ML20141C855
| ML20141C855 | |
| Person / Time | |
|---|---|
| Issue date: | 06/19/1997 |
| From: | Rossi C NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | Martin T, Sheron B, Zimmerman R NRC (Affiliation Not Assigned) |
| Shared Package | |
| ML19317C431 | List: |
| References | |
| NUDOCS 9706260053 | |
| Download: ML20141C855 (4) | |
Text
..
,{
f>R REQ g*
UNITED STATES
{
4
)
NUCLEAR REGULATORY COMMISSION f
WASHINGTON, D.C. 2066H001
- * **
- p
W
,n c,n.,n
~
4 June 19, 1997 MEMORANDUM TO:
Roy P. Zimmerman, Associate Director, NRR/ADPR Thomas T. Martin, Acting Associate Director, NRR/ADT Brian W. Sheron, Director, NRR/DE Frank P. Gillespie, Director, NRR/ DISP Bruce A. Boger, Director, NRR/DRCH Marylee M. Slosson, Acting Director, NRR/DRPM Gary M. Holahan, Director, NRR/DSSA Lawrence C. Shao, Director, RES/DET Bill M. Morris, Director, RES/DRA M. Wayne Hodges, Director, RES/ DST James T. Wiggins, Director, RGN-l/DRS Johns P. Jaudon, Director, RGN-II/DRS John A. Grobe, Acting Director, RGN-Ill/DRS Arthur T. Howell, Director, RGN-IV/DRS FROM:
Charles E. Rossi, Directo oI M
Safety Programs Division Office for Analysis and Evaluation of Operational Data
SUBJECT:
SPECIAL STUDY: FIRE EVENTS - FEEDBACK OF U.S.
OPERATING EXPERIENCE - FINAL REPORT Attached for your information is the final report (special study) on fire events covering operating experience from 1965 through 1994, This repoil characterizes the frequency and nature of fire event data from U.S. operating nuclear power plants and examines the potential impact this updated data could have on fire risk assessments.
I The report updates the SANDIA fire events database (1965 through mid-1985) with fire event data from Licensee Event Reports (LERs), the proprietary EPRI fire events database (1965-
/
1988), and fire-related component failure histories from the Nuclear Plant Reliability Data System (NPRDS) database (also proprietary). Appendix A - Tables I and 11, "Overall Fire
,g Events Data" contain proprietary fire event data from the EPRI fire events database (coded in h f
)
w n, the tables es "++"). The consolidated fire events database and other appendices developed for this study provide a comprehensive and up-to-date compilation of information on fire SW events, their calculated frequencies, and severity as they have affected U.S. nuclear power plants.
2 250083 8{
The study reviewed and compared plant location fire frequencies with those used in selected probabilistic risk assessments (PRAs), including the Control Room, the Cable Spreading Room, the Auxiliary Building (Pressurized Water Reactor plants only), the Reactor Building gh (Boil g W t Reactor plants pnly), the Switchgear Room, and the Turbine Building.
. S,wa erLW L% d!
v
. -n a
q-
, yN$m, e,wmum dA U2i ddAdiIU 7
e e
104 e1 FRf CEaliER COPY LM &> I I + pp op Owrm& =xr.
k N h d il k h
]
The final report includes resolution of comments from within the NRC and from industry (EPRI). As a result of resolution of these comments and follow-on discussions, the following significant changes were incorporated in the final report:
l Clarification of "overall fire events" definition and addition of a definition for " smoke events."
Addition of Smoke Events tables for the two periods (1965-1995 and 1986-1994),
including pie-charts for smoke event density, smoke event causes, and smoke event locations for information. No risk significance was found from the review of these smoke events.
i Expansion of the Power Operations Fire Events Severity Grouping from three categories to five categories. The two new categories were provided to: 1) group fire events where a SCRAM or Reactor Trip occurred, regardless of the fire's duration, but no loss of a safety-related train occurred (Category C) and 2) group fire events that resulted in the loss of one train of fire safe shutdown equipment, regardless of the fire's duration, but no loss of a safety-related train (Category D). The original Category C (balance of fire events) was reidentified as Category E.
Reidentification of fire events and determination of estimated fire frequencies occurring in the Auxiliary Building (PWR only) and the Reactor Building (BWR only).
Modifying fire duration estimates to be within equal 5 minute intervals, with the maximum duration used in the report remaining at 100 minutes. These duration estimates are intended as qualitative information of the apportionment at plant locations during power operations and shutdown.
l This report identified the following major findings and conclusions:
i A comparison of fire events in the pre-Appendix R period (1965-1985) with fire events in the subsequent period shows that event frequencies have declined slightly, while the safety significance of events has also been lower. The most significant fire event occurred at Browns Ferry in March,1975 and was a pivotalincident in the recognition of fire safety concems. It resulted in a scram and propagated without supprassion to affect multiple
^
redundant trains of safety equipment. Since the implementation of Appendix R modifications and other industry activities (1986-1994), there were no fire events with similar sa:ety significance. There were only two fire events resulting in a scram and loss of one safety related train or loss of offsite power (LOOP) during this period compared to 10 events previously. Other fires have been severe in terms of the magnitude and duration of combustion (such as some turbine building fires), but their severity in terms of challenges to safety systems operation has been limited. However, such fires could be important if redundant safety trains or decay heat removal systems were dependant on equipment located there.
iL iiJ 1
... _ ~.
4 4
PROPRiflARYINFORMATION
' 4 The fire durations during power operations were generally short (less than 10 minutes). The information available on these short duration f!res was not sufficient to evaluate probability of fire detection and suppression used in recent PRAs.
c The fire durations during shutdown were also generally short (less than 10 minutes).
Shutdown durations in plant locations that contain systems necessary for decay heat removal during shutdown, were the same or lower than fire durations for the same plant locations during power operations.
I A sensitivity study, based solely on changes to the initiator frequencies, did not indicate the potential for substantial changes to the overali CDF due to fires. Other aspects of fire analyses may be more critical to their risk assessments, including: the mechanics of L
combustion, combustible loading, and means of detection and suppression. The data in this report was not suitable for addressing these issues.
i l
For the 1986-1994 period, the shutdown fire frequencies varied in comparison to the fire i
frequencies at power for most risk significant plant locations used in PRAs. Since some plant locations were higher (Containment, Reactor Building, Auxiliary Building, Switchgear Room, and Diesel Generator Building) and no PRAs were available that addressed fires during 1
shutdown, a more detailed review of shutdown fire events was performed. This review indicated the following:
)
Containment fires were predominantly caused by welding operations and did not affect j
There were a limited number of fire events that affected the functional operability of Residual Heat Removal (RHR), Decay Heat Removal (DHR), and Emergency Diesel 1
Generator (EDG) system trains.
Therefore, the operating experience indicates that the frequency and duration of shutdown fire events appears to be similar or less significant than for fire events occurring at power operation. This finding is somewhat tentative considering the limitation in treatment of fires in j
currently available shutdown risk assessments.
Although archival data is retrievable from the NPRDS database, no new nuclear plant failure I
history data is available through the NPRDS database after 12/31/96. It is not clear whether j
there will be an industry initiative to replace it with another data source that would be useful in compiling fire event and smoke event data.
w, 7 7.v F M n m
' Q g
- -m
4 N$NIEbb!!! dNhbNN$N$N l :
i If you have any technical questions on this material, please contact James R. Houghton of my staff at 415-6353.
+
i
Attachment:
As stated THIS MEMORANDUM IS NOT PROPRIETARY WHEN DETACHED FROM TABLES I AND ll OF PROPRIETARY APPENDIX A l
cc w/ attachment of report, including nonproprietary Appendix A instead of proprietary Appendix A:
V. K. Chexal, EPRI Pat Lewis, INPO Debbie W. Queener, NOAC i
l Distribution w/att.:
~ NRC PDR (report with nonproprietary Appendix A in lieu of proprietary Appendix A)
File Center DHickrnan REmrit,RES JChen,RES SPD RF JRosenthal RSavio, ACRS JFlack,NRR l
Dross AChaffee,NRR EChelliah,RES SWest,NRR FCongel RWessman,NRR PMadden,NRR LMarsh,NRR KRaglan MMayfield,RES RBorchardt,NRR JMurphy,RES JRosenthat EJordan,DEDE JMitchell,OEDO PKadambi,RES T0 Martin, OED0 DOCUMENT NAME: H:\\FIRESTAF.LTR l
i To receive e copy of this document, indicate in the box "C" copy w/o attach /enci "E" copy w/ attach / encl "N" no copy OFFICE RRAS/RRAB g
RRAQ/RRAB /
RRAB f/
SPD 6
NAME JHoughton C [ N SMaysY PBarank CERo@A DATE 6 /13/9 7
/// \\
hh97 fff97 W
Y w o u-w w w m....s