ML20196J060
| ML20196J060 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom, Indian Point, FitzPatrick  |
| Issue date: | 11/04/1987 |
| From: | Specter H POWER AUTHORITY OF THE STATE OF NEW YORK (NEW YORK |
| To: | Ryder C NRC |
| Shared Package | |
| ML20195G598 | List: |
| References | |
| RTR-NUREG-1150 NUDOCS 8803140139 | |
| Download: ML20196J060 (22) | |
Text
_. __. __ _-.
- 123 Mcin Street Whte Rasis. New YcA 10001 914 681.6200 00 CME TED USNRC M G F ?SOCE0cREsna i
- b NewYorkPower 4v Authority r "" :
.a November 4, 1987 Mr. Christoper Ryder U.
S.
Nuclear Regulatory Commission Mail Stop NL-005 Washington, D.
C.
20555
Dear Mr. Ryder,
Enclosed for your use are:
(1)
An errata sheet for the more important errors in our September 28, 1987 submittal on NUREG-ll50, (2)
Retypad pages, where these errors have been corrected, (3)
A draft version of an executive summary that we are working on.
There were a number of minor typographical errors in our original submittal, whicn have been corrected in a master copy at the Power Authority.
If you wish, this revision could be sent to you as well.
We are working on the bibliographic summaries you requested.
This will be forwarded to you as soon as possible.
Sincerely, f'
.2 M Herschel Specter Technical Advisor to the Executive Vice President 8803140139 071104 PDR NUREG PDR 1150 C m
.~
t.
e ERRATA SHEET COMMENTS ON NUREG-ll50 DOCKET NOS. 50-333 AND 50-286 ATTACHMENT I Page 8, Section 1.3.7.
line 4, change "containment" to "reactor" Page 14, Section 2.2.1.2.
line 1, after the words "MARCH 3 code does", add the word "not",
Page 31, first full paragraph, line 7, after the words "Both the RMA and STCP analyses", add the words "do not assume" Page 48, second paragraph, line 1, change "TOl" to "TC3" l
j
pool host removal capacity and oparator cctiono to control water level during an ATWS event.
STCP improvements permit l
plant-specific design features and operator actions to be accurately modeled.
i 1.3.7.
Pressurized Water Reactors In NUREG-ll50, direct containment heating is a major contributor to uncertainty for plants with large, dry containments.
Early containment failure as a result of direct heating requires that two conditions coexist: high reactor pressure coincident with reactor vessel bottom failure due to melting core.
Authority analyses show that these two conditions are highly unlikely to occur together.
Rather, other primary system locations (like the hot leg) are likely to fail before the vessel.
This rapidly depressurizes the reactor removing one of the two early containment failure prerequisites.
Uncertainty associated with large dry containments should be reduced to reflect this.
1.3.8.
Emercency Preparedness Models Early fatality calculations are very sensitive to the models used for emergency response.
NUREG-1150's assumption that five percent of the population does not respond to emergency notification is the determining factor for early fatalities.
This technical basis for this assumption should be re-examined and the value re-established.
l i
i l
.g.
i
.=
tho boiling watcr reactor.
Nsvortholons, ths MARCH 3 coda doss not realistically analyze boiling water reactors.
Specifically, the following problem areas have been identified.
2.2.1.2.1.
Treatment of the Reactor System as One Volume As a carryover from the PWR calculation methodology, the BWR reactor system is modeled as a single node saturated system.
The net effect is an overprediction of the steaming rate and, therefore, the increased transport of fission products away from the reactor core.
In reality, there can be considerable subcooling in the downcomer and lower plenum regions of the reactor vessel.
The STCP should be modified to model BWRs as i
multi-volume systems.
i 2.2.1.2.2.
Problems Associated with the Treatment of Emercency Core Codina (ECC) Systems Two problems have been identified.
The first is the manner in which ECCS is assumed to be initiated.
In a BWR, initiation i
of the ECCS pumps as well as the ADS system is determined by the reactor vessel downcomer water level.
The Source Term Code l
~
Package does not treat the water level in the reactor system in a realistic manner and, therefore, is unable to relate the initiation of the ECCS and ADS systems to downcemer water level.
The STCP finds an "equivalent" water level and does not
[
account for different water temperatures in the reactor vessel.
Therefore, the analyst must input the time at which the ECCS is l
initiated and is turned off.
Similarly, the analyst must input 4
J I
i !
[
FigurOc TC1.3 cnd TC1.4.
Stecming proccurizca th2 drywall cgain due to suppression pool heatup and the passage of non-condensibles from the core to the suppression pool and finally to the drywell.
This recloses the ADS valves and the reactor system begins to repressurize.
The drywell pressure again begins to decrease, but not fast enough to reopen the ADS valves before the reactor vessel fails.
At 225 minutes the core slumps into the lover head, followed almost immediately by lower head failure.
Because the reactor system is at high pressure where the reactor vessel fails, the reactor system depressurization causes the drywell pressure to rapidly exceed the containment failure pressure, and the containment fails.
Although the reactor vessel has failed in the RMA analysis, the CRD pumps would continue to deliver water, which ends up in the reactor cavity.
The flow rate of water is sufficient to partially quench the core debris.
This delays significant core / concrete interaction until the inventory of the condensate storage tank is depleted and the CRD pumps are assumed to trip.
Both the RMA and STCP analyses do not assume that operator action to replenish the condensate storage tank is taken.
The probability of such an action being taken is actually very high.
The RMA results differ significantly from those produced by the STCP.
The largest contributors to these differences are the assumptions dealing with uninterrupted delivery of water by the 1 i P
II In ths RMA.analy0is, the cosium hydroxido is readily removed by chemisorption processes onto stainless steel surfaces because of f
the relatively long residence time in the reactor system during i
the period of time between 200 and 250 minutes.
A similar effect is observed in the behavior of tellurium.
The results presented in Table 3.11 show that a moderate portion of cesium iodide is retained in the reactor system, whereas cesium hydroxide is removed in the reactor coolant system to a much large extent.
Consequently, the majority of cesium iodide ends up in,the suppression pool, whereas only a small portion of cesium hydroxide and a very small portion of tellurium ends up in the suppression pool water.
The RMA analysis of the TC3 sequence shows that the thermal-hydraulic behavior of the reactor system is not doninated by artificially induced flow tates.
Therefore, adequate residence times exist for the chemisorption processes to remove cesium hydroxide and tellurium.
The behavior of cesium iodide is affected by the thermal behavior of the reactor system and highly influenced by the flow behavior following core slump and collapse.
In view of the fact that the STCP calculation for the TC3 sequence predicts containment venting prior to vessel head failure, it is not surprising that Table 3.9 shows that the suppression pool plays a major role in reducing the source term to the small quantities shown.
As' discussed earlier, early containment venting is an error and, therefore, the results in Table 3.9 are also in error.
These STCP errors overpredict t
D W T Copy COMMENTS ON NUREG-1150, REACTOR RISK REFERENCE DOCUMENT 1.
EXECUTIVE
SUMMARY
1.1 PURPOSE THE PURPOSE OF THIS EFFORT IS TO PROVIDE COMMENTS ON THE DRAFT FOR COMMENT OF NUREG-1150.
THIS SUBMITTAL CONSISTS OF TWO REPORTS, ONE OF WHICH IS THE FOLLOWING CRITIQUE Or NUREG-1150 WHICH CONTAINS SPECIFIC COMMENTS.
THE SECOND COMPANION REPORT, IS A TWO VOLUME DOCUMENT WHICH IS A SOURCE TERM ANALYSIS OF SELECTED ACCIDENT SEQUENCES AT THE PEACH BOTTOM ATOMIC POWER STATION.
THIS SECOND REPORT IS PART OF THE TECHNICAL UNDERPINNING OF THIS CRITIQUE.
BECAUSE OF THE IMPORTANCE ATTACHED TO THIS WORK, THE NEW YORK POWER AUTHORITY (NYPA)
SPONSORS THIS EFFORT AND REQUESTS THAT THE NRC CAREFULLY REVIEW ITS CONTENTS.
1.2 BACKCROUND THE NRC'S PUBLICATION OF THE DRAFT FOR COMMENT OF NUREG-1150, REACTOR RISK REFERENCE DOCUMENT, REPRESENTS A MAJOR MILESTONE IN NUCLEAR TECHNOLOGY.
BY UTILIZING ANALYSES OF A GROUP OF 4
REFERENCE PLANTS, WHICH TOGETHER COVER A WIDE SPECTRUM OF LWR
i 4
i REACTOR AND CONTAINMENT DESIGNS IN USE IN THE UNITED STATES, NUCLEAR RISKS ARE CALCULATED.
A VERY LARGE NUMBER OF HYPOTHETICAL ACCIDENT SEQUENCES ARE EXAMINED INCLUDING THEIR LIKELIHOOD OF OCCURRING AS WELL AS THE TYPES AND AMOUNTS OF RADI0 ACTIVE MATERIAL THAT MIGHT BE RELEASED TO THE ENVIRONMENT FROM SUCH ACCIDENT SEQUENCES [THE SOURCE TERMS 3.
WITH $UCH INFORMATION FURTHER ANALYSES CAN BE MADE OF POSSIBLE OFFSITE PUBLIC HEALTH AND ECONOMIC CONSEQUENCES FROM THESE HYPOTHETICAL ACCIDENTS.
THE COMBINATION OF THE LIKELIHOOD OF A PARTICULAR ACCIDENT OCCURRING AND ITS ASSOCIATED CONSEQUENCES IS A MEASURE OF THE RISK FROM THIS PARTICULAR ACCIDENT.
BY ADDING THE RISK CONTRIBUTIONS FROM ALL OF THE HYPOTHETICAL ACCIDENTS, AN OVERALL RISK STATEMENT CAN BE MADE FOR A SPECIFIC PLANT AT A SPECIFIC SITE.
THIS PROCESS ALSO PERMITS COMPARISONS OF THE RISK PROFILES OF DIFFERENT NUCLEAR PLANTS.
RISK STUDIES OF THE TYPE IN NUREG-1150 HAVE BEEN MADE IN THE UNITED STATES AND ELSEWHERE, PARTICULARLY SINCE THE 1975 PUBLICATION OF WASH-1400, THE REACTOR SAFETY STUDY.
- HOWEVER, NUREG-1150 IS DISTINCTIVE FROM THESE EARLIER EFFORTS IN THAT IT IS BROADER IN SCOPE, UTILIZES ADVANCES MADE IN BOTH SEVERE ACCIDENT COMPUTER CODES AND IN PROBABILISTIC RISK ASSESSMENT TECHNIQUES, EMPLOYS A LARGER PLANT OPERATING EXPERIENCE DATA 2-
I BASE, AND BENEFITS FROM A CONSIDERABLE INCREASE IN EXPERIMENTAL EVIDENCE ON SEVERE ACCIDENT PHENOMENOLOGY.
ADDITIONALLY, EFFORYS HAVE BEEN MADE BY THE NRC TO QUANTIFY THE UNCERTAINTY ASSOCIATED WITH THE VARIOUS RISK PREDICTIONS CONTAINED IN ITS REPORT.
ONCE ISSUED IN FINAL FORM, NUREG-1150 IS LIKELY TO HAVE A PROFOUND EFFECT ON THE PERCEPTION OF NUCLEAR RISKS FROM LWR'S IN THE UNITED STATES, AS WELL AS IMPACTING ON NUCLEAR REGULATION AND LEGISLATION FOR MANY YEARS TO COME.
VARIOUS REGULATORi ACTIVITIES SUCH AS OFFSITE EMERGENCY PLANNING, EQUIPMENT QUALIFICATION, EMERGENCY PROCEDURE GUIDES, TECHNICAL SPECIFICATIONS, AND SITING CRITERIA COULD BE INFLUENCED BY THE RESULTS IN THE FINAL VERSION OF NUREG-1150.
BECAUSE OF THE CENTRAL RDLE THAT NUREG-1150 WILL HAVE IN THE FURTHER DEVELOPMENT OF NUCLEAR TECHNOLOGY IN THE UNITED STATES, IT IS ESSENTIAL THAT THE DRAFT FOR COMMENT OF NUREG-1150 BE T
CLOSELY SCRUTINIZED AND CONSTRUCTIVELY CRITIQUED.
THIS PROCESS OF GENERAL AND PEER REVIEW HAS ALREADY BEEN INITIATED BY THE NRC, WHICH HAS EXTENDED AN INVITATION TO ALL TO COMMENT ON THE DRAFT WITHIN A DESIGNATED TIME PERIOD.
NUMEROUS EFFORTS TO RESPOND TO THIS INVITATION, INCLUDING THE ATTACHED REPORTS, HAVE BEFN UNDERTAKEN WITHIN THE NUCLEAR INDUSTRY AND ELSEWHERE.
. 4 I
, -,, ~ - -
9 SEVERAL YEARS AGO THE NEW YORK POWER AUTHORITY (NYPA), AND RISK MANAGEMENT ASSOCIATES (RMA) BEGAN A DETAILED EXAMINATION OF THE NRC'S (BMI-2104) SOURCE TERM SUITE OF CODES.
VARIOUS CODING ERRORS WERE DETECTED AND CORRECTED.
THE INDIVIDUAL CODES WITHIN THIS SUITE OF CODES WERE THEN JOINED OR COUPLED TOGETHER IN AN INTERACTIVE MANNER, AS MORE REFINED COMPUTER MODELS OF PLANT SYSTEMS OR SEVERE ACCIDENT PHENOMENOLOGY BECAME AVAILABLE, THEY WERE INCORPORATED INTO THE CODE SUITE.
THE RESULTANT CODE SUITE, IDENTIFIED AS "THE RMA CODE SET" IN THIS REPORT, HAS BEEN UTILIZED TO ANALYZE ACCIDENT SEQUENCES AT THE INDIAN POINT 3, PEACH BOTTOM, FITZPATRICK, SHOREHAM AND BELEFONTE PLANTS AND IN A COMPARATIVE STUDY OF DIFFERENT ACCIDENT SOURCE TERM METHODOLOGIES SPONSORED BY THE EMPIRE STATE ELECTRIC ENERGY RESEARCH CORPORATION (ESEERCO).
PRESENTATIONS ON THE RMA CODE SET HAVE BEEN MADE BEFORE THE AMERICAN PHYSICAL SOCIETY, THE SUBCOMMITTEE ON CLASS 9 ACCIDENTS OF THE ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS), THE NRC STAFF, THE NRC ConMISSIONERS, A RECENT EPRI-LED INDUSTRY PEER REVICW GROUP, AND AT VARIOUS NATIONAL AND INTERNATIONAL SOURCE TERM CONFERENCES.
THE RHA CODE SET HAS SEEN USED TO PREDICT THE OUTCOME OF SOURCE TERM EXPERIMENTS IN EPRI'S LACE PROGRAM, WITH FAVORABLE RESULTS.
COMPARISONS OF RMA CODE SET ANALYSES TO THE LOFT EXPERIMENTAL PROGRAM RESULTS AND TO TMI ACCIDENT OBSERVATIONS ARE UNDERWAY.
, m
l BECAUSE THE RMA CODE SET HAS ITS ROOTS IN THE SAME CODES NOW UTILIZED BY THE NRC IN ITS OWN SOURCE TERM CODE PACKAGE (STCP),
IT IS POSSIBLE TO MAKE POINT BY POINT COMPARISONS BETWEEN SPECIFIC SOURCE TERM ANALYSES CONTAINED WITHIN NUREG-1150* BY RECALCULATING THE ACCIDENT SEQUENCES UTILIZING THE RMA CODE SET.
SIMILARITIES AND DIFFERENCES RETWEEN THE NRC AND RMA ANALYSES FOR THE SAME ACCIDENT SEQUENCES CAN BE IDENTIFIED AND EXPLAINED.
THIS UNIQUE DIAGNOTIC CAPABILITY HAS BEEN UTILIZED TO ANALYZE SEVERAL ATWS (ANTICIPATED TRANSIENT WITHOUT SCRAM)
AND STATION BLACKOUT SEQUENCES AT THE PEACH BOTTOM PLANT, NUREG-1150'S REFERENCE MARK I BWR, AND COMPARISONS ARE PROVIDED IN CHAPTER 3.
THE GREAT MAJORITY OF PEACH BOTTOM'S CALCULATED PISK IN NUREG-1150 COMES FROM ATWS AND STATION BLACKOUT i
SEQUENCES.
THE SEQUENCES SELECTED FOR EXAMINATION ARE AMONG FIVE BASIC ACCIDENT SEQUENCES USED IN NUREG-1150'S ANALYSIS OF PEACH BOTTOM.
THESE FIVE SEQUENCES FORM THE ANALYTICAL FOUNDATION OF NUMEROUS EXTRAPOLATED SOURCE TERMS LATER UTILIZED IN NUREG-1150.
THE CORRECTNESS OF THE PEACH BOTTOM PLANT DESCRIPTION USED AS INPUT IN THESE ANALYSES HAS BEEN VERIFIED BY THE PHILADELPHIA ELECTRIC POWER COMPANY.
i 4
- FoR SIMPLICITY, THE NUREG-1150 DRAFT FOR COMMENT AND ITS ASSOCIATED DOCUMENTS IS HENCEFORTH REFERRED TO AS "NUREG-1150" I
l ;
j i
f 1
5 IN ADDITION TO THE INSIGHTS GAINED BY COMPARING THE RMA CODE SET TO THE NRC'S STCP ANALYSES FOR THESE PEACH BOTTOM ACCIDENT J
4 SEQUENCES, SOME OBSERVATIONS ARE AVAILABLE FROM THE ANALYSES OF l
THE OTHER PLANTS LISTED ABOVE.
.THuS COMMENTS ON NUREG-1150 ARE PRESENTED FOR BOTH BWR'S AND PWR'S.
i 1.3 MAJOR CONCLUSIONS P
1.3.1 IMPORTANCE OF THE STCP i
OUR REVIEW OF NUREG-1150 CONCLUDES THAT THE STCP HAS f
l IMPORTANT DIRECT AND INDIRECT IMPACTS ON NUREG-1150 I
I I
RESULTS.
THEREFORE, STCP DEFICIENCIES SHOULD BE OVERCOME j
i AND RISK DOMINANT SEQUENCES REANALYZED.
1.3.2 UNCERTAINTY f
A MAJOR CONCERN ABOUT NUREG-1150 IS THE VERY WIDE UNCERTAINTY BANDS AROUND THE STCP RESULTS.
THESE WIDE i
UNCERTAINTY BANDS WERE GENERATED BY AN EXPERT OPINION I
PROCESS; ONE WHICH HAS OFTEN BEEN CRITICIZED.
IT IS l
j IMPORTANT THAT THE PROCESS BY WHICH EXPERT OPINION IS i
GATHERED AND UTILIZED TO GENERATE THESE UNCERTAINTY BANDS BE REVIEWED AND IMPROVED, AS NECESSARY.
HOWEVER, ACCORDING 1
t l
I
__,_.-__.-.,.m.
,__y
__.m_,__,y,
l-TO THE NRC, A PORTION OF THIS UNCERTAINTY IS GENERATED BY PHENOMENOLOGY THAT ARE NOT ADEQUATELY ACCOUNTED FOR IN THE STCP.
FURTHERMORE, ALL PHENOMENA THAT ARE ABSENT IN THE NRC'S STCP ARE THOUGHT TO POTENTIALLY INCREASE THE SIZE OF THE SOURCE TERMS.
OUR REVIEW INDICATES THAT MANY OF THESE MISSING PHENOMENA ARE ALREADY ACCOUNTED FOR IN THE RMA CODE SET.
CONSEQUENTLY, THE SOURCE TERMS CALCULATED BY THE RMA CODE SET SHOULD HAVE NARROWER UNCERTAINTY BANDS AND SHOULD BE LESS SENSITIVE TO THE LIMITATIONS OF ANY EXPFRT OPINION PROCESS.
SHOULD THE STCP BE MODIFIED AS SUGGESTED IN A NUMBER OF PLACES IN THIS REPORT, REVISED NUREG-1150 RESULTS WOULD EXHIBIT NARROWER UNCERTAINTY DANDS THAN THE PRESENT ANALYSES 00.
FURTHER NARROWING OF THE UNCERTAINTY MAY OCCUR AS AN INDIRECT RESULT OF IMPROVED STCP RESULTS.
JUDGMENT IS EXERCISED BY THE ANALYSTS THROUGHOUT THE WHOLE NUREG-1150 PROCESS IN THE SELECTION OF SEQUENCES, 1HE CLUSTERING OF PLANT DAMAGE STATES, AND IN NUMEROUS OTHER WAYS.
JUDGMENT IS ALSO EXERCISED BY THE EXPERTS IN THE WEI5HTING FACTORS 1
THEY ASSIGN TO UNCERTAIN EVENTS OR PHENOMENA.
THESE JUDGMENTS ARE INFLUENCED BY PERCEPTIONS OF HOW ACCIDENT SEQUENCES EVOLVE.
TO THE EXTENT THAT THESE JUDGMENTS ARE MODIFIED BY AN IMPROVED CAPABILITY TO CALCULATE ACCIDENT !
i
i CHRONOLOGIES, THE HEATUP OF STRUCTURAL SURFACES, THE OPERATION OF PLANT SYSTEMS, THE DEVELOPMENT OF PRF.SSURE DIFFERENTIALS, THE TIME DEPENDENT LOCATIONS OF 1
RADIONUCLIDES, ETC., THE OVERALL UNCERTAINTY MAY CHANGE.
t ANALYSTS MAY STRUCTURE THE NUREG-1150 PROCESS DIFFERENTLY AND EXPERTS MAY SELECT DIFFERENT WEIGHTING FACTORS BASED ON THEIR HIGHER LEVEL OF KNOWLEDGE.
t ADDITIONAL INSIGHTS HAVE BEEN GAINED THAT ALSO HAVE A MAJOR BEARING ON THE UNCERTAINTY ISSUE.
WE HAVE EXPLORED 4
VARIATIONS OF BASIC NUREG-1150 SEQUENCES IN WHICH WE L
ASSUMED THAT CERTAIN ACTIONS WERE TAKEN, SUCH AS REFILLING THE CONDENSATE STORAGE TANK, EXTENDING THE TIME DURING WHICH THE AUTOMATIC DEPRESSURIZATION SYSTEM (ADS) IS OPEN, 1
AND CONTINUING WATER FLOW THROUGH THE CONTROL ROD DRIVE SYSTEM OR THE REACTOR COOLANT INJECTION SYSTEM.
AS DESCRIBED IN THE BODY OF THIS REPORT, SUCH ACTIONS CAN I
LEAD TO THE PREVENTION OF A CORE MELI, PROVIDE MORE TIME FOR RECOVERY ACTIONS, OR REDUCE SOURCE TERMS SHOULD THERE l
?
BE A MELTED CORE.
FOR EXAMPLE, ASSUMING THE CAPABILITY EXISTS TO KEEP THE ADS VALVES OPEN, EVEN IN A 3TATION BLACKOUT SEQUENCE WHERE BATTERY POWER IS IMMEDIATELY 4
UNAVAILABLE, WILL RESULT ZN A.*?OUCEO PROBABILITY OF
]
i DRYWELL LINER MELTTHROUGh, A MAJOR CONTRIBUTOR TO BWR t
)
UNCERTAINTY.
-8
[
]
r THE IMPLICATION OF THE ABOVE ALTERNATIVE ANALYSES IS THAT THERE ARE A NUMBER OF "SUCCESS PATHS" WHICH MAY NOT HAVE BEEN CONSIDERED IN NUREG-1150.
SuCH ANALYSES CALL FOR A REEVALUATION OF THE ASSUMED FREQUENCIES OF THOSE SEQUENCES THAT ARE NOW IMPORTANT DETERMINANTS OF THE RISK IN~
IT MUST BE REESTABLISHED WHAT FRACTION OF THE TIMI' AN ACCIDENT PROGRESSES DOWN ONE OF THESE SUCCESS PATHS VERSUS THE PATH CHOSEN IN NUREG-1150.
TO THE EXTENT THAT
-l THE FREQUENCIES OF THE PRESENT DOMINANT SEQUENCES IN i
NUREG-1150 ARE THEN REVISED DOWNWARD, THE IMPORTANCE OF THE UNCERTAINTY ISSUES ASSOCIATED WITH SUCH SEQUENCES IS CORRESPONDINGLY REDUCED.
THEREFORE, IMPLEMENTATION OF THE SUGGESTIONS IN THIS REPORT i
SHOULD REDUCE THE WIDTHS OF THE UNCERTAINTY BANDS.
1.3.2 PLANT SYSTEMS THE STCP HAS UNREALISTIC REPRESENTATIONS OF VARIOUS BWR SYSTEMS SUCH AS THE AUTOMATIC DEPRESSURIZATION SYSTEM, THE SAFETY RELIEF VALVES, AND THE EMERGENCY CORE COOLING SYSTEM.
FURTHER, THE STCP DOES NOT ADEQUATELY DESCRIBE THE RELATIONSHIP BETWEEN REACTOR POWER LEVEL AND CORE WATER LEVEL DURING ATWS EVENTS. i
1.3.3 INPUT ERRORS A NUMBER OF IMPORTANT INPUT ERRORS WERE MADE IN NUREG-1150 IN THE PEACH BOTTOM ANALYSES.
SOME OF THESE ARE STRAIGHT FORWARD NUMERICAL ERRORS SUCH AS TWICE THE CONTROL ROD DRIVE PUMP FLOW RATE OR THE TRIP SET POINT OF PEACH BOTTOM'S RCIC THIRTY POUNDS TOO LOW.
OTHER INPUT ERRORS ARE MORE IN THE NATURE OF MODELING ERRORS.
ONE SUCH ERROR APPARENTLY IS THE RESULT OF TREATING A BWR CGMTAINMENT CAVITY LIKE A PWR CONTAINMENT SUMP.
THIS ERROR RESULTED IN OVERLOOKING TENS OF THOUSAND3 0F POUNDS OF WATER ON THE DRYWELL FLOOR IN A NUMBER OF MAJOR SEQUENCES.
THIS WATER AFFECTS THE WHOLE ACCIDENT PROGRESSION ONCE THE REACTOR VESSEL FAILS.
ANOTHER MODELING ERROR IS THE ASSUMED CONTINUING OPERATION OF PEACH BOTTOM'S HIGH PRESSURE COOLANT (HPCI) INJECTION SYSTEM FAR BEYOND PHYSICAL LIMITS OF THE SYSTEM.
HERE THE STCP INPUT ON THE HPCI TRIP POINT NEEDS TO BE CORRECTED.
CONTROL ROD DRIVE PUMP FLOW AFTER REACTOR VESSEL FAILURE WAS ALSO IMPROPERLY MODELED.
IN ADDITION TO THESE INPUT MODELING ERRORS, THERE ARE MODELING ERRORS INTERNAL TO THE STCP WHICH AFFECT THE CALCULATION OF THE TRANSPORT OF FISSION PRODUCTS WITHIN THE REACTOR SYSTEM AND THE CONTAINMENT.
THIS CLASS OF MODELING ERRORS IS ALSO DISCUSSED IN THE BODY OF THIS REPORT.
1.3.4 OPERATOR ACTIONS AS A CONSEQUENCE OF 1.3.2 AND 1.3.3, ABOVE, IT IS DIFFICULT FOR THE STCP TO EVALUATE THE SAFETY SIGNIFICANCE OF OPERATOR ACTIONS.
FOR EXAMPLE, STCP LIMITATIONS CAN UNDERESTIMATE T.1E IMPORTANCE OF REACTOR WATER LEVEL MANAGEMENT DURING CERTAIN ATWS SEQUENCES.
POOR WATER LEVEL CONTROL CAN AGGRAVATE ACCIDENT CONDITIONS.
ON THE OTHER HAND, THESE STUDIES SHOW THAT FURTHER RISK REDUCTION OPPORTUNITIES MAY BE GAINED WHEN THE PLANT'S RESPONSE TO ACCIDENT CONDITIONS IS PROPERLY ACCOUNTED FOR.
EXAMPLES OF THIS INCLUDE CONDENSATE STORAGE TANK REFILL AND EXTENDED USE THE ADS.
THEREFORE IMPROVEMENT IN THE STCP WOULD BE VALUABLE IN PREVENTING ADVERSE OPERATOR ACTIONS AND IN RECOGNIZING RISK REDUCTION OPPORTUNITIES.
1.3.5 EXTERNAL EVENTS ONGOING EFFORTS WITHIN THE NRC ARE DIRECTED TOWARD EXPANDING THE SCOPE OF NUREG-1150 TO INCLUDE THE EFFECTS OF LARGE EXTERNAL EVENTS, SUCH AS EARTHQUAKES AND HURRICANES.
SOME PREVIOUS PROBABILISTIC RISK ASSESSMENTS HAVE CONCLUDED THAT, FOR THE PLANTS ANALYZED, EXTERNAL EVENTS DOMINATE THE RISK.
PEACH BOTTOM IS ONE OF THE PLANTS THAT THE NRC HAS SELECTED TO EXPLORE THE EXTERNAL EVENTS QUESTION,
4 4
IT WAS FOUND THAT EXTENDED OPERATION OF PEACH BOTTOM'S AUTOMATIC DEPRESSURIZATION SYSTEM, DESCRIBED LATER IN THIS REPORT, IS USEFUL BOTH IN REDUCING MAJOR UNCERTAINTIES IN STATION BLACK 0UT SEQUENCES AND IN REDUCING RISKS FROM EXTERNALLY INITIATED EVENTS.
1.3.6 DESIGN FEATURES IF THE STCP WERE MODIFIED AS SUGGESTED HEREIN, THEN IT WOULD BE POSSIBLE TO MORE PRECISELY DETERMINE THE VALUE OF PLANT DESIGN FEATURES.
FOR EXAMPLE, AT THIS TIME THE STCP CANNOTDISCERNTHEDIFFERENCEBETWEENTWOjTAGEAND THREE-STAGE ADS VALVES AND THE IMPACT OF THESE DIFFERENT i
j VALVE DESIGNS ON THE COURSE OF CERTAIN ATWS EVENTS.
i l
ANOTHER EXAMPLE OF THE IMPORTANCE OF BEING ABLE TO EVALUATE i
j SPECIFIC PLANT DESIGN FEATURES IS THE RELATIONSHIP BETWEEN SUPPRESSION POOL HEAT REMOVAL CAPACITY AND OPERATOR ACTIONS TO CONTROL REACTOR WATER LEVEL DURING ATWS EVENTS.
AN
^
IMPROVED STCP IS NECESSARY TO EVALUATE SPECIFIC OPERATOR
- l ACTIONS AT INDIVIDUAL PLANTS.
i i
1
(
i l
l C
AS THE NRC MOVES BEYOND REFERENCE PLANT ANALYSES AND BEGINS
~
TO INVESTIGATE THE CHARACTERISTICS OF INDIVIDUAL PLANTS, IT NEEDS AN ANALYTICAL TOOL OF SUFFICIENT SOPHISTICATION TO BE l
ABLE TO EVALUATE THE PARTICULAR FEATURES OF EACH PLANT.
TO l
00 THIS, THE STCP NEEDS TO BE UPGRADED.
1.3.7 ADDITIONAL SAFETY FEATURES A BYPRODUCT OF THE RMA ANALYSES OF THESE PEACH BOTTOM ACCIDENT SEQUENCES IS THAT FURTHER LIGHT IS SHED UPON THE DESIRABILITY OF VENTING CONTAINMENTS AND INSTALLIllG DEBRIS BARRIERS.
BASED ON THESE ANALYSES, THE SAFETY WORTH OF i
SUCH FEATURES APPEARS TO BE QUITE LIMITED.
1.3.8.
PRESSURIZED WATER REACTORS I
WITH RECsARD TO PRESSURIZED WATER REACTORS, NUREG-1150 ASSIGNS DIRECT HEATING AS A MAJOR CAUSE OF THE UNCERTAINTY IN THE RISKS FROM PLANTS WITH LARGE, DRY CONTAINMENTS.
DIRECT HEATING IS POSTULATED TO PRODUCE ACCIDENT CONDITIONS WHERE THE CONTAINMENT COULD FAIL RATHER EARLY BECAUSE OF LARGE PRESSURE SPIKES IN THE CONTAINMENT.
DIRECT HEATING REQUIRES A COMBINATION OF HIGH REACTOR PRESSURE AT THE TIME WHEN MOLTEN CORE MATERIAL WOULD HAVE CREATED A PATHWAY THROUGH THE BOTTOM OF THE REACTOR VESSEL.
OUR ANALYSES SHOW THAT HIGH SYSTEM PRESSURE AND PENETRATION OF THE i
REACTOR VESSEL BY MOLTEN CORE MATERIAL ARE VERY UNLIKELY TO CO-EXIST.
OTHER LOCATIONS IN THE PRIMARY SYSTEM BOUNDARY, SUCH AS AT THE HOT LEG, ARE LIKELY TO FAIL BEFORE THE REACTOR VESSEL FAILS WITH A RESULTANT RAPIO DECREASE IN PRIMARY SYSTEM PRESSURE.
UTILIZATION OF THIS INFORMATION WOULD REDUCE THE UNCERTAINTY BANDS ASSIGNED TO LARGE, DRY i
PWR ANALYSES.
1.3.9 EMERGENCY RESPONSE MODELING THE CALCULATED EARLY FATALITY RISK IS HIGHLY SENSITIVE TO 1
i EMERGENCY RESPONSE MODELING.
IN PARTICULAR, NUREG-1150'S ASSUMPTION THAT 5% OF THE POPULATION DOES NOT TAKE A TIMELY
- RESPONSE, I.E.,
CONTINUES ON WITH NORMAL ACTIVITIES, IS IHE DETEAMINING FACTOR FOR THE EARLY FATALITY RISK.
THIS i
ASSUMPTION SHOULD BE CLOSELY SCRUTINIZED AND A VALUE WHICH l
HAS A TECHNICAL BASIS SHOULD BE BE USED.
I e
I I i I
t
,tirp_*f.f7 o~
/
ROUTING AND TRANSMITTAL SLIP TO: (Nemo, o#nce symbol. rooro numeer, initiate este bul%. A0entya%et) 1.
{(
3.
4.
Action File Note and Retum Approval For Cloerence Per Conversation Aa % : _ 2 For Correction Propero Reply
)
Clrevnete For Your Information See Me Co.n.n.
in.e.fi.ei.
. w u,e lCoord6nstion JustW RtMARES
< V V0 %
"[h ese cs ye Glee +.c %
W 9/CP/E ?
/e #4 -
The a.ccen%
h % bn-1 gy -f1c. a vigi%l
/e -//e,
2 fr 9 '7/O '2/ O QC) 9 -
.I 00 NOT wee this form as a RECORD of approvals, concurrences, diepoeals.
]
clearances. and selter actions PROM: (Name, orp. symbol. Apexy.PoafJ Room No.--81dg.
l k.
2 Jr t* V 8881 W u a oeo un+4si n4 roca4 O]P l,03ALg4W 41(hev.179) e FPese totYres tet 11q
REGULATORY INFORMATION DISTRIBUTION SYSTEM (RIDS)
ACCESSION NDR: 8710210209 DOC. DATE: 87/09/23 NOTARIZED: NO DOCKET O FACIL: 50-000 Generic Doc ket 05000000 50-286 Indian Point Station, Unit 3, Power Authority of Stat 05000286 50-333 James A.
FitzPatrick Nuclear Power Plant, Power Autho 05000333 AUTH.NAME AUTHOR AFFILIATION DRONS.J.C.
New York Power Authority (formerly Power Authority of the S RECIP.NAME RECIPIENT AFFILIATION Division of Rules & Records (Post 870413)
SUBJECT:
Comment opposing draft NUREG-1150, "Reactor Risk Ref f
Document." Reduced ur. certainty in risk assessment found to be significant w/ respect to NUREG-1150.NUREO also does not consider value of operator actions.Addl comments enc 1.
DISTRIBUTION CODE: DSO9D COPIES RECEIVED: LTR a ENCL -( SIZE:
TITLE: SECY/DSD Dist: Public Comment on Proposed Rule (PR)-Misc NoticeJReg G NOTES:Lpdr 2cy PDR Do c u m e n t r..
05000?86 RECIPIENT COPIES RECIPIENT COPIES ID CODE /NAME LTTR ENCL ID CODE /NAME LTTR ENCL 4
1
- g wA- -
1 1
OGC/SAIP 1
1 G FILE 1
1 RES DIR 1
1 REs7PMPDAS 1
1 EXTERNAL: NRC PDR 01 1
1 NOTES:
2 2
Od d ' C. h dd2.
a Q}{hhhk to
/ 0 TOTAL NUMBER OF COPIES REQUIRED: LTTR
-f ENCL'
-4 "
F
,