ML20087G562
| ML20087G562 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 03/12/1984 |
| From: | Scott Freeman TENNESSEE VALLEY AUTHORITY |
| To: | Markey E HOUSE OF REP., INTERIOR & INSULAR AFFAIRS |
| References | |
| NUDOCS 8403200059 | |
| Download: ML20087G562 (2) | |
Text
-
w wt..naremq P
~
TENNESSEE VALLEY AUTHORITY
!"Z KNOXVILLE. TENNESSEE 2 '
-,-i 00LKETED y
"SW C
'====!
.....r 1
- OFFICE OF THE BOARD OF DIRECrORS March 12 1984 84 MR 19 A10d5 v.
Y A","
7 ;i C+ 5i..F.t.i.-
=..
a DCU-ilyg & St.P u -
BRANCH
........-y
~
The Honorable Edward J. Markey
_1 E-; Chairman..
~"
' { Subconnitt' e.on Oversight and
~.7. g,... f
' " ' T.[
e
=
--- Investigations v'
Committee.on Interior and
~ " ' ' ' '
M. M d
-r Insular. Affairs
=
- .=
- - U.S. Bossi of Representiatives p. M 9f
~ " "
~Washingtoni D.C.
20515'
~
~T r
~-'. Deur Mr...ZCh. airman:
~
~
~
ThankyouforyourresponsktomyFebruary 27, 1984 letter. 'Ve appreciate your concerns' on this issue and welcome the opportunity to provide addi-1 tional perspective on TVA initiatives in probabilistic risk assessment (PRA).
At abou[th'e time ~ Browns Ferry was selected as an Interim Reliability Evalu-
...]
~
ation Program- (IREP) study plant, TVA decided to perform its own risk asseser-ment study of Browns Ferry. TVA recognized the use of risk assessment tech-niques in the nuclear. safety decisionmaking process was likely to increase n
and thatf as a nuclear industry leader, we should develop capabilities for
~
these assessments. In addition, we recognized an independent study would provide an. important comparison basis of the IREP program.
- =. -
14t:1me. assure you that TVA did not ignore the NRC's IREP report on Browns
,a
- Ferry.. R were of course.conestned abeut its apparent findings and TVA -
m engineers.thus conducted an informal review of the Browns Ferry IREP report
==. =is
)
(NUREG CA/2802) when it was issued in July 1982. Our review revealed some
' REC'D Cstasificant variations between the plant design capabilities asssumed in
' ~ ~
yhe report.and the actual plant design. Ihese variations would substantial-
~" ]
~_,
ly alter the IREP results. For example, IREP did not consider the cross-4 WR 84 3Je between the decay heat removal systems on the three units at Browns
.x
.7erry, and it erroneously assumed that failure of a minimua flow valve to operate would result in a system failure.
Our review was sufficient to R
convince us that had the report taken into account the actual plant design capabilities and configuration, it would have been able to conclude that
-" ~~~
Browns Ferry would meet the' interim secondary safety goal on core-melt
- " " ~
frequency'.7 Engineers involved in the TVA study did not do a detailad.re-
~
~
~"
view of thi IREP study to ensure we did not compromise the independence
[.....
of our own study. The TVA study will be completed later this year. We
[-..
believe the comparison of these TVA and IREP studies will be a valuable Lm step in the, continued development of PRA tech.iques.
e 8403200059 840312 PDR ADOCK 05000259 e...._.-
O PDR gg
- "" q An Equal Opportunity Employer
]
mug m.=._
,,m...
.m:we l
=
~
V N
C The Honorable Edward J.. Markey 2-r.._.
O With regard to the incident at Browns Ferry on February 14, 1984, the shutdown croling valve which' failed to open is part of a nonsafety operaticg mode cf the Residual Heat Removal System.
Safety analyses for the plant assume this func-e tion always fails for the purposes of accident evaluation.
In fact, tiis valve has design features which will prevent it from being opened if an accident con-
[
dition exists to ensure the shutdown cooling mode will not interfere with safety cystems. Shutdown cooling is one of the normal methods for decay heat removal.
The other method, the main condenser, remained operable throughout the event, allcwing the plant.to be brought to a cold shutdown condition. The IREP study W
r2 quires failure of both normal systems plus the redundant, safety-grade resid-H ual heat temoval systems as a precursor to a core melt.
In light of these
[""2"
[
fccts, we believe it is apparent why TVA does not consider any changes to the s
plant or its operation necessary in response to the February 14 event.
L I gather that we both believe that probabilistic risk assessment must be used with caution.
It can become a valuable tool in performing neelear plant safety evaluations because such assessments can show where improvements are needed and can assist in evaluating the cost and the safety. benefit of proposed improve-
'" "[.
ments. However, PRA cannot supplant the traditional methods for safety analy-cis which use the knowledge and judgment of experienced specialists in nuclear ecfety analysis. Also, there is dangar if too much credibility is given to h=
the accuracy of specific numbers which result from PRA. The TVA program is cimed at providing a useful tool to improve safety analysis methods and to improve the safety of our nuclear plants without putting undue emphasis on the accuracy of specific numbers.
N Wa believe that Browns Ferry is being operated well within safety requirements today and believe NRC agrees with this conclusion.
But we are not satisfied and we are making an all-out effort to reduce the number of violations and risks. TVA is determined to be a leader in the advancement of safety at its nuclear facilities. We welcome the opportunity to provide input to your dio-cussions on this issue and would be glad to answer any additional questions
]
y:u may have.
m..._
L<
Sincerely
{
w==n
' 1==i ammt S. David Freeman
{i{
l Director TheHonorableNunzioJ.Palladino[
l cc:
Chairman U.S. Nuclear Regulatory Commission 1717 M Street, NW.
]
l Washington, D.C.
20555 22.n v.
wan
$5 prin W
- y
. 5F5G" ila:7.
... cnaam;,;6_J"E. iG Eu s unumuid u a "I
7~ 7N
. : 7-""
v=. e "w
m w
..r--