heat sink, although it is the preferred heat sink.
The authors of the report estimate that inclusion of the condenser as a heat-sink would reduce the probability by two orders of magnitude.
2.
No credit is given for non-safety grade equipment (e.g. condensate pumps) which may be used to supply large quantities of makeup water to the core and the suppression pool.
3.
The RCIC is safety grade but not part of the ECCS.
No consideration is given to using the RCIC system to pump water to the suppression pool.
XA Copy _Hd{3een sent to PDR 3
1 le>
M l0
(
(
f*
7
~2-fdAR 2 4 1331 A.
Water may be fed from the fuel pool to the suppression pool by gravity J
feed, but no consideration is given to this mode of supply water to the suppression pool.
,/
5.
The emergency procedure guidelines which are being used to develop emergenry procedures for Grand Gulf and other BWRs call for venting the containment before allowing the containment to fail.
Ilo consideration
)
is given to such action.
t 6.
As a last resort, water can be supplied from the ultimate heat sink directly to the core suppression pool.
Tio consideration is given to this source of water.
In addition to the omissions which I have listed for the dominant sequence, there are several other questionable aspects of the report which impact other sequencts.
These include:
1.
The LOCA model considers only two types of LOCA; large break and small break.
There is no distinction between a steam break and a water break.
(The depressurization rate, inventory loss and systems required to mitigate are significantly different for the steam break and the vater break).
For example, sequences involving a stuck open relief valve are treated as small LOCAs and the ADS is assumed to be required to depressurize the system.
However, a stuck open 4
relief valve will depressurize the system without ADS as has been dcmonstrated at operating plants.
,O 2.
The interfacing systems LOCA sequence assumes the failure of a check
/alve in one RHR line and the opening of a normally closed MOV for testing.
fio consideration is given to the fact that leak testing is d1 one at reduced pressure.
Interlocks prevent the opening of these valves at high pressure. Also, there is no consideration of isolating the LOCA and no apparent credit for cooling from the core spray system.
Because of the anticipated public reaction to failure probability numbers which I consider to be everly-pessimistic and misleading, I recomnend that the RSSMAP report for Grand Gulf not be published until it is expanded to include most, if not all, of the systems which are available for cooling the cort and/or containment.
Any study of this kind should provide a best judgement of realistic probability numbers; not overly " conservative" numbers which create the potential for unjustified public concern.
Y Wayne liodges Reactor Systems Branch Division of Systems Integration, NRR cc:
See page 3
k
(
ii:'F 2 4 1931
+
Jim Curry -
cc:
R. Bernero F. Rowsome D. Ross P. Check T. Speis J. fiazetis B. Sheron V. Panciera W. Hodges e
W 4
O e
e