ML20245F951
| ML20245F951 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 06/29/1989 |
| From: | Fitzpatrick R BROOKHAVEN NATIONAL LABORATORY |
| To: | Nilesh Chokshi NRC |
| Shared Package | |
| ML20245F930 | List: |
| References | |
| NUDOCS 8908150141 | |
| Download: ML20245F951 (26) | |
Text
. _ _ _
ENCLOSURE 4 BROOKHAVEN NATIONAL LABORATORY ASSOCIATED UNIVERSITIES, INC.
Upton. Long Island. New York 11973 (516) 282s Deoccment of Nuciear Energy FTS 666' 7204 June 29, 1989 Nilesh Chokshi U.S. Nuclear Regulatory Commission Mail Stop NL/S-372 Washington, DC 20555
Subject:
June 13-15 PRA Neeting With PG&E.
Dear Nilesh:
This letter is a follow-up to the subject PRA meeting held in San Francisco. As a result of this meeting, we believe there are three action items that need to be addressed in the short term.
The first action item was for BNL to provide a greater level of documentation into the specific details, assumptions etc. that we used to generate the "BNL2" results in Letter Report 04/Rev.1 on the Diablo Canyon Auxiliary Saltwater System. This information package is enclosed.
In order to expedite the dissemination of this material, I have, per your request, sent copies directly to R. Theirry at PG&E and to H. Rood, the NRC Project Manager.
The second action item deals with the reduced model of the internal events sequences. While reviewing aspects of these sequences during the meeting, a draft document was exhibited by PG&E. This draft document contained notes on at least some of the sequences detailing some of the idiosyncracies built into them.
This documentation would be extremely beneficial to have when we perform our leading sequente review.
If this information can be made available, BNL would need it soon.
The third action item relates to the ability of a single ASW pump to supply the full cooling needs of both units (under all modes) given the other three pumps are not available.
Our letter report requested an analysis to back up this assumption in the DCPRA. The pump capability was briefly discussed at the meeting but no analysis was presented. We feel this is an important item that is necessary to help complete our review of ASW.
Overall, the meeting provided significant new input which addressed many of our concerns on the following three systems: 1) solid state protection 8908150141 890807
@ ADOCK 05000275 pNU
q.-
' 1
-b 1
(Letter to'N. Chokshi'from R. Fitzpatrick pune 29, 1989
'Pago~2 system (ESFAS portion), 2) auxiliary feedwater system, and 3) auxiliary saltwater system. We are now reviewing this new input.
Sincerely,
,n NC R.G. Fitzpatrick, Group Leader Risk Evaluation Group RCF/dm/1 Enc.
cc:
R. Theirry w/cnc..
H. Rood a
'W.
Pratt w/o enc.
R. Eari
-W. Kato' i
nn -
ss s
UPDATED BNL SENSITIVITY CALCULATION ON THE DIABLO CANYON INITIATOR FREQUENCY, LOSW ss s
ss 4
In reviewing the calculation of the initiator frequency, " Total Loss of Auxiliary Saltwater LOSW," in the DCPRA a sensitivity calculation (denoted by "BNL2") was performed by BNL to see the impact on the initiator value of modifying some of the assumptions used in the DCPRA and including some features of the ASW system's operation and configurations apparently not modelled in the DCPRA. A condenced description of the modifications used in the BNL calculations and the obtained results were given in BNL Letter Report-1 04/Rev.1.
The purpose of this note is to provide some additional information about the details of this calculation as agreed upon during the June 13-15 meeting.
In contrast with the DCPRA's approach, the BNL2 sensitivity calculation assumed that a Unit 2 ASW train cannot provide (share) cooling water to the Unit 1 CCW heat exchangers if the other Unit 2 train is down due to 1) failure,- 2) scheduled maintenance (overhauls) during Unit 2 cold shutdown or refueling periods, or 3) when maintenance is performed on supporting equipment to a Unit 2 ASW train (e.g., unscheduled maintenances, demusseling or tests).
This assumption was prompted by the lack of convincing evidence (calculational or experience) that a Unit 2 ASW train can sufficiently share its flow with Unit 1 at any time upon demand, The. modified fault trees reflecting these assumptions are given in.
Notice the new basic event, COLD, which describes the scheduled outages (overhauls) and tests and the event MFDG which includes the unscheduled maintenances as well as the total duration of Unit 2 ASW train's demusseling.
l In addition to the above modifications, the BNL2 fault trees include the following:
Updated BNL. June 29, 1989
t N
a.
Failure ~of the Unit 1 train-to-train crosstie during the maintenance period of a Unit 1 pump (basic event: BKDC).
~
- b. Plugging of the travelling screen at Unit 1 (basic event: BKA;in spite of the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> attendance at the ASW, it was assumed that plugging may occur unnoticed, mainly during night periods; a representative time period.of 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> (per year) was taken during which screen plugging may occur).
Demusseling periods of the standby trains at Unit 1 (modified duration c.
of the unscheduled maintenance: MFDC).
The leading cutsets and their quantified values obtained by using the
. SETS-code are listed in Attachment 2.
The listing of the software used to quantify the basic events and the basic event "value-block" are given in Attachment 3.
The listing of the software includes comments that provide additional details about the quantification.
The present version of the calculation compared with that given in Letter Report-04/Rev.1 does not contain provisions for unavailability contributions due to 1) maintenance of the unit train-to-train valves (because, on reflection, it was judged that whether or not a given specific maintenance
)
activity.would involve valve isolation, it should be quickly recoverable) and 1
- 2) maintenance and deplugging duration of the travelling screen of Unit 2.
To avoid the possibility of double counting, the present version of the calculation also does not contain the basic event, BKA, in the fault tree i
l i
describing pump rotations.
l The sum of cutsets from the full pump rotation fault tree was multiplied by a factor of two, corresponding to the expected number of pump rotations during the total outage periods of Unit 2 ASW trains (about one month). The j
sum of the cutsets of the pump rotatien fault tree without the basic event, COLD, was multiplied by a factor of 24, the expected number of pump rotations when Unit 2 trains are operating.
l l
Updated BNL... June 29, 1989 l
l 1
q l
L._-___:-__._________-
p*,
v.
9="
l'~
The new results.obtained for'the initiator frequency are:
LOSW contribution during normal operation (total):
5.097-4/yr.
I LOSW contribution during pump rotation (total):
1.146 6/vr.
Total:
5.108-4/yr.
Thus, in Table 2.3 of the Letter Report-04/Rev 1, the set.of values associated with the LOSW initiator should be. replaced by the new set of values
.given below:
Top Event Case Calc.
TTL HW HWI HWD Comment LOSW ASI BNL2 5.108 4 5.108-4 4.581-4 5.270-5 Present BNL2 5.616-4 5.616-4 5.411-4 2.038-5 LR-04/Rev.1 BNL believes that these values still somewhat underestimate the LOSW initiator because the model does not take into account:
the dual unit common cause failure (plugging) of the traveling screens, a.
b, the expected higher failure rates of the ASW components due to the harsh environmental effects, and the probability that a Unit 2 ASW train will randomly fail in the time c.
window of one unscheduled pump maintenance preceding the complete loss of both Unit 1 ASW trains.
Updated ENL. June 29, 1989
= _ - _ _ _ _ _ _ _ _ _ _ _ _
s b
4 4
') #
-t e
a.
l-i ~
s.
/
l d
c 4
y y:
p 4
E y
/
i, ATTACHMENT 1
~
r s
u i
s y >,
s
/
/
's s.
m i
l
)
l 1
I
. Lrs w hc n 8 ArrscHHenT.1 r.
ii y _:
_ _. 4
.,g
._.._.__m.
u 2 ur u
- s.. - _..
..._ z..
t E -$
c d-m 1
u y
fr e e 8 CO o>
.N
' ('
iod I
6 t
p yu*
- J i
s
__.....__'__'___..___.j._._,.;._
..?
.,._.L_._..
_ C I
p'._...._.__._._._..__.-_-_____.__..__.._
+
e 2
u, g x
2 N
i i
_..._.._x..-._...._...._-.._._.-.--.-
_ gq
._ s gu :q:!._;--_..
c w:
a.
_. _. _ _ ~.., _.. _. -. -. _ _. _ _
h g g.
CQ H.C
~
a
__v*.*
p2 i
3* e.
.a g g L...,....
. 'i l
- C N $
wh t
- i. _......_._..
K
. m/
q *h <.
sw.
l
%'D
% v' (2 NR 2
a.
1 v~ p
+
46 d
1
.g 2 w
'I e +
.a'=
3wa
. $ n,- 1
_..u J s o
x 1
Su.
_n.
< b g__
g......_. _ 2 1
+
3
. _.d, j
g,,,,
.u d'
.. _ used r
<nu
... n.
a g ur uu a
o 1
- .a 3 3 [~fi i
t.
~
. C~
5 t[~ l
^
40 t
.db
.b.
y*
g c --a p g-y (
2 m
c4
.. o _ _.
3 o
g
__ ow (
/,
9st se i
g p
u.e n
o < <::t :.:
2y
'D u, D
=>
t)
Cd' o E tu c
t:
2
- 2. 2 *.:n e
yg O
c.,
< C. A W2._. _
.'3.0 o ei 5
6 to a
o 3
W*2 i
i V
365
- . g -_K g
1
(
q
.gV v.y a3.
g.: W f
~ ~~
v
- - sE l
gb Q,_
s~h-.
^ ~. y g
~
O.
.o p N-3
$h o 2.,n o 3 ;p 0 :r '2 e o-s
. ? % f.
$$w
" D<
h Iy
%p<,<
y O 'f.'L3 3$
h
@3t,
=v
< e m bb
- /. C C '
kl g ta p)
$ga Vd B n
g =d a e
v M <3 3
spr i
3 O
p can p a f.u
-a
.~ O m<.
--"d,'
4-)
ar.
a 0
- t A Cs
[
.l y
dub 4
E. D.2 Q.c.7 (w h s hENL)..._ =.,.N I
cr ek 4, 9
~
..