ML20023D602
| ML20023D602 | |
| Person / Time | |
|---|---|
| Site: | 05000000, Grand Gulf |
| Issue date: | 02/17/1981 |
| From: | Cesare G NRC |
| To: | Jeanette Curry NRC |
| Shared Package | |
| ML20023A436 | List: |
| References | |
| FOIA-83-123 NUDOCS 8305240260 | |
| Download: ML20023D602 (5) | |
Text
,
n w I.< 7l
~ ~ -
p i._,_.
.,..- a..-
=
b i
.A
..,. t s
'l J
k p,. s.
r
, so, ' W ",o,cca5iDX:-Bg..
?;. -
e A..~ '.'
- .s
~
e-r ---. :..r
~
.- {' I 5 .1 '
'. ' y(.bf3T"
... r.
V-
'~TO*,,n.,
> c.., : an.,,.n:::_:
q.X.~.z., e : -~v z.:*s.e:...c e W:s ; :? ='r
- 1 q- : - - ~.,,...: :*,. :' ',
~;. ; ;i +'.:
z..
r ; M
,v..,c. w..r.-Z 9. *s. '. A. ~'. ~ < ;v-t.-. -.
.4
-
- A. u. s. _
,f
.:..+t.W.-
6
%.,u,-.,,.3-....4..,.:..,_.
s :
.,.. m..,..
A,.
. e..<
.. 3g ;. v. ',.,.
w.
. w, I y..,,.,.
. w,.,., :
g'-~
a.
- ~.
-j,l [. -
. w.,,. a... - a
.e...
p n
_ ll*:- k -?
~ h_
~?
" h k ' h}.
NM -
.., [,
w....:
e
.. ' ' g..:.
... r,-
~,..
...s-f' - g.--
. ' =.
.. m'w.c.,.
1
,v
. s E
~
.- u.
...~:...s.-
--s:. g
-r.#.- 4.:-
.1
... ~
.,,m.,r.
.~
. (p ti 7.,.,-
g
,a,,.p k. -t.,.; ku
, Q.T~
u --.
,?
&, f
- m
'Q
~
(
g
,. A -
- . n c..,.
- %. :c;C'
=
J-
- ^
w
~,~"',-A.
i j. W p,..+
c,v..x m
'e.-.
q e.%
~,
L
,)
SMLe %
e
[.?
g
'S O'
_.c-
.'?. W; ': ~} [.~Q (w:>> f? '
~ ?*
.L
,.g..
,-o..
_,z..
4 se
..-m-7 3
s.> _-
..;;.... :.a -,., -
~
^; M~.J.-
Ak T;~-6t.
.,gn.
m-k w.
f.
- ..g l
- g wcansv Q w.
%EA M. A L-c
.s..
a..
i l
SHOLLY83-123 PDR e
r
.e
a h
0 e
o 6
%g M-
.~a~
., ~ - -
- m --- ; y ~~ q
...-,-.o-
....~..
p
,y.
y,
. n. o.,,
i, g
.+ s
- -,.'e
, e;,.
3 s
f.e '
4..
.. = i' 3-r* '; *.
- ~
s
, *y.
..g j.
,, /,*
A - _,
]
.--f hC Ekb
.h,.., Ob.
~ $
j...
<*'A.,,
- .,e, a.
't.
9: e^ r = ' MV
-.s on.,. s ~o 4.
l* e. 1, m, t...- e.,%
.'w',f..~
a-r r y, eR+%,s
,.jw
.s 9
r* *gQ.
f,f *K.**.e
- R', y A
c_ej f..%.,tf.f..g mm
%.C,. ? M S* *^ O">,Y*. s M ~,f.
,,...s,,,-
a
>? W ~&l**,~'*Y%*d:
- 3l*C.&l-?r'
.'. 2
- S!
,.. "N.MSC.# F WNj,
- e. %,s w, -:~_*;e,
, m...>v,,,... ;.. m..,,.,;7-5.. -
g,,~,...w......s.,n,r~.,py...~- - +, j w, p~w -
s
... -". V.
? >
. A. u...
-s, c
-w.<.,..*e.-
m, s.
- g... 4.
g s-g a. <,,.
,g g
4
- e.,
s g7 e
- .~.
m..
a.
.y*.. _. ~,
.f,.,-
. y
.. - a
~
~
'D t,...
,,.[ e.g~
- =',.. e*.
1*Q * = j.,,.
~..,..
g.
.~.
-,,,.,.'o','*,k.
m.
- d.** [ *...- '.....
"#*, (..)^ * *,. *** *
='
g "t
~
.g
- **I d. g."* *g. " t
-* *i ',
- g, 4
g
- r+'.'
^ ** y. -
.. - _j e,
.g f[. q, e
=,,.*' l s [ #. -ja
,, ',. ' ' +-r (
t.,
,,.5
. ~y = ; *.. *e
- r. A.
i
%'d.<. e:.*,.e 4
,*,+., -
. ~-
i
.w
..-0,.
.~
- m w.".
.s
-.., *.~~',g,-
ar
.e.
-
- c. *.a'm~1 v.
,, M ~7..,,g. 4 aa.a*. v
.. -+ -
4
..b m.
g
~ **'g"**-,.
(
y " e k;s, ]. m. ],g% W'"h;*r-*.g%
a..
- -84{f '.m E p y."'g'.\\, r M-.
t=.
p
, * 'g a "*,
- e. *.J',~. 3 i
d.. -.. - A,_ %o
~
a., s: #
L~
..n,,,._.-,,.,.m,..,.,c.- wew..m,,
.,.. m.
..t *r....
e* *.
e-
. '<-d.'
' y r, i,.* ;
e..
sp,
g m
.e d h
- Qq v,.,.
- :, (,.
' *% 4.;e,
- k. '" *
(ye.J '..-.y". * *.1. * ; 4 's A.
s
- a...,. ~, '
- s
.u..'
%*d.
g 7
- =.
6 I,* s.
=,~
.e. * -**
.e f..b
. s t
,.".p.
( 4,,e
.,, e
..,ee
- "' ~,
y,=~
- . +f
.x
/ Q t'm7.*f 3_ m (Q,;;~."q..,*
,1, b., n
- * ;i,
_y,.g -
- s. -
),,h O,,
t.'.i '.
/'e
- 4k
- 7 f:. ;,. ' -,, E ^% ;. ~,.
t
- ',',e_.*,s.~
', ~. :$ p
-=
66
- ['"'
[^
e,-[
- J,t.m, e-
,.6-
's
- i"I' -
[
4~
e g.
,-* *,..s. sa *.g.n.,. -
y
.s,
=,,
~
q.'.-
e*
g -
y e
g e
g O
a.
g g
.O"-
gP 7 '. g p
,y*
" 4 a'
,i
,,e, + -. m n
s6 t
b,4W -. -g',
,,O
,a.s..
a.., -
J
.*4
- 4
.s.
-0
,?4 2,,e 1,_(
9 6
% 'A'
.' g,4 A *.
.....r, y.
,i
. - 4
- S..
g.,
F.
g 4'
4
.;y
.,t%,~
r D,* '.,;..
.,y
-'A".".
e g
g
~
~._. ;
r b
O
,~
- 8 *%y*.,.
g'\\.
" - qF " ',
p.
6
- .4
+.-..
.g
' T
,.6,.
.,, ro g
,.,..,.'..?.
w
. r
. =,
a k
b e
t
- 4,.
g
-T.,
g J^
. j.
g;,.
s
~, -
~
>..e *.
.g. %, -,.
e
- .w
~
g r
==
- i-g
, O
,, 7 g
b I/*
,,./,,'.
. a.
~*T
- J
.,.o zsa,
,e
,% s
.., 1 e
+,
~,
a-
. 1..
. u.-N...
e i
a w 4
8 g
n 4
D
=
,8'
-.+
4
,b 141.LtJ11 iV:
SS._- Jim Cur y /;ee
(
).q,e 1 of 3
.ite:e FROM:
Sam H. Hobbo February 17, 1981
~
,r TRELIMINARY COMMENTS ON DRAFT VOLUME 4
/
NUREG/CR - 1569 s
RSSMAP CRAND GULF STUDY The Overall conclusion from a brief review is that the reported results misrepresent the risks associated with the Crand Gulf BWR/6 plant design.
Very conservative assumptions have been made in the analysis which lead to an unrealistic assesscent of the absolute risk of the Grand Culf plant and the relative difference in risks between the Crand Gulf plant and the WASH-1400 BWR baseline plant (Peach Bottom 2).
\\.s Although the report indicates several arens where the BWR/6 plant systems 4
q are more reifable than the WASH-1400 BWR baseline plant systems, Figures 6-1 and 6-2 indicate that the Grand Gulf plant has an overall core melt frequency
'7
[j which is a factor of 25 higher than Peach Bottom.
The primary reason for
, j':
this difference i,s the unrealistic assumption made for the accident sequence represented by a transient (T) followed by a safety / relief valve sticking
'r; 3,
open (P). The assumption gives no credit to the availability or restoration
'" /
of the Main Condenser (Power Conversion System) for the residual heat removal 3
function. The accident sequence frequency is increased by approximately two orders of magnitude using this assumption. The Main Condenser is recoverable during a transient event involving a stuck opea relief valve.
The probability t
of the Main Condenser being unavailable for a'long duration (2 to 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br />) during an accident nequence of this type has been estimated to be 6x10-3/
shutdown.
It is strongly recom. mended that the assumption associated with 9
the' transient-stuck open relief valve-loss of residual heat removal accident J'
)
sequence be reevaluated in light of more realistic and available information.
\\
The following are other specific comments from the subject review:
- 1.
A complete review of the impact of the success criteria used in the s
report was not possible due to time limitation.
Further review and contact with the report authors would be helpful in identifying and resolving other areas of concern or conflict.
[ 2.
The report assumes that failure of the Reactor Protestion System 1 cads to a core molt.
No credit is taken for manual boron injection /recircu-
,j il lation pump trip as was done in the WASH-1400 study. This nakes the ATWS core melt frequency for Grand Culf at least a factor of 5 higher than the WASH-1400 BWR baseline plant.
I 3.
On Page 3-12 of the report it is stated that the Crand Culf Residual Heat Removal System (RHRS) unavailability is assessed to be an order of 4
e i
(
ww,
l l
Preliminary Comm(ents on Draf t Volume 4 February 17e 1981
.+
magnitude higher than for Peach Bottom (WASH-1400).
This asnessment is not adequately substantiated, particularly since the WASH-1400 RHRS un-availability is dominated by operator error and a common mode in the service water system and not the hardward configuration.
4.
The report mades the assumption that loss of heat removal leads to a core melt.
While this assumption may be realistic for a BWR/4 (WASH-1400) plant, it is not for a BWR/6 (Grand Gulf) type plant, since the BWR/6 Emergency Core Cooling System pumps are rated for 212 F at 15 psi.
A better estimate is that only a fraction of the loss of heat removal events would actually result in a core melt for the BWR/6 Mark III plants.
5.
The CRD system discussed in the appendix on Reactor Protection System has an instrument volume separate from the scram discharge header; this is not the Grand Gulf design.
6.
The Event Tree Interrelationships are oversimplified, vague and don't reflect the total picture.
7.
The RHR unavailability numbers appear to be too high. particularly in view of the steam condensing made of operation available at Grand Gulf and not at Peach Bottom.
In addition, no credit was taken for use of heat exchanges in LPCI made.
8.
In several cases, theprobabilityofvalve$beingleft closed after maintenance is unusually high especially in light of the fact that the systems have to be tested prior to being returned to service and de-clared operable.
9.
In several cases, stat'e'ments, if taken out of context,'could have a very detrimental meaning, i.e., the statement " Grand Gulf System's 'has a failure probability of five times greater than Peach Bottom System A" on' Page 3. 2.
This cautionary intent of this statement is appropriate but the statement should be reworded.
10.
The reference list is missing from the report in several instances.
This hampered a meaningful review.
11.
Use of a single event tree to represent plant response to a LOCA may oversimplify the assessment and cause risk to be assessed unrealistically high.
12.
Considering the fundamental difficulties with event tree and fault tree modelling, the use of a " survey and analysis" technique to identify system failure modes which dominate risk is questionabic.
Further O
e
- .L.
L.
_ *-+
Telecopy to:
\\ - Jim Curry / Joe Martore Page 3 of 3 Preliminary Comments on Draft Volume 4 February 17, 1981 justification of this " systematic approach" is needed to defend that the approach is repeatable in independent studies and that the dominant risks will be fully identified.
How can the analyst be sure that "all" doubic or triple f tilure modes have been identified.
13.
Since tolerance bounds were not estimated, it should be recognized that large tolerance bounds would cast doubt on the assessment of relative risk (where the tolerance bounds could overlap).
14.
The use of Peach Bottom meteorology and population densitj data may be useful for comparing the Sandia team's assessment of BWR/6 safety to BWR/4 safety in the same circumstances, but it is not helpful in assesing the Grand Gulf risk.
15.
The statement that the dominant VSS failure-at Grand Culf is for steam to bypass the suppression pool through Icaks in the drywell penetrations does not seem plausibic.
Elaborate or revise.
16.
The statement that the Grand Gulf SSWS and the Peach Bot tom HPSWS un-availabilities are similar ceems implausible in view of the fact that the Peach Bottom HPSWS unavailability is dominated by potential faults in the single output v51ve.
17.
Last paragraph on page 4-4 is confusing. Reword to read:
Transient sequences with event P, i.e., a Failure of any open S/RV's to rescat, can be treated as a small LOCA since the systems responding to these LOCA's are identi-cal to those required for a small (S) LOCA.
These "tran-sient induced LOCA's" are therefore. transferred to the LOCA event tree upon event P.
18.
Concerning Section 4.1.4 Interfacing Systems LOCA, consider the follow-ing comments:
(a) Could the MOV's be reclosed?
(b)
Is the use of the ADS considered effective in mitigating the event?
19.
In general, it is not cicar whethat or not credit was given for system restoration or repair where the duration of sequences is long enough to allow for that.
_