IR 05000237/1991007
| ML17202V037 | |
| Person / Time | |
|---|---|
| Site: | Dresden  |
| Issue date: | 03/13/1991 |
| From: | Lougheed V, Phillips M NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | |
| Shared Package | |
| ML17202V036 | List: |
| References | |
| 50-237-91-07-EC, 50-237-91-7-EC, EA-91-014, EA-91-14, NUDOCS 9103260025 | |
| Download: ML17202V037 (38) | |
Text
U. S. NUCLEAR REGULATORY COMMISSION
REGION III
Report N /91007(DRS)
Docket N EA 91-014 License No. DPR-25 Licensee:
Commonwealth Edison Company Opus West III 1400 Opus Place Downers Grove, IL 60515 Fa~ility Name:
Dresden Station Unit 2 Meeting At:
U. S. Nuclear Regulatory Commission Region III Office Glen Ellyn, IL 60137 Meeting Conducted:
February 14, 1991 Telephone conference conducted:
March 4, 1991 Telephone results of review presented:
March 12, 1991 Type of Meeting:
Enforcement Conference Inspector:
/~~,/
V: P. ugheecJ Approved by:~~
M. P. Phillips, Chief, Operations Branch Meeting Summary:
3//3/9,1 Date Meeting on February 14, 1991, Conference call on March 4, 1991 (Report No. 50-237/91007CDRS))
Matters Discussed:
Apparent violation of Technical Specification 3.7.A.2:
Failure to maintain containment integrity when the reactor is critica Unit 2 operated from February 1989 to September 1990 without containment integrity due to insufficiently tightened bolts on the inboard flange of a torus purge exhaust containment isolation valv When discovered during the conduct of the containment integrated leak rate test in December 1990, the flange leaked at a measured rate of 24.6 weight percent of containment atmosphere per day at 14.6 psig, with design pressure being 48 psig. This was consi£!~rably in excess of the Technical Specification defined allowable of 1.6 wE?<Pght percent per day at design pressur The specific d.eficiency is descri-bed in detail in Inspection Reports No. 50-237/91006(DRS);
No. 50-249/91006(DRS).
Information regarding the causes, corrective actions, and significance of the calculated leak rate were discusse The licensee submitted additional information regarding its calculational methodology to 9103260025 910315 PDR ADOCK 05000237 G
the NRC on February 22, 199 NRC review of that material is discussed in this repor The results of the NRC review indicate that the containment would not have been capable of performing its safety function in the event of a large break loss of coolant accident during the last operational cycl Disposition of the apparent violation will be presented in subsequent communications.
- *
DETAILS Enforcement Conference Attendees Commonwealth Edison Company D. Galle, Vice President, BWR Operations E. Eenigenburg, Station Manager, Dresden T. Kovach, Nuclear Licensing Manager H. Massin, BWRSD Superintendent J. Brunner, Technical Staff Support Superintendent D. Van Pelt, Assistant Superintendent, Maintenance, Dresden K. Brennan, Engineering and Construction Regulatory Assurance Supervisor K. Peterman, Regulatory Assurance Supervisor, Dresden M. Strait, Technical Staff Supervisor, Dresden S. Stimac, Nuclear Licensing Administrator
- M. Richter, Nuclear Licensing Administrator, Dresden J. Lockwood, Regulatory Assurance Supervisor, LaSalle
-*B. Viehl, Design Supervisor, Nuclear Engineering, Dresden D. Bucknell, Assistant Technical Staff Supervisor, Quad Cities D. Schumacher, Technical Staff Engineer, Quad Cities G. Edgar, Attorney M~ Turbak, Superintendent of Onsite Safety Groups R. Geier, Master Mechanic, Dresden
- E. White, Engineering and Construction Regulatory Assurance R. Radtke, Nuclear Licensing P. Barnes, Nuclear Licensing J. Glover, Production Services NUS Corporation
- D. Studley, Mechanical Systems Engineer U. S. Nuclear Regulatory Commission A. B. Davis, Regional Administrator, Region III
- C. J. Paperiello, Deputy Regional Administrator, Region III T. 0. Martin, Director, Division of Reactor Safety (DRS)
G. C. Wright, Acting Deputy Director, DRS W. D. Shafer, Chief, Branch I, Division of Reactor Projects
- M. P. Phillips, Chief, Operations Branch, DRS
- B. L. Siegel, Dresden Project Manager, NRR C. H. Weil, Enforcement and Investigation Coordination Staff (EICS)
F. A. Maura, Reactor Inspector, DRS
- Denotes those individuals participating in the March 4, 1991, conference call.
U.S. Nuclear Regulatory Commission (Cont'd)
M. S. Peck, Resident Inspector, Dresden Site J. D. Monninger, Resident Inspector, Dresden Site
- P. R. Pelke, EICS K. Salehi, Reactor Inspector, DRS K. M. Shembarger, Operator Licensing Examiner, DRS
- J. Y. Lee, Radiation Protection Branch, NRR
- J. C. Pulsifer, Plant Systems Branch, NRR
- C. F. Gill, Senior Reactor Programs Specialist, DRSS
- C. D. Pederson, Director, EICS
- B. L. Burgess, Chief, Section 18, DRP
- Denotes those individuals participating in the March 4, 1991, conference cal.
Enforcement Conference As a result of an apparent violation of NRC requirements, an Enforcement Conference was held at the Region III Office in Glen Ellyn, Illinois on February 14, 199 The preliminary findings, which were the bases for the apparent violation of NRC requirements, were documented in NRC Inspection Report 50-237/91006(DRS); 50-249/91006(DRS).
The attendees of this enforcement conference are denoted in Paragraph 1 of this report.
The purposes of the enforcement conference were to discuss the apparent violation, the cause, and corrective actions; to review the licensee's conclusions regarding the calculated leak rate and subsequent significance; to determine whether there were any mitigating circumstances; and to obtain any other information which would help determine the appropriate enforcement actio Following introductory remarks, the NRC staff opened the conference by stating the purpose of the conference, presenting a brief description of the apparent violation, and summarizing the inspection finding The licensee did not contest the information presented by the NR The licensee was requested by the NRC to discuss the following:
o The root cause(s) of the event; o
The corrective actions to prevent recurrence; o
The safety significance in terms of dose releases following a loss of coolant accident; and o
The basis for the underlying calculations for the dose releases, including extrapolation of the leak rate data from 15 to 48 psi The licensee presented the results of their review concerning the apparent violation and addressed the above discussion topic In its analysis, the licensee described the corrective actions initiated to
- .
preclude repetition of the apparent violation, and presented its determination of the safety-significance of the even *
(See Enclosure 1)
Several questions were raised by NRC staff member concerning the rationale for the calculational methodologies employed in determining the events potential safety significance. The licensee was unable to present this information during the meetin The licensee representatives committed to submit additional information in regard to
~he rationale and final results to the Region III staff within one week of the conferenc Enforcement Conference Conclusion At the conclusion of the conference, the NRC thanked the licensee for the presentation and agreed to review the supplemental material regarding safety significance when it was receive The final evaluation and disposition of the apparent violation will be presented in subsequent communication Results of ~eview of February 21. 1991, Licensee Submittal On February 22, 1991, the licensee provided additional information*
relating to what the extrapolated leak rate would have been at design test pressure, and what the offsite consequences concerning dose would have been had a large break loss of coolant accident (LOCA) occurred during the last operational cycle, i.e., whether exposures to the public would be below the limits specified in 10 CFR Part 100. A conference call was held on March 4, 1991, to clarify information contained within the February 21, 1991, submittal, and to determine the effects of iodine loading on the efficiency of the standby gas treatment system (SBGTS). Extrapolated Leak Rate Calculation The measured leak rate at 14.6 psig was 24.6 wt%/da The design pressure for leakage measurement is 48 psig. The licensee presented an extrapolated leak rate of 25 wt%/ day for choked fl ow, 31 wt%/day for turbulent flow, and 66 wt%/day for laminar flo During the enforcement conference the licensee representatives stated that the best representation for the extrapolation was choked flow; however, turbulent flow conditions were used in the dose calculations. The licensee assumed in its calculations that the same flow conditions existed at both 14.6 and 48 psig. The resulting calculated Reynolds number of 21,600 at 48 psig was utilized to justify using the turbulent flow valu The licensee also concluded that, based on the composition of the flange material, the area of the leak pathway would not change with increasing pressur No experimental data was available to support these conclusion An Atomic Energy Commission report, TID-20583, which was based on experimental data, states that "in the absence of leakage test data at the higher pressure, it appears that for upward extrapolation the laminar flow may be considered as most
- *
conservative while extrapolation based on orifice flow will provide the least conservative value for the leakage rate." Given conditions observed at the time of the leakage measurements th NRC staff believes it is most appropriate to assume that the formula would reflect a combination of both laminar and choked flo Laminar flow conditions could be expected to occur from the actual test pressure up to approximately 25 psig, beyond which choked flow conditions would be expecte This methodology yields a leakage rate of 44 wt%/day at 48 psi Dose Consequences Calculations The licensee determined that the 25 Rem whole body and 300 Rem thyroid dose limits at both the site boundary and low population zone (LPZ) specified in 10 CFR Part 100 would not be exceeded only if non-design basis assumptions were utilized. Similarly, the control room doses would not exceed the General Design Criteria 19 limits of 5 Rem whole body, 30 Rem thyroid, and 30 Rem skin dose only if non-design basis assumptions were employe The licensee utilized the following assumptions, which it felt were more realistic than those utilized in the design of the facility, in its presentation of dose consequences:
1)
Assume 25% of the equilibrium iodine and 100% of the equilibrium noble gas fission products are immediately available for release, with the iodine fractions at 91% elemental, 5% particulate, and 4% organi )
Bypass leakage through the MSIVs is assumed to be at a rate of 276 standard cubic feet per hour for each steamline, with a transit time through the piping of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. This results in no bypass contribution to exposure until 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after the accident begin )
Assume the relative concentration for ground level and elevated release are the same as those provided in the NRC's January 5, 1982, accident analysis for SEP Topic XV-1 )
A two compartment model for containment leakage, namely, leakage from the torus to secondary containment at 5)
31 wt%/day, followed by assumed complete mixing in secondary containment, and a subsequent secondary containment leakage to standby gas treatment at a rate of 100 wt%/da This allows for holdup and decay of short-lived radionuclides in secondary containment prior to release to the environmen These leakrates are assumed to be constant for the duration of the acciden Iodine retention occurs in the suppression pool such that elemental and particulate activities are reduced by a decontamination factor of 5. This is in accordance with
the guidance contained in Section 6.5.5 of the current NRC Standard Review Pla )
Standby gas treatment system efficiency is taken as that measured by the licensee during its last test, namely 98%,
rather than the design and technical specification limits of 90%.
The licensee did not evaluate the reduced efficiency effects caused by iodine and contaminant loading of the SBGTS which would result in a higher percentage of iodine being released to the environmen Utilizing the above described assumptions and methodology, the licensee computed the following doses for a torus leak rate of 31 wt%/day:
2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> Site 30 day 30 day Boundary LPZ Control Room Thyroid Dose 1.6 Rem 39.2 Rem 18.7 Rem Whole Body Dose 1.3 Rem 1.2 Rem 1.4 Rem Skin Dose N/A N/A 24.6 Rem The NRC staff considered assumptions 4 and 6 above to be overly non-conservative, and therefore calculated the dose consequences based on the more conservative design basis assumption The result was that Part 100 limits would be exceeded for the whole body site boundary dose, and the control room limits specified in GDC-19 would be exceeded for whole body, thyroid, and skin dos The above assumptions utilized by the licensee that were incorporated into the NRC analysis were numbers 1, 2, 3, and _Assumption 4, the two compartment model with mixing, is contrary to regulatory position l.e of Regulatory Guide 1.3 (1974).
This regulatory position states that the primary containment should be assumed to leak at the specified leak rate for the duration of the accident and that the leakage should be assumed to pass directly to the emergency exhaust *system without mixing in the surrounding reactor building atmospher If this assumption is not utilized, the licensee calculated the associated doses to be as follows:
2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> Site 30 day 30 day
- Boundary LPZ Control Room Thyroid Dose 39.5 Rem 50.6 Rem 490.0 Rem Whole Body Dose 41.4 Rem 9.0 Rem 24.2 Rem Skin Dose N/A N/A 395.0 Rem
_
These doses still take credit for a 983 efficiency for iodine removal for the SBGTS, and result in the Part 100 limit of 25 Rem for the whole body site boundary dose being exceeded, and all GDC-19 dose limits for the control room being exceede However, based on the volume of iodine and other impurities that. would be released for this postulated accident, there* is insufficient assurance that the SBGTS will continue to perform at a 98%
efficiency. Given a calculated leakage rate of 31 wt%/day, a nominal mass of 15 kilograms of iodine within the core (per NUREG-0956), and a decontamination factor of 5 for elemental and particulate iodine within the suppression pool, the total mass of iodine that could be expected to be released to the torus atmosphere would be approximately 870 gram Over time, all of this material would subsequently be released to secondary containment. Section 5.3.2.5 of the Dresden Updated Final Safety Analysis Report states that the carbon bed capacity of the SBGTS is approximately 100 grams of iodin However, based on discussions with licensee representatives on March 6, 1991, one train of SBGTS contains approximately 600-720 pounds of carbo Per the methodology described in Regulatory Guide 1.52, approximately 766 pounds of carbon would. be sufficient to contain 870 grams 6f iodine assuming no other contaminants were presen The licensee was unable to ascertain why the UFSAR value was only 100 gram However, based on the uncertainties between the UFSAR number and the calculational methodology, and the need to assure that no impurities were present in the post-accident environment which could also load the charcoal, there does not appear to *be sufficient justification for utilizing the test results for SBGTS efficiency rather than the design and technical specification values. Therefore, the staff extrapolated the effect on dose of lowering the SBGTS efficiency to 903 for the 31 wt%/day leakrat The result was as follows:
2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> Site 30 day 30 day Boundary LPZ Control Room*
Thyroid Dose 195.0 Rem 91.7 Rem 2,000 Rem Whole Body Dose 41.4 Rem 9.0 Rem 24.2 Rem Skin Dose N/A N/A 395.0 Rem
- - Neglects dose contribution from MSIV pathway Given these assumptions, the same limits are exceeded regardless of the SBGTS efficiency, the only difference being by how much they are exceede If the leak rate is taken as 44 wt3/day, as the NRC extrapolation would indicate is more appropriate, ~he above whole body doses,
- *
two hour thyroid dose, and control room skin and thyroid doses would be increased by a factor of 1.42, giving a resultant whole body dose at the site boundary of 58.8 rem, and even larger doses to the control room staf Conclusions Based on the above calculations, the staff concludes that the Dresden containment would not have been capable of performing its safety function given a large break loss of coolant accident during the last operational cycle. *These results were discussed with Mr. Milt Richter of the licensee's staff on March 12, 1991.
Enclosure 2 FEBRUARY 14, 1991 DRESDEN ENFORCEMENT CONFERENCE UNIT 2 PRIMARY CONTAINMENT ISOLATION V Al VE INBOARD FLANGE NOT LEAK RATE TESTED AFTER VALVE REPLACEMENT AGENDA
- INTRODUCTION
- EVENT CHRONOLOGY AND CORRECTIVE ACTIONS
- LEAK RATE CALCULATION D. GALLE E. EENIGENBURG J. GLOVER
- POTENTIAL SAFETY SIGNIFICANCE H. MASSIN
- OVERALL CONCLUSIONS E. EENIGENBURG
- SUMMARY D. GALLE
- ULD770/l
- CECO OVERVIEW
FAILURE TO MAINTAIN PRIMARY CONTAINMENT INTEGRITY DURING THE LAST OPERATING CYCLE ON UNIT 2 WAS A VIOLATION OF A TECHNICAL SPECIFICATION REQUIREMEN THE EVENT OCCURRED AND WAS IDENTIFIED AND REPORTED
. IN THE COURSE OF CONDUCTING AN INTEGRATED LEAK RATE TEST (ILRT). PROGRAM CHANGES MADE IN RESPONSE TO A QUAD CITIES EVENT ENSURED THE IDENTIFICATION OF THE LEA CAUSES THE LEAKAGE FROM THE INBOARD FLANGE OF THE VAL VE WAS DUE TO INSUFFICIENT TIGHTENING OF THE FLANGE BOLTS FOLLOWING THE VALVE REPLACEMENT. THE VALVE REPLACEMENT PROCEDURE DID NOT PROVIDE SUFFICIENT INSTRUCTIONS FOR FINAL TIGHTENING OF THE FLANGE BOLT THE POST VALVE REPLACEMENT TESTING WAS INEFFECTIV :Zf1LD770/:
- CORRECTIVE ACTIONS AFTER THE LEAK WAS IDENTIFIED, PROMPT, COMPREHENSIVE CORRECTIVE AND PREVENTIVE ACTIONS WERE TAKEN. WHILE THE LEAK WAS AN ISOLATED EVENT, POTENTIAL GENERIC IMPLICATIONS WERE CONSERVATIVELY BOUNDED BY CORRECTIVE/PREVENTIVE ACTION POTENTIAL SAFETY SIGNIRCANCE
THE POTENTIAL SAFETY SIGNrFICANCE OF THE AS-FOUND LEAK RATE WAS EVALUATED USING REALISTIC ASSUMPTIONS AND METHODOLOGIES THAT ARE JUSTIFIED BY REFERENCE TO CURRENT REGULATORY PRACTICES. PART 100 AND GDC-19 LIMITS WOULD BE MET, AND CONTAINMENT WOULD HAVE PERFORMED ITS INTENDED SAFETY FUNCTION, GIVEN THE REALISTIC ASSUMPTIONS USED.
ZNL0711)/3
- OVERALL CONCLUSION WE RECOGNIZE THE IMPORTANCE OF MAINTAINING CONTAINMENT INTEGRITY AND TAKE THIS EVENT SERIOUSL THIS PRESENTATION IS NOT MEANT TO EXCUSE THE
. SERIOUSNESS OF THE EVENT BUT RATHER TO PUT IT INTO PROPER PERSPECTIVE. IN THAT REGARD WE HAVE CONCLUDED THAT ALTHOUGH THE EVENT REPRESENTS AN ISOLATED CASE BASED ON OUR HISTORY OF SUCCESSFUL ILRT'S, IT IS AN EVENT DEMANDING PROMPT AND THOROUGH CORRECTIVE ACTIO *
FURTHER WE HAD RECOGNIZED POTENTIAL.WEAKNESSES IN OUR APPENDIX J PROGRAM AND IN 1988 INITIATED COMPREHENSIVE SELF ASSESSMENT EFFORTS WHICH HAVE *
IMPROVED OUR PROGRAM. WE BELIEVE THE NRC ALSO RECOGNIZES THAT IMPROVEMENT. OVERALL, WE BELIEVE OUR PROGRAM IS ADEQUATE AND THAT OUR ONGOING SELF-ASSESSMENT PROGRAM CONTINUES TO BE AN EFFECTIVE VEHICLE FOR FURTHER IMPROVEMENT.
"Z11LD770/4
- INSIDE SUPPRESSION CHAMBER (PRIMARY CONTAINMENT)
Simplified Flow Diagram (2-1601-20A) Leak Path
..,.._ ___ LINE TO A0-2-1601-56 A0-2 LEAK ---.
-1601 LOCATION
\\
-20A
TEST TAP OPEN TO
RX BLD 'PENETRATION
\\_ 2-1601-JlA LINE TO LEVEL TRANSMITTER LT 2-1641-58 LINE TO A0-2-1601-208
REPLACED AIR OPERATED BUTTERFLY VALVES
---~
DRYWELL PURGE FAN 2JAOl
~
...___g;021 __ B __ ~
.-
i)1
...
REACTOR BUILDING I-l/'1 REACTOn BUILDING
- BACKGROUND
THE 18 INCH AND 20 INCH PRATT N2Fll CONTAINMENT BUTTERFLY VALVES HAD A HISTORY OF LOCAL LEAK RATE TES1 (LLRD FAILURES. THE VALVES HAD TO BE RETURNED TO THE VENDOR FOR SEAT REPAIR *
THE EIGHT VALVES FOR UNIT 3 WERE REPLACED WITH PRATT 1200 SERIES VALVES IN JUNE 1988 USING A MAINTENANCE/MODIFICATION PROCEDUR BOLTING INSTRUCTIONS INDICATED "SNUG TIGHT PLUS l /4 TURN".
-
SNUG TIGHT WAS DEFINED ASTHE TIGHTNESS ATTAINED BY A FEW IMPACTS OF AN IMPACT WRENCH OR THE FULL EFFORT OF A MAN USING AN ORDINARY SPUD WRENCH.
Z!ILD770/5
- BACKGROUND CCONTINUED)
- .
A NEW PROCEDURE, DMP 1601-2 "DRYWELL AND TORUS AIR OPERATED BUTTERFLY VALVE MAINTENANCE" WAS IMPLEMENTED IN SEPTEMBER 1988. THIS PROCEDURE ADDRESSED A VARIETY OF VAL VE MAINTENANCE ACTIVITIES, INCLUDING SEAT MAINTENANCE, OPERATOR MAINTENANCE, AND VALVE PACKING, AS WELL AS VAL VE REPLACEMENT. TIGHTENING INSTRUCTIONS INDICATED:
-
"TIGHTEN THE BOLTS IN A GRADUAL OPPOSED PATTERN. SEVERAL PASSES WILL BE REQUIRED TO COMPRESS THE GASKET".
-
THE "SNUG-TIGHT PLUS l /4 TURN" REQUIREMENT WAS NOT INCLUDE Z!lLD/70/6
- EVENT CHROGNOLOGY FEBRUARY, 1989
THE EIGHT CONTAINMENT BUTTERFLY VALVES, INCLUDING THE A02-1601-20A VALVE, WERE REPLACED DURING THE UNIT 2 OUTAGE USING A WORK PACKAGE CONTAINING PROCEDURE DMP 1601-THE WORK PACKAGE ALSO CONTAINED A SHEET WITH THE DEFINITION OF "SNUG-TIGHT".
-
THE VALVE FLANGE BOLTS WERE INSTALLED PER
PROCEDURE AND RELIED ON CRAFT CAPABILIT *
-
THE WORK PACKAGE SPECIFIED AN LLRT BE PERFORMED FOLLOWING REPLACEMENT.
. *
THE LLRT PERFORMED AS PART OF THE POST MAINTENANCE TEST (PMD CHALLENGED THE VALVE SEAT AND OUTBOARD FLANGE, BUT NOT THE INBOARD FLANGE.
ZNLD77f)f7
EVENT CHRONOLOGY <CONTINUED)
FEBRUARY, 1989 (CONTINUED)
THE INBOARD FLANGES WERE NOT PART OF THE STATION LLRT PROGRAM; THEREFORE, NO PROVISIONS OR PROCEDURE EXISTED FOR THEIR TESTIN THE INBOARD FLANGES WERE CHALLENGED DURING PREVIOUS ILRTS. NO LEAKS HAD BEEN IDENTIFIE *
NO ILRT WAS REQUIRED THIS OUTAG DECEMBER, 1990
THE AS-FOUND UNIT 2 ILRT FAILE A LEAK WAS DISCOVERED ON THE INBOARD FLANGED CONNECTION OF A02-1601-20A THAT WAS DUE TO INSUFFICIENTLY TIGHTENED FLANGE BOLT ALL SIXTEEN BOLTS WERE CHECKED FOR TIGHTNES ONLY TWO ADJACENT BOLTS REQUIRED TIGHTENING WITH A SLUGGING WRENCH UNTIL A SNOOP TEST INDICATED ZERO LEAKAG *
THE ILRT WAS SUBSEQUENTLY PERFORMED AND PASSED.
ZIJLD71n,9
- RESULTS OF INVESTIGATION CAUSES OF LOSS OF PRIMARY CONTAINMENT INTEGRITY
THE LEAKAGE FROM THE INBOARD FlANGED CONNECTION OF VALVE A02-1601-20A WAS DUE TO
. INSUFFICIENT TIGHTENING OF THE FLANGE BOLTS FOLLOWING VALVE REPLACEMEN THE PROCEDURE USED DID NOT PROVIDE SUFFICIENT INSTRUCTIONS FOR FINAL TIGHTENING OF THE FLANGE BOLT *
THE LEAK RATE TESTING PROGRAM DID NOT DIFFERENTIATE BETWEEN PMT AND REQUIRED PERIODIC TESTING AND THEREFORE DID NOT ASSURE THE PERFORMANCE OF AN APPROPRIATE PMT ONCE THE FLANGE WAS DISTURBE * THE INBOARD FLANGED CONNECTION WAS NOT DESIGNED FOR LLR ::;uLD77*:*. 9
- CORRECTIVE ACTIONS CORRECTIVE ACTIONS TO ADDRESS BOLTING DEFICIENCIES
ALL FLANGE BOLTS ON THE 18 INCH AND 20 INCH CONTAINMENT BUTTERFLY VALVES ON BOTH UNITS WERE CHECKED AND FOUND TO BE ADEQUATELY TIGHTENE *
A TEMPORARY PROCEDURE CHANGE TO DMP 1601-2 WILL BE MADE PRIOR TO ITS NEXT USE, TO INCLUDE REQUIREMENTS FOR ADEQUATE TIGHTENING OF THE FLANGE BOLT *
THE STATION WORK ANALYSTS' PRE-JOB CHECKLIST WILL BE REVISED BY FEBRUARY 28, 1991 TO ENSURE ADEQUATE GUIDANCE IS PROVIDED IN WORK PACKAGES CONCERNING
BOLTING REQUIREMENT *
-
EXISTING MAINTENANCE PROCEDURES WILL BE REVIEWED FOR ADEQUACY OF BOLTING INSTRUCTIONS AS THEY ARE INCORPORATED INTO WORK PACKAGES.
ZNLL' 1 1('1I10
. CORRECTIVE ACTIONS (CONTINUED)
CORRECTIVE ACTIONS TO ADDRESS BOLTING DEFICIENCIES (CONTINUED)
A NUCLEAR OPERATIONS DIRECTIVE (NOD) ON BOLTING PRACTICES IS UNDER DEVELOPMENT AND WILL BE ISSUED FOR COMMENT BY FEBRUARY 28, 1991 AND ISSUED DURING THE SECOND QUARTER, l 99l. THIS NOD IS BASED ON THE EPRI MANUAL, "GOOD BOLTING PRACTICES".
-
THE INFORMATION IN THE DRAFT NOD HAS BEEN PROVIDED TO MAINTENANCE WORK ANALYSTS TO USE AS GUIDANCE IN WORK PACKAGE PREPARATION PENDING ISSUANCE OF THE FINAL NOD.
ZllLD770/ll
- LEAK RATE TESTING CHRONOLOGY
MARCH, 1986
IE INFORMATION NOTICE 86-16, "FAILURES TO IDENTIFY CONTAINMENT LEAKAGE DUE TO INADEQUATE LOCAL TESTING OF BWR VACUUM RELIEF SYSTEM VALVES", WAS ISSUED. *
-
DRESDEN REVIEWED THE DEFICIENCIES IDENTIFIED IN THE NOTICE. THE CONTAINMENT BUTTERFLY VALVES ARE TESTED LOCALLY BY PRESSURIZING BETWEEN THE ISOLATION VALVES. DURING TYPE "A" ILRT'S, THESE VALVE FLANGES HAD NOT EXCEEDED ALLOWABLE LEAKAG THE CURRENT TESTING METHODS WERE CONSIDERED ADEQUATE AND NO CHANGES WERE INITIATE APRIL 1989
A DRESDEN PRIMARY CONTAINMENT LEAK TESTING PROGRAM ASSESSMENT WAS CONDUCTE THIS WAS A GENERAL REVIEW OF PROCEDURES AND TEST METHODS TO ENSURE CONSISTENCY AT ALL CECO STATIONS.
Z!ILD77Q/1Z
LEAK RATE TESTING CHRONOLOGY (CONTINUED)
JUNE, 1989
IN RESPONSE TO THE APRIL 1989 ASSESSMENT, DRESDEN AND CORPORATE ENGINEERING DETERMINED THAT A MORE DETAILED EVALUATION OF PRIMARY CONTAINMENT PATHWAYS WAS NECESSARY. THE LLRT IMPROVEMENT PROGRAM WAS INITIATE APPENDIX J CORRESPONDENCE WAS GATHERED AND REVIEWE EACH PRIMARY CONTAINMENT PATHWAY WAS EXAMINED, INCLUDING DRAWING REVIEW AND SYSTEM WALKDOWNS, TO IDENTIFY ALL CONTAINMENT PATHWAYS THAT REQUIRE PERIODIC TESTIN AUGUST, 1989
THE BWROG POSITION FROM NED0-31722,
"STANDARDIZED TESTING RECOMMENDATIONS FOR CONTAINMENT INTEGRITY TESTING", WAS ISSUE UNTESTABLE LEAKAGE PATHS SHALL BE EITHER SUBJECTED TO FUNCTIONAL TESTING OR MADE PART OF THE TYPE "A" TEST BOUNDAR AT DRESDEN, THE CONTAINMENT BUTTERFLY VALVES'
INBOARD FLANGES WERE TESTED AS PART OF THE PERIODIC TYPE "A" ILR ZULD11*)/13
- LEAK RATE TESTING CHRONOLOGY (CONTINUED)
DECEMBER, 1989
- *
AS A RESULT OF THE QUAD CITIES PRIMARY CONTAINMENT PATHWAYS EVALUATION, DRESDEN TOOK ACTIONS TO TEST THREE SYSTEMS NOT THEN INCLUDED IN THE DRESDEN LLRT PROGRA JUNE, 1990
THE QUAD CITIES ILRT /LLRT NRC INSPECTION REPORT WAS ISSUE AN UNRESOLVED ITEM WAS IDENTIFIED REGARDING THE REQUIREMENT TO PERFORM lYPE "B" TESTING OF FLANGE ZHLD??0/14
LEAK RATE TESTING CHRONOLOGY (CONTINUED)
JULY, 1990
THE DRESDEN PRIMARY CONTAINMENT PATHWAYS EVALUATION REPORT IDENTIFIED PATHWAYS NOT PREVIOUSLY TESTE COMPLETION OF THE REVIEW IDENTIFIED FURTHER DISCREPANCIES WHICH RESULTED IN TESTING, INTERIM MEASURES, AND A TEMPORARY TECHNICAL SPECIFICATION AMENDMEN PERIODIC TESTING OF THE INBOARD FLANGE WAS EVALUATED AND FOUND TO MEET THE BWROG POSITION THAT THESE BOUNDARIES ARE ADEQUATELY CHALLENGED DURING TYPE "A" TESTING.
ZllL0770/15
.
LEAK RATE TESTING CHRONOLOGY (CONTINUED)
SEPTEMBER, 1990
DURING THE START OF THE UNIT 2 OUTAGE, DRESDEN RECEIVED THE QUAD CITIES RESPONSE TO THE UNRESOLVED ITEM WHICH ADDRESSED TYPE "B" TESTING OF FLANGE *
BASED ON QUAD CITIES UNRESOLVED ITEM, DRESDEN CONCLUDED THE FOLLOWING:
UNIT 2
-
AN ILRT WAS SCHEDULED TO BE PERFORMED
-
DURING THE OUTAGE, NO ADDITIONAL TESTING WAS REQUIRE A REQUIREMENT TO SNOOP TEST INBOARD VAL VE FLANGES WAS ADDED TO THE ILRT PROCEDURE, DTS 1600- *
UNIT 3 ZllLD770/16
-
A ILRT WAS PERFORMED DURING THE PREVIOUS
- OUTAGE. NO ADDITIONAL TESTING WAS REQUIRED.
CORRECTIVE ACTIONS TO ADDRESS LEAK RATE TESTING
TESTING METHODS ARE BEING EVALUATED WHICH WILL ALLOW LOCALIZED PMT OF THE SUBJECT FLANGES. THE EVALUATION WILL BE COMPLETED BY APRIL 15, 199 *
A TEMPORARY PROCEDURE CHANGE TO DMP 1601-2 WILL BE MADE PRIOR TO ITS NEXT USE TO SPECIFY AN APPROPRIATE PMT WHENEVER THE INTEGRITY OF THESE FLANGES ARE DISTURBE *
A COMPREHENSIVE REVIEW OF PRIMARY CONTAINMENT PATHWAYS HAS BEEN COMPLETED AND WILL BE USEDBY *
THE MAINTENANCE DEPARTMENT WORK ANALYSTS TO
- TRIGGER APPROPRIATE LEAK RATE TESTING REVIEW BY THE TECHNICAL STAFF. THIS WILL BE IMPLEMENTED PRIOR TO THE UPCOMING UNIT 3 REFUELING OUTAG *
DAP 14-5, "LEAK RATE TESTING PROGRAM," WILL BE REVISED TO CAUTION AGAINST INAPPROPRIATE APPLICATION OF STANDARD PERIODIC LLRT LINEUPS FOR PMT *
A REVIEW OF IE INFORMATION NOTICE 86-16 WILL BE PERFORtv1ED TO DETERMINE IF ADDITIONAL ACTIONS ARE NEEDE *
A SUMMARY OF THIS EVENT WILL BE SUBMITTED TO NUCLEAR NETWOR Z!ILD17*l/l 7
CORRECTIVE ACTIONS TO ADDRESS LEAK RATE TESTING (CONTINUED)
ON PAGE 12 OF THE INSPECTION REPORT IT IS INDICATED THAT A COMMITTMENT WAS MADE REGARDING TESTING OF THESE FLANGES. BECAUSE THERE MAY HAVE BEEN A MISCOMMUNICATION, WE WOULD LIKE TO SCHEDULE A NRC/CECO TECHNICAL MEETING TO DISCUSS THIS ISSUE.
ALONG WITH THE OUTSTANDING BWROG APPENDIX J PERIODIC TESTING TECHNICAL ISSUES.
Z!lLD770/18
- POTENTIAL SAFETY SIGNIFICANCE LEAK RATE CALCULATIONS
A 24.6WT%/DAY LEAKAGE WAS OBTAINED FROM TEST DATA AT 14.7 PSI *
THE 25WT%/DAY LEAKAGE WAS CONVERTED TO AN EQUIVALENT MASS FLOW RATE USING:
-
(%/DAY)* (Mo/2400) = 437.44 LBM/HR
-
WHERE Mo IS THE TOTAL DRY AIR MASS IN CONTAINMENT AT THE START OF THE CALCULATION *
- - TECHNICAL SPECIFICATIONS REQUIRE A 48 PSIG PRESSURE
DIFFERENTIA *
EXTRAPOLATION FROM 14.7 TO 48 PSIG IS NECESSAR *
THREE POSSIBLE TYPES OF FLOW WERE CONSIDERED, LAMINAR, TURBULENT, AND CHOKED ORIFICE.
ZUL077Q/19
POTENTIAL SAFETY SIGNIFICANCE <CONTINUED)
LEAK RATE CALCULATION CCONTINUED)
THE EXTRAPOLATED LEAKAGE RATE FOR EACH TYPE OF FLOW IS SHOWN BELOW:
FLOW MULTIPLIER LEAKAGE RATE %/DAY
@48 PSIG CHOKED 2.137
TURBULENT 2.65
LAMINAR 5.762
- CECO DETERMINED THAT FLOW WAS CHOKED (TURBULEN THE REYNOLDS NUMBER CALCULATED FOR THIS EVENT WAS ON THE ORDER OF 10 REYNOLDS NUMBERS GREATER THAN 2000 EQUATE TO TURBULENT FLO THE SIREN-LIKE SOUND OF THE LEAK WAS INDICATIVE OF TURBULENT FLOW, SPECIFICALLY SONIC FLOW.
ZULD770/2Q
POTENTIAL SAFETY SIGNIFICANCE CCONTINUED)
LEAK RATE CALCULATION (CONTINUED)
FRANKLIN RESEARCH AND NRR HAVE USED A LAMINAR CORRELATION FOR EXTRAPOLATION OF MSIV AND AIR LOCK DOOR LEAKAGE RATES BECAUSE THESE PATHWAYS HAVE LEAKAGE ACCEPTANCE CRITERIA LESS THAN 20 SCF LAMINAR FLOW MAY BE EXPECTED FOR SUCH SMALL LEAK *
WE BELIEVE THIS CORRELATION IS INAPPROPRIATE IN THIS CASE BECAUSE:
-
THE LEAKAGE WAS APPROXIMATELY 12,000 SCF CONCLUSION
- BASED ON CHOKED FLOW, WE CALCULATED THE LEAKAGE RATE TO BE APPROXIMATELY 25%/DAY
- BASED ON TURBULENT FLOW, THE LEAKAGE RATE WAS 31%/DA *
DOSE RATE CALCULATIONS HAVE BEEN DETERMINED CONSERVATIVELY USING 31 WT°lo/DA Y.
WLD170/21
POTENTIAL SAFETY SIGNIFICANCE (CONTINUED)
DOSE RATE CALCULATIONS QUESTIONS FOR ANALYSIS: WOULD CONTAINMENT HAVE PERFORMED ITS INTENDED SAFETY FUNCTIONS IN THE EVENT OF A DESIGN BASIS LOCA WITH THE CALCULATED AS-FOUND LEAK
- RATE OF 31.0%/DAY?
r
.. l\\LCULATIONS WERE PERFORMED USING CURRENTLY ACCEPTED METHODOLOGIES (I.E. REGULATORY GUIDE AND STANDARD REVIEW PLAN SECTION 6.5.5), KNOWN STATION OPERATING DATA AND CONSIDERATION OF HOLD-UP OF ACTIVITY IN THE SECONDARY CONTAINMEN *
THE RESULTS INDICATE THAT 10 CFR 100 AND GDC-19 LIMITS WOULD HA VE BEEN MET.
POTENTIAL SAFETY SIGNIFICANCE <CONTINUED)
DOSE RATE CALCULATIONS CCONTINUED)
SITE BOUNDARY & LOW POPULATION ZONE
. 10 CFR 100 LIMIT THYROID CREM)
300 WB CREM)
SITE BOUNDARY DOSE W /
SRP 6. LOW POPULATION ZONE DOSE
.W/ SRP 6..6 3 INPUT PARAMETERS.2
- CONTAINMENT LEAKAGE @ 31.0%/DA Y FOR 30 DA VS
- HOLD-UP OF ACTIVITY IN SECONDARY AT 1 VOLUME/DAY (5/11/83 SER)
- 98% SBGTS EFFICIENCY (TEST RESULTS)
- ALL OTHER INPUT PARAMETERS IDENTICAL TO 1982 NRC DRESDEN ACCIDENT ANALYSIS CONCLUSION: PART 100 LIMITS WOULD BE MET, AND CONTAINMENT WOULD HAVE PERFORMED ITS'
INTENDED SAFETY FUNCTION, GIVEN THE REALISTIC ASSUMPTIONS USED.
- POTENTIAL SAFETY SIGNIFICANCE CCONTINUED)
DOSE RATE CALCULATIONS CCONTINUED)
. THYROID CREM)
SKIN CREM)
WB CREM)
GDC-19 LIMIT CONTROL ROOM DOSE
. WI SRP 6..7 INPUT PARAMETERS
2.4
- CONTAINMENT LEAKAGE @ 31.0%/DA Y FOR 30 DA VS
- HOLD-UP OF ACTIVITY IN SECONDARY @ 1 VOLUME/DAY (5/11 /83 SER)
- 98% SBGTS EFFICIENCY (TEST RESULTS)
- 40 MINUTE ISOLATION (REANALYSIS, FSAR)
CONCLUSION: GDC-19 LIMITS WOULD BE MET, AND CONTAINMENT WOULD HAVE PERFORMED ITS'
INTENDED FUNCTION, GIVEN THE REALISTIC ASSUMPTIONS USED.
- CONCLUSIONS
FAILURE TO MAINTAIN PRIMARY CONTAINMENT INTEGRITY WAS A VIOLATION OF A TECHNICAL SPECIFICATION REQUIREMEN *
THE EVENT OCCURRED AND WAS IDENTIFIED AND REPORTED DURING THE UNIT 2 ILRT. PROGRAM CHANGES MADE IN RESPONSE TO A QUAD CITIES EVENT ENSURED THE IDENTIFICATION OF THIS LEA *
- AFTER THE LEAK WAS IDENTIFIED, PROMPT, COMPREHENSIVE
. CORRECTIVE AND PREVENTIVE ACTIONS WERE TAKEN TO ADDRESS BOLTING AND LEAK RATE TESTING ISSUE *
-
FLANGE BOLTS WERE VERIFIED TO BE TIGH THE WORK ANALYSTS' CHECKLIST WAS REVISED FOR THE UNIT 3 OUTAGE TO ENSURE ADEQUATE BOLTING GUIDANCE IS PROVIDED IN WORK PACKAGE THE COMPREHENSIVE REVIEW OF PRIMARY CONTAINMENT PATHWAYS WILL BE USED BY THE WORK ANALYST DURING THE UPCOMING UNIT 3 OUTAGE TO TRIGGER APPROPRIATE LEAK RATE TESTING REVIEWS BY THE TECHNICAL STAF THE LEAK RATE TESTING PROGRAM IS BEING REVISED TO CAUTION AGAINST THE APPLICATION OF STANDARD PERIODIC LLRT LINEUPS FOR PMT Z!ILD770/:S
CONCLUSIONS CCONTINUED)
WHILE THIS LEAK WAS AN ISOLATED EVENT, POTENTIAL GENERIC IMPLICATIONS WERE CONSERVATIVELY BOUNDED BY THE CORRECTIVE/PREVENTATIVE ACTION *
THE POTENTIAL SAFETY SIGNIFICANCE OF THE AS-FOUND LEAK RATE WAS EVALUATED USING REALISTIC ASSUMPTIONS AND METHODOLOGIES THAT ARE JUSTIFIED BY REFERENCE TO CURRENT REGULATORY PRACTICES.. PART 100 AND GDC-19 LIMITS WOULD BE MET, AND CONTAINMENT WOULD HAVE PERFORMED ITS INTENDED SAFETY FUNCTION, GIVEN THE REALISTIC ASSUMPTIONS USED.
Z!ILD7 7Q/2 ~
""**,
Enclosure 3 PRIOR RELATED VIOLATIONS CONTAINMENT INTEGRI1Y. POST MAINTENANCE, OR POST-MODIFICATION TESTING ZION BYRON FAILED TO ADEQUATELY TEST MODIFICATIONS. (1!24/89)
POST-MOD TEST PERFORMED DID NOT CONFIRM ACCEPTANCE CRITERIA ME'[
(NCV) 4/13/89 DRESDEN FOR CRD CHARGING WATER CHECK VALVES, INADEQUATE TEST PROCEDURES. (4/21/89)
DRESDEN BATTERY TESTING NOT IN ACCORDANCE WITII POST-MOD TEST-6 EXAMPLE (5/8/89)
DRESDEN VENTILLATION HATCHES IN U-2 DRYWELL LEFf IN AN IMPROPER POSITION DUE TO INADEQUATE MAINTENANCE & IMPROPER SURVEILLANCE. (7/31/89)
ZION UNTIMELY CORRECTIVE ACTION ON KNOWN PROBLEMS IN POST-MAINTENANCE TESTING, TEMPORARY MODIIFICATIONS AND WORK CONTROL. (9/1/89)
ZION FAILURE TO DO A POST-MOD OR POST-INSTALLATION TEST RESULTING IN NORMAL POWER BEING LOST TO ALL U2 ANNUNOATORS. (9/'l1J/89)
LA SALLE NO TESTING PERFORMED AFTER REINSTALLATION OF A RELAY FOLLOWING CALIBRATION. (11/20/89)
LA SALLE FAILURE TO TEST "B" DIESEL FIRE PUMP AFTER CORRECTIVE MAINTENANCE PERFORMED. (5/'l3!90)
ZION POST-MAINTENANCE TESTING SPECIFIED IN WORK REQUESTS STILL IACKED ACCEPTANCE CRITERIA. * (6/14/90)
BRAIDWD. FAILED TO TEST AIRLOCK WITHIN 3 DAYS AFTER CONTAINMENT ENTRY. (7/18/90)
DRESDEN LOSS OF CONTAINMENT INTEGRITY DUE TO PROCEDURAL CHANGE WITHOUT 5059 REVIEW (11128/90)
DRESDEN PERFORMANCE OF POST-MAINTENANCE TEST ON 12/8/90 RESULTED IN SEVERAL U2 GROUP II ISOIATION VALVE CLOSURES. (1117/91)
'\\'\\..._ ________________________________________________________________......