ML19347F022
| ML19347F022 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 05/05/1981 |
| From: | Tedesco R Office of Nuclear Reactor Regulation |
| To: | Hanauer S, Mattson R, Murley R Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8105150099 | |
| Download: ML19347F022 (8) | |
Text
_ _ _ __ _ - __
h.
(
h 7
M.
g DISTRIBUTION NS SY 5198j LB#4 Reading EAdensam Docket able n..desco
~
RHVoll.ner DEisenhut/RAPurple MEMORANDUM FOR:
S. Hanauer. Director Division of Human Factors safety C[?
R. Mattson, Director Division of Systems Integration
//
b R. Nrley, Director
/'/4r y
O 44 P g/A -
Division of Safety Technology m.
i FROM:
R. Tedesco, Assistant Director for Licensing s.
Division of Licensing e
SUBJECT:
SEQUOYAH SSER input ON RSSMAP STUDY Enclosed is the proposed input for the sequoyah SSER on the RSSMAP Report.
The report was developed primarily upon the rer.ponse from TVA and the myiew by DST. Your coments and review are required by C.O.B., Friday, May 8,1981, and should include such matters as:-
1.
adequacy that the Evaluation as presented by the six recomendations is sufficient and that no other dominant effects appear to be at issue; and i
2.
some indication of the next step in the evaluation program on a longer tenn basis unless we do not intend to do any further work on the report i
et al (we could indicate that this action would be deferred pending receipt of the TVA risk assessment).
Odshubesned by Bohe*L, Tederco Robert L. Tedesco, Assistant Director for Licensing Division of Licensing nY 0${0f As tate
- "' C ' >1
.DL%;a..
' DL:.LB#4...
DL:LB#4..
' "'^"'> CStahlb ;.lb.
.EAdansam.,
.RLTe3esco..
5/
fai s/ /81
,, 3,/81 s/
sac reau aie oc sce nacu c:4o OFFICIAL RECORD COPY
' *o ' *-32+e 810M Rka
24.0 Reactor Safe *.y Study Methodology Applications Program In February 1981,' the Sandia National Laboratories, issued the results of a study, entitled, Reactor Safety Study Methodology Applications Program; Sequoyah #1 PWR Power Plant, (NUREG/CR-1659 Volume 1).
This report is the first of several reports to be published by Sandia on the results of analysis performed in the Reactor Safety Study Methodology Application Program (RSSMAP).
RSSPMP analysis utilizes the methodology developed in the Reactor Safety Study to identify the accident sequences that dominate accident risk for a variety j
of light water reactor power plants representative of the current population of plants.
The Sequoyah analysis by Sandia was conducted primarily with information available from the early versions of the Sequoyah Final Safety Analysis Report, Technical Specifications and selected plant procedures.
It is acknowledged that a substantial portion of the Sand', report was written prior to the plant modifications and procedural changes which have been implemented at Sequoyah as a result of the TMI-2 accident. Also, it is acknowledged that the analysis in the report was not upgraded, to reflect the new efforts in plant reliability analysis by both the nuclear industry and the Nuclear Regulatory Commission.
Neverthaless, the NRC staff, including the Division of Risk Analysis, RES, has concluded that while more rigorous methodologies may provide more detailed information on the causes of system failures, the most significant conclusions of the Sandia RSSMAP study of Sequoyah Unit No. I are accurately sunaarized s
in section 4.2 Risk Comparison and Conclusions.
The conclusions stated are listed below:
1.0 An important accident sequence occurring for the Sequoyah plant results from the potential for blockage or closure of the drains between the upper and lower compartments. This causes a common-mode failure of the ECRS and CSRS when the sump runs dry (sequences S HF and S HF). The 1
2 probability of these sequences could be reduced by improved checking procedures and improved fault detection capabilities.
5-1 Sequoyah SSER#5/ Stable 1/K
-4
,.,,,.-,---,-,-.--w,,,
+y,
2.0 Failure of the ECRS alone caused by component failures other than the drains also results in some important accident sequences.
[..
. ~..~
.~
3.0 Sequence V,,which'. valve failures causes the high pressure primary coolant to fail the low pressure piping outside containment, remains'on important sequence for Sequoyah.
This sequence could be improved by a more strategic testing procedure of the check valves over the limited testing capability which now exists.
4.0 Unlike larger containments, core melting caused by failure of ECIS or ECRSfaiithelongerpressure,smallericecondensercontainmentby overpressure even though the containment cooling system continues to operate properly.
The analysis of accident processes by Battelle Columbus Laboratories ravealed that the smaller containment pressure and volume design would not withstand the pressure exerted by the nonconde'nsible gases generated in the core meltdown accidents.
(This result was similar to the RSS findings for the RSS BWR design.)
5.0 Sequence TMLB' 9, which was important for the Surry plant as analyzed in the RSS, does not appear to be as significant to risk for Sequoyah due to the lower unavailability of on-site ac power.
6.0 Failure of the containment cooling system causing core meltdown following a small LOCA (the S C sequence in the RSS) does not appear to lead to 2
core meltdown at Sequoyah due to the difference in sump water temperature at the time containment failure.
On April 24, 1981, TVA responded to the Sandia RSSMAP.eport which is provided
)
in its entirety in Appendix F.
%t &
The staff concludes that TVA's responses are satisfactory.--- 5::7 P e remedial 2
measures hase 5--- t i n for Sequoyah in response the conclusions of this p
I report c-N s. k by D v
b.isect y.%tSt) 5-2 Sequoyah SSER#5/Stahle 1/K
--.w-
9 In particular, responses to Sandia conclusions 1 and 3 have resulted in stringent Sequoyah Technical Spccifications requirements.
Technical Specification for the potential for blockage or closure of the drains between the upper and lower compartments (conclusion 1), are:
Each refueling canal drain shall be demonstrated OPERABLE.
Prior to increasing the Reactor Coolant System temperature above 200*F a.
after each partial or complete filling of the canal with water by verifying that the. plug is removed from the drain line and that the drain is not obstructe'd by debris, and b.
At least once per 92 days by verifying, through a visual inspection, that the plug is removed and there is no debris that could obstruct the drain.
For Sequenca V (conclusion 3) whereby valve failures causes the high pressure primary coolant to fail the low pressure piping outside containment, the Technical Specification limits the leakage to 1.0 gpm from any RCS pressure isolation valve specified in a table.
Surveillance requirements forthese valves are that each valve shall be demonstrated operable by verifying leakage to be within itslimit:
a.
At least once per 18 months.
b.
Prior to entering MODE 2 whenever the plant has been in COLD SHllTDOWN for 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> or more and if leakage testing has not been performed in the previous 9 months.
c.
Prior to returning the valve to service following maintenance, repair or l
replacement work on the valve.
d.
Within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following valve actuation due to automatic or manual action or flow through the valve.
l l
5-3 Sequoyah SSER#5/ Stable 1/K
-v--9
-*7w-
No other issues in the Sandia report warrant discussion in the Sequoyah supplementary safety evaluation report.
l 5-4 Sequoyah SSER#5/Stahle 1/K
Appendix F TVA RESPONSE TO SANDIA RSSMAP REPORT TVA has reviewed the subject document and agrees with the final conclusion of the report that the public risk associated with a nuclear power facility having an ice condenser containment can be expected to be similar to the public risk associated with a similar facility having a dry containment.
TVA stresses the fact that a substantial portion of the report has written prior to the many plant modifications arid administrative changes which have been implemented as a result of TMI-2.
The final report was not revised to account for these changes.
In general, both the draft and the final report do not reflect the latest in:
(1) Plant design (2) Risk assessment technology (3) Human factors technology (1) Plant Desian - Since the time of the draft report, plant modifications and administrative changes ha.ve been implemented which should reduce the system failure probabilities as stated in the draft report. This, in turn, could potentially change the dominant sequences as stated in the draft.
(2) Risk Assessment Technology - The report examined the differences between the Surry and Sequoyah plants and then qualitatively determined whether the Sequoyah system designs provided improved or degraded system reliability as compared to the counterpart Surry system. The system probabilities of failure were then adjusted accordingly and the magnitude of the change was based primarily on. engineering judgement.
These facts should be kept in mind wnen reviewing the report and the quantitative conclusions should not be taken to represent a true indication of the plant risk, even though the methodology is demonstrated.
TVA has in progress a full scale nuclear plant safety and availability analysis being performed by Kasan Sciences Corporation (KSC).
The study i
i 1
Sequoyah SSER#5/Stahle 1/L 9
e-
,r f..,,
..,ww--------,.p p,
y.--,,,g,,,oye,%,-wgyw.y.m-
ein-
++w m,
+*.
.'+o
-e....<e 1mm TEST TARGET (MT-3) f 1.0 s a E4
- 5E E u \\l52 MB
~
}
I.8 1.25 1.4 1.6
/
4 6"
4' *#q sp 'OO LO
'*%f>s,,,,,,
43<gt,$
ov77777 i
i.
3 l
d>
'4 '$'
/
TEST TARGET-(MT-3)
O l.0 5 Lu L24 y"3 IE 1
1.1 SJ.* llE l '"
l
.8 I__
l.25 IA 1.6 l
=
6"
=
l l
I//
l&
l#4 4$ 4
- kf,,,,h,#'
gy%h 5
)3 4
1
is enected to be completed at the end of 1981.
The TVA/KSC study is being funded by EPRI and is being performed using the GO probabilistic methodology, whereas the Sandia report based its analyses on fault tree methodology.
The GO methodology is a success-oriented system analysis which uses standardized logic operators to model the particular plant systems.
A more accurate plant design is being modeled by the TVA/KSC study and it is expected to result in a different set of dominant sequences from those indicated in the Sandia report (3) Human Fahter Technology - Since TMI-2, the area of human factors has gained promir.ence resultin in significant advances in knowledge.
The report does not reflect this current knowledge.
Rather, it relies heavily on human factor data as supplied in WASH-1400, which in many cases was a coniervative estimate at best.
Since human error frequently represents the dominant factor in the overall failure of a system, TVA feels that application of current technology will have significant impact of the failure probabiitties and could potentially impact the dominaat sequences.
TVA reviewed the criginal draft report and has taken steps to correct the potential deficiencies it identified in its conclusions. We have also reviewed the conclusions of the final report and have noted that its conclusions are essentially the same as those found in the original draft.
The Sandia report resulted in six conclusions deemed significant enough to warrant specific identification:
(A) Two conclusions described common mode failures of two systems; (B) One conclusion described significant independent failures of. heat removal systems; (C) One conclusion expressed concern over the ability of the ice condenser containment to withstand certain events due to generation of noncondensibles; and (D) Two failure sequences which were dominant contributors to risk fcr the Surry plant
. were found not to be significant contributors for Sequoyah.
(A) Common Mode Failures of ECRS and CSRS and Check Valve Failures - The
- potential common made failure of the blockage or closure of the l
l 2
Sequoyah SSER#5/ Stable 1/L L
drains identified by the draft and final report for Sequoyah was recognized by TVA as items which required specific attention. As such, TVA decided to minimize this potential failure mode and provided administrative measures which would given assurance that the drain plugs will not be closed or blocked.
TVA decided n 1977 to minimize the potential for check valve failure by more testing.
Sequoyah Technical Specifications were written to include new surveillance requirements.
(B) Independent Failures of Recirculation System - The dominant failures forJthis sften are represented primarily by human errors.
TVA feels that if present human factors data and technology are applied, these failures may be reduced and they may no longer be dominant. failures.
(C) Ice Condenser Containment - Since the draft report was written, TVA has taken steps to mitigate the effect of certain noncondens bles, such as hydrogen, in a degraded core event.
The ultimate capability of the containment has also been found to exceed the design value by a large margin. Accordingly, the dominant sequences, as identified in the report, may be ef.fected.
(D) AC Power Availability and Containment Cooling System - TVA agrees with the conclusion of the report that the sequence of loss of electric power with subsequent loss of normal and emergency feedwater and the sequence of a small LOCA followed by failure of the containment cooling system are not significant contributions to risk at Sequoyah.
l I
+
3 Sequoyah SSER#5/ Stable 1/L 4
y, e --