BSEP-92-0045, Responds to NRC Ltr Re Violations Noted in Insp Repts 50-324/92-27 & 50-325/92-27 on 921026.Corrective Actions: Sampling of Reverification Program Has Been Implemented to Justify Insp of Welds W/More Slag Remaining in Place
| ML20117A537 | |
| Person / Time | |
|---|---|
| Site: | Brunswick  |
| Issue date: | 11/25/1992 |
| From: | Morgan R CAROLINA POWER & LIGHT CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| BSEP-92-0045, BSEP-92-45, NUDOCS 9211300126 | |
| Download: ML20117A537 (21) | |
Text
. _
Carolina Power & Light Company Ilrunswick Nuclear Project P.O. Ilex 10429 Suthport, North Carolina 28461 November 25,1992 SERIAL: BSEP-92 0045 United States Nuclear Regulatory Commission ATTENTION: Document Control Desk Washington, DC 20555 BRUNSWICK STEAM ELECTRIC PLANT, ' JIT IJOS.1 AND 2 DOCKET NOS. 50-325 & 50 324/ LICENSE NOS. DPR 71 & DPR-02 REPLY TO NOTICE OF VIOLATION Gentlemen:
On October 20,1992, the Nuclear Regulato;y Comraission issued a Notice of Violation for the Brunswick Steam Electric Plant, Units 1 nnd 2. Details of the underlying NRC inspecticns are provided in inspection Report Nos. 50 325/92 27 and 50-324/92-27 dated October 2f,,1992.
Carolina Power & Light Company hereby responds to the Notice of Violation. Enclosu re 1 to this letter piovides CP&L's reply to the Notice of Violation in accordance with the provisions of 10 CFR 2.201. Enclosure 2 provides CP&L's reply to the inspection Report's request for additional information.
Please refer any questions roarding this submittal to Mr. S. D. Floyd at (919) 457-2404.
Yours very truly,
/
3 R. E. Morgan Interim Site Manager, Brunswick Nuclear Project GMT/gmt (\\wpiea92027)
Enclosures cc:
Mr. S. D. Ebneter Mr. R. H. Lo Cr. R. L. Prevatto k
/[, \\
9211300126 921125 PDR ADOCK 05000324 G
=
Document Control Desk NLS 92112 / Papa 3 bec:
Mr. M. D. Bradley (BNP)
Mr. E. M. Northeim Mr. R. K. Buckles (LIS)
Mr. D. B. Waters Mr. S. H. Callis (Bf4P)
Mr. H. A. Pollock Mr. M. S. Calvert (BNP)
Mr. R. 8. Richey Mr. R. M. Coats Mr. R. E. Morgan (BNP)
Mr. S. D. Floyd (C'JP)
Mr. J. M. Brown (BNP)
Mr. J. L. Harness Mr. R. S. Stancil Mr. E. M. Harris Mr. R. B. Starkey, Jr.
Mr. J. D. Heidt Mr. G. H. Warriner (DNP)
Mr. W. G. Hindman (2)
Mr. R A. Watson-Mr. E. E. Willett (BNP)
File: OC/A-4 Mr. A. M. Lucas File: B-X-0544.1 File: B-X 0544.2 1
6 4
'b t
2 e
h i
0 1
4
~~,.-..i,,,..-v
-..,w.,*
e.
~,-.
,.--.Yw,.-
-., ~..-ro,.,......wi.-U.,...-v.,
-.,.n.,,,
ENCL,OSURE 1 BRUNSWICK STEAM ELECTRIC PLANT, UNITS 1 AND 2 NRC DOCKET NOS. 50-325 & 50 324 OPERAT!NG LICENSE NOS. OPR 71 & DPR 02 REPLY TO NOTICE OF VIOLATION Mt.AT10N:
I During an NRC insraction conducted on September 14-18 and September 24 25, 1992, a violation j
of NRC requiremer.o was identified. In accordance with the " General Statement of Policy and Procedure for NRC Enforcement Actions," 10 CFR Part 2, Appendix C, the violation is listed below:
10 CFR 50, Appendix B, Criterion V, requ'.<es that activities affecting quality be prescribed -
by documented instructions, procedures, or drawings, of a type appropriate to the circumstances and shall be accomplished in accordance with these instructions, procedures, or drawings. Procedures and instructions were either not appropriate (adequate) or were not followed for inspectic. of structural steel in the drywell and for performance of structural steel design acdvities as described below.
1.
Paragraph 11.2.3.1 of Bechtel Procedure WDP-002, Phase ll Walkdown Procedure for Reactor Building Miscellaneous Steel and Drywell Platform Steel, requires Phase 11 inspection personnel to evaluate welds using criteria specified in Appendix A of the procedure. Appendix A requires that weld size and all weld attributes be verified by welding engineers.
Contrary to this requirement, Phase 11 welding engineers classified wolds at connections B1B and 84B on the elevation 17'-10 1/4" drywell platform, azimuth 99 deg. to 122 deg, incorrectly as partial penetration welds The correct classification for these welds in fillet.
- ]
1 2.
Paragraph 11.2.4.1 of procedure WDP 002 requires the Phase il walkdown
)
l personnel to compare the number of bolts in each connection with design information and record differences on Exhibit G in the walkdown documentation for each connection.
Contrary to this requirement, Phase 11 walkdown personnel failed to identify and document on Exhibit G a missing bolt in connectir 1 B 5A on the elevation 17',0 1/4" drywell platform, azimuth 99 deg. to 122 deg.
3.
Appendix A to Procedure WDP-002 requires welding engineers to document all weiding attributes on Exhibit A 1.
Contrary to this requirement, the welding engineers failed to document that welds at connection numbers 03b and BBB at azimuth 270 deg. to 349 deg. and welds at -
connection numbers B4A and B3B at azimuth 90 deg. to 157 deg. o, elevation 80 drywell platform were covered with slag, and that inspection of these welds for cracks, lack of fusion, and other irregularities had not been performed.
l~
E1 1 yt'+-w,e
-gi--
erw--me-me a
- 't'-
- -+
it v'-N
- us8 T'
- G" f't"T*F'99'1 M-w ser'e ' T 4m i-
's$r'7~
--Wv 2 eMMg-
- W
@_wv-y-nWW'wm+-7y-wTi't-i-y-ia*dge tw Fve-F
='A--99'*--f-
-s="Y'*
h w
4.
Bechte! Procedure EDP 4.27, Design Verification, and EDPI 4.37, Design Calculations, requirs design calculations be checked to verify the calculations are correct and accurate.
Contrary L, snese requirements, the calculation checkers f ailed to identify an error in the weld length on page 16 of 47 in calculation number 2RB2+1113, and the failure i
to perform evaluation of the irregularity at connection number B78 in package 2-RB-i D E160-1 (P-S/21R 22R) in calculation number 2RB21010.
5.
The welding inspection instructions in procedure WDP-002 were not appropriate to accomplish visualinspections in accordance_with referenced NCIG Visual Welding Accep*. ace Criteria (VWAC). The procedure permitted inspection of welds covered with excessive slag and acceptance of groove welds with five percent lack of fusion. VWAC requires removal of slag to perform visualinspections and permit 0 percent lack of fusion in groove wolds.
This is a Severity Level IV violation (Supplement 1).
BESPONSE TO VIOLATIOB:
Mg)]jision or Denial of Violation:
Carolina Power & Light Company admits this violation, t
5 i
i El 2 we.=
_..,y..i TFe-
-F--DT'8'CM' W-F
'***W-
""+18*"*v3.ag.-m--e g mpvv y--
g-e n r-
+em gi rv gemw g 'uoy (-
f y
y
RESPONSE TO SPECIFM VIOLATION ISSUES lssue No.1 Paragraph 11.2.3.1 of Bechtel Procedure WDP-002, Phase 11 Walkdown Prucedure for Reactor Building Miscellaneous Steel and Drywell Platform Steel, requires Phase 11 inspection personnel to evaluate welds using criteria specified in Appendix A of the procedure. Appendix A requires that weld size and all weld attributes be verif'ed by welding engineers.
Contrary to this requirement, Phase 11 welding engineers classified welds at connections B1B and 34B on elevation 17' 10%* Drywell platform, Arimuth 990 to 1220, incorrectly as partial penetration welds. The correct classification for these welds is fillet.
A.
Resoonse To Violati.qa Bechtel Procedure WDP-002 was specifically developed for conducung Phase it walkdown activities for the Miscellaneous Steet Verification Program (MSVP) at the Brunswick Steam Electric Plant (BSEP). This walkdown procedure, of which Revision 2 was in effect at the time of the NRC inspuction, provides appropriate technical and quality assurance program requirements for the walkdown activities.
Appendix A of WDP-002, which is based on EPRI Document NP 5380, Visual Weld Acceptance Critoria (VWAC), specifically interprets VWAC and provides requirements to be used at BSEP for inspection of existing, completed welds. However, VWAC and Appendb, A do not specifically address criteria for inspection of partial penetration welds. In addition, neither discusses joint preparation verification which is an in-process inspection item.
However, Appendix A does allow classifying appropriate welds which cannot be verified at "not obtainable (NIO).* Appendix A of WDP-002 provides two basic options to the welding engineers.
- 1) Identify the weld (using a reasonable best effort basis) and confirm the identification with the second welding engineer in the team. This additional check (second welding engineer) is in excess of VWAC requirements, or
- 2) classify the weld as inaccessible for verifiotion.~
The welds in question are diffluult to assess due s ih tight constraints of the joint. Special I
measures, including puint removal, were required to kv.ifirm that two of the six partial penetration welds questiened by the NRC duieng the site visit were essentially filiat welds contrary to the originalinterpretation by the welding engineers. The welds, having no other visible irregularity, were judged to be partial penetration welde consistent with the design-information, and not fillet welds. Consideration should be given to the degree.of difficulty involved in the verification; in addition, both par 6al penetration welds (with caps) and fillet -
we'ds have similar external attributes.
We believe the procCure was appropriate for anticipted conditions and reasonable efforts were exercised in interpreting and evaluating the in-place welds under difficult conditions.
El-3 e
J This process provides high conNdence that the judgement applied provides appropriate verification.
D.
C.grrective Actions. Steos Taken. anifklutt Achieved To strengthen the ability of welding pctsonnel to validate weld configurations and joint verification, Appendix A was revised to permit additional usage of chipping hammers and other common tools for additional partial paint or slag removal. This process adds selective flexibility to the welding inspection process; however, some constraints exist for ALARA considerations and minimization of airborne contamination. Weld attributes, which cannot be confirmed, continue to be classified as inaccessible.
To assure that appropriate action was taken on the partial penetration weld population, additional reviews were conducted. A total of 107 connections wh'ch appeared to have partial penetration welds were identified in the Unit 2 Drywell. Of this population,52 connections are of types that havc been addressed, or will be addressed without taking any uedit fo< partial penetration welds. For the remaining 55 joints, a reinspection program was conducted to assess joint preparation to confirm the partial penetration weld quality.
Consistent with the expected difficulty of this insnection, none of the welds could be totally
\\
verified. Some appeared to be partial penetration, some appeared to have a level of reinforcing fillet, and others could not be classified beyond the visible fillet welds.
As a conservative action, CP&L determined that a restoration of the welds to their original equivalent capacity (generally with fillet welds), conservatively noi;lecting the unverifiable partial penetration (or other) welds, would be completed prior to start-up to close the issue.
This action, when completed, will positively and conservatively address De structural acceptability of the partial penetration weld issue.
C.
CorredignLio Avoid Further Violations The revised WDP-002 Appendix A (Revision 3) was issued October 26,1992 to specifically require verification of significant weld attributes, and specifi: resolution where engineering evaluations require validation of structuralintegrity. Welds that cannot Le confirmed will f
continue to be classified as "NIO" (not obtainable). Engineering will not use weld capacity of unverified partial penetration wilds unless a fillet weld is added for load capacity.
Evaluations for this condition will be based on the verified fillet weiu capacity only.
D.
Date of Full Comoliance Modification sketches for partial penetration welds have 17en issued and reinforcing welds have been completed to conservatively ensure the origina :esign requirements. WDP-002 was revised on September 8,1992 and October 26,1992 with enhancements and agreements recommended in meetings with the NRC. Rcsponsible personnel have been trained in the revised requirements and WDP-002 is in compliance with NRC reauirements.
J El-4 l
Llilus No. 2 Paragraph 11.2.4.1 of Procedure WDF-002 requires the Phase ll walkdown personnel to compare the number of bolts in each connection with design information and record differences on Exhibit G in the walkdown documentation for each connection.
Contrary to this requirement, Phase 11 walkdown personnel failed to identify and document on Exhibit G a missing bolt in connection B-5A on the elevation 17' 10%" Drywell platform, Azimuth 99 to 122".
A.
Snagme To Violation Issue A walkdown inspection by the NRC of an in-process work package identified a tangential beam connection at Azirnuth 99' at elevation 17'-10%" (Drywell) with a bolt missing. This missing bolt had, at the time of the NRC walkdown, not been documented in the work package.
The walkdown process incorporates specific steps to match the data sheets and the photos and this process is formal, documented, and uses a flowchtri and routing slip. These reviews, closely comparing the photos and data sheets, had nct toen nmpleted on the subject package pending receipt of the developed photos. The photos were received on the day of the NRC walkdown, the missing bolt was noted, and the data sheet was revised accordingly.
11 should be recognized that the subject package was selected for the NRC walkdown inspection to minimize the need for HP coverage and grating removal, even though it was an in process package.
B.
Corrective Eleos Taken and Results Achieved Review of the package was completed using the photos and the existing defined process.
The data sheet was modified to identify the missing bolt and the photographs were included in the package.
Package reviews have been completed and the package has been issued for analysis. No program changes were dee.ned necessary since the package was in-process and the documentation process used by the program made and documented the tecnni9al corrections. This process readily identifies isolated discrepancies with high confidence levels.
The package identified was completed in the normal review and documentation cycle anri issued for use.
C.
Corrective Steos to Avoid Further Violations As added emphasis, additional instructions were provided verbally during group meetings of walkdown personnel reinforcing the process to reconfirm data sheets with the photos prior to sign-of f.
Since the waldowns require collection of an extensive amount of data, many times in unfavorable conditions, the walkdown procedure was developed to prove multiple leveis of checks and revie.cs to ensure accuracy of the data. This includes walkdowns by two-person El 5
teams, reviews of packages including comparisons with and use of photographs by walkdown and engineering personnel, independent verifications, and audits and surveillances, by both CP&L and Bechtel personnel. Thus, the walkdown procedure process provides sufficient controls to minimize errors in the walkdown data and provides for continuous imorovement through feedback, additionallevels of training, and procedure revisions.
D.
Date of Full Comoliance The program is in compliance with these NRC requirements.
E1-6
wr u
~
+
' GLN2J Appendix A to Procedure WDP 002 requires welding engineers te ument all wel@g attributes on Exhibit A-1.
y3
.<y is this requirer:mt, the weldirig engineers fa: led to do: a nent that welds at conno-Von
~ M and 88B at Azimuth 270' te 349' :.nd welds at connection numbars B4A and 838 s
I O' to 157" on the elevation 86 Drywell platferm w0re covere 1 with slag, and that
,9 -
- these welds for cracks, lack of fusion, and other irregularities had not been g -
s.
}
re' r r
aconse To Violation r
Mt. Il V.atkdown Procedure V.T K002 was specifically developed fo conducting Phase il
=
wardowns f ar the Miscellancous St nel Verification Program (MSVP) at BSEP. It has been cr the practice of the project team to share draft versions of this procedure with NRC representatives during the course of this project. This process is consistent with CP&L's come..ument to permit adjustments to the program to meet changing requirements as the (f
work progressed. This approach, outlined in the Jaly 27,1992 submittal (Serial NLS-92 J
203), commits to continued strengthening of the program wre u cerience is obtained
.i which offers opporturd y for improvement. The <
' aad enVo monts to WDP-002, t
incluclog Appendix A, should be dewed in this c.
Appendix /, v s N on the guidelines of VWAC and identified the weld verification criteria b Je used fo.
. ?. During the course of prograrn developroent, with special ccncern for LARA and airbcrne contam: nation on a large scale, the enrineering team and technical t
specialists cDblished a verification process using the best Jetermination possible without significant removal of coat:ngs or slag. While both objectives are importaat, they w.. 'k against each other and a balanced approach was s,lteted. For these reasons, slag te-oval was not identified as a requirement in Appencu A and the strict adoption of VWnt was not considered appropriate for thi* rpplication, large amounts of surface slag were not anticipated in the. plant during the development of the acceptance criteria.
The program as defined balances severalimpottant program objectives and the presence of g
j =
slag in this application has not been a major detriment in overall assessment weld quality, At elevations 38' 5" and 17'-10", slag has been confirmed to be a non-issue or aeavily f
loaded connections. At elevation 80'-0", slag is evident in greater amounts; hove /er, these platforms are minor and Whtly loaded. The issue of inspecting welds with surface slag present was presented to TechNs. Advisory Committee (TAC)in their August 27-28,1992 meeting. It was suggested that technical reasons be assembled to provide a basis for 1
acceptabihty. The technical reasoning was assembled and became a part of Appendix B to WDP-002 issued in Revision 2 on September 8,1992. The inspect:ng of welds with slag was piesented again to TAC in the September 17,1992 meeting (after 3e NRC mentioned it as an issue) and in the Govember 13,1992, meeting TAC concurred that the results of the slag inspection program provide an adequate basis for concluding that surface slag has fiot been s major factor in inhibiting sound weld verification in this application.
B.
Corrective Actions. Steos Taken, and Results Achieved El-7
Several walkdowns by Bechtel welding specialists cor. firm that generally, at connections where slag has been removed, welds are of good quality.
However, recognizing the continued goal to enhance the program, WDP 002 and Appendix A were revised on September 8,1992 to include selected discretionary sicg or paint removal to strengthen validation of weld inspection. This additional flexibility offertd the welding inspector does not connote extensive slag removal that would be detrimentalit, ALARA and increase airborne contamination.
C.
Corrective Steos to Avoid Further Violations Th snhancement to Appendix A of the walkdown pro':edure specifically allows the removal of paint and/or slag as judged necessary by the weld evaluator to confirm weld quality. As part of the procedure revision process, welding engineers performing the inspections have been retrained to the enhanced provisions allowing them to remove slag, as required, to comp'ete the inspection.
h' A sampling and reverification program has been implemented to justify inspection of welds k'
with more slag remaining in place than is permitted by VWAC, and to demonstrate that the presence of slag does not adversely affect the weld eva vations, This program is discussed i
in issue 5.
D.
Qate of Full Comoliance The program is in compliance with these NRC requirements and is to be validated by the slag samplin; program discussed in issue 5.
El-8
luue No. 4 Bechtel Procedure EDP-4.27, Design verificatian, and EDPI-4.37 01, Design Calcelations, require j
design calculations be checked to verify the calculations are correct and accurate.
Contrary to these requirements, the calculation checknrs failed to identify an error in the weid length on page 16 of 47 in calculation number 2RB_1113, and the fa:
' to perform evaluation of 1
the irregularity at connection number B8 in package 2-GO-E100-1 (I.
1R 22R) in calculatica r umber 2RB2-1010.
j l
Note: Calculation 2RB2-1113 (Reactor Building calculation number)is for the Dryweil and subsequent to NRC review, the calculation number was revised to 2 RIP 1017 +o be consistent with the CP&L numbcnng system for Drywell calculations. The correct walkdown package number correspond %g to calculation 2RB21010 is 2 RB-D-EL60-1 (P-S/21R-22R).
A.
Eqsoonse To Violatica Issue fring an NRC site visit to walk down structural steel welding and review the procedures for v91 ding related issues, an NRC inspector reouested a separate meeting to roview the process ur d for the disposition of welding irregularities, including weld derating, and methodology.
The meeting was held and Bechtel described the interfacing processes between walkdown data collection, weldin0 nspection, document cor (ol, anc procering through the evaluation i
group. In the course of these discussions, the NRC inspector reqvsted a listinn <
1 calculations and their wtus 9echtel clearly indicated that al! evaluation work was in-procees, that some of the work had been checked, but none had been design-verified or reviewed by the evaluation supervisoc. None of the packages had been released to CP&L for an Owner's Review. The NRC inspector asked if he could review some of the in-process calculations.
The two errors identified were among at ' east 22 calcuhtion issues boing addressed in the two documents. The specific details are as follows:
The error associated with calculation 2RB21113 was a transposition of weld length, the l
actual condition (4" weld) is clearly stated 'n ine calculatlan sheet. However, the welded member length (5" clip angle) was erroneously used in computing the weld properties. The calculation errc s pon correction indicates the connection capacity te be 50% above its loading.
I The error associated with calculatior: 2RB21010 was not specihcally addressing the weld irregularity for the clip angle attacament to the web of the connecting bearn (very.
j similar weld configuration / loading to that of the clip angles to the embed plate which l
was addressed in the calculation). This calvJiation error upon correction indicates the connection capacity to be 260% above its loading-In summary, while the errors identified are acknowledg6e the calculations were in-process.
The overa!! assessment of quality, completeness, and con.pliance to the procedures had not been performed. Correction of these errors has not changed the acceptability of the irregularities.
El-9 l-
=
4 B.
Cpngetive Stoos and Results Achievnd Both calcuiations have been revised, design verified, and signed by the Bechtel engineering supervisor. Both have significant margins and are a':ceptable in their final form.
The following process changes and additional steps have been instituted to ensure and enhance quality:
Irregularity evaluation calculations required for Unit 2 restart, which were originated and checked prior to October 2,1992, were reviewed in depth and data was obtained to establish the overall quality of the calculations. Pareto charts were developed to ihtify areas deserving attention and training sessions 'ield on those areas where enhancements -
could be realized. CP&L personnel were involved to ensure that there is a clear understanding of what is expected in terms of format, preferences, and level of detail in providing justifications for conclusions reached (especially where engineering judgement is involved). Selected improvements in the work procuses were incorporated to assess the quality of work in various in-process stages. In addition, weekly meetings on technical and proceduralissues have been instituted to provide a vehicle to share contemporary technical issues and resolutions among the engineering evaluation team members.
The ca'culations are proceeding with the Bechtel Engineering Supervisor's review, design verification, and administrative review. Additional reviews are being perfotmed by the Bechtel Chief Civil / Structural Engineer's office, and CP&L on an ongoing basis to further ensure a consistent and high quality product is being provided.
When prr act calculation quality issues we s raised by the NRC and CP&L, an a>sessment c
was perfonned with the goal of developing a specific calculation quality improvement plan.
A number of factors wara identified which adversely affected quality. These can be grouped into two fundamental causes that, taken together, identified the corrective actions that were subsequently taken.
Evolving project technical approaches and methodologies Technical prob! ems were identifi 1 as work progressed that required expert judgement and guioance to the l-conditions being evaluated for Brunswick. ~As a consequence, initial calculatioas were started without sufficiently detailed guidance and clear definition or expectations.
Rapid manpower id! dup for a demanding t,chedule - Staffing of engineers progressed at L
-a rate that (1) out-paced physical resources resulting in a difficult working environment p
and (2) created a less than-ldeal supervisory span of control, in order to validate these original causes and identify any further specific causes, collection of data on calculation errors or discrepancies has been initiated, The discrepancies are being l:
~ categorized to allow rapid identification of trends to address in a continuing improvement plan.
The first set of evaluation calculation was formally issued by Pechtel to Cf'&L in November j
1992. The finalized calculations are in compliance with applicaule project procedures, t
i El-10 l
.. ~~ _
C.
Corrective Steos to Avoid Further Violations 4
2
. Wsth the continued application of the project review process prior to issuance of calculations
- (checker, Engineering Supervisor'e.eview, design verification, and administrative review),-
gathering data and evaluating it in relation to quality issues, providing training as required, and use of the Bechtel Chief Engineer's reviews and CP&L owner's, reviews es deemed appropriate, discrepancies will be minimized.
j D.
Qa),q,of Full Comolianet The Miscellaneous Steel Verification Project (MSVP) including the engineering evaluation process is in compliance with these NRC requirements.
l I
6 l
P E1-11 4
m m
v rr
-..wa w r Inue No. 5 The welding inspection instructions in procedure WDP-002 were not appropriate to accomplish visualinspections in accordance with referenced NCIG Visual Welding Acceptance Criteria (VWAC).
The procedure permitted inspection of welds covered with excessive slag and acceptance of groove welds with five percent lack of fusion. VWAC requires removal of slag to perform visual inspections and permit 0 percent lack of fusion in groove welds.
A.
Resoonse To Violation Issue The propcsed violation issue presumably is based on the assumption that NCIG VWAC criteria was the referenced criteria and basis for weld inspection in WDP-002. While many of the provisions of VWAC were accepted, selected deviations wer3 adopted to maet the specific conditions to minimize airborne contamination and exposure to personnel. Two contributions form the basis of the above violation issue:
- 1) Acceptance of groove welds with up to 5% lack of fusion.
- 2) Complete removal of slag.
The allowance of 5% lack of fusion for groove welds was introduced in Revision 2 of the procedure and has since been removed by Revision 3. As discussed with NRC representatives, this provision was never used and does not impact the data collacted,-nor the waluation of an, welds. Panial penetration groove welds have been upgrad< cr reconfirmed as described in issue No.1 as a conservative action to improve quality by CP&L.
Regarding the second contributor, the issue of appropriate levels of slag removal hac received extensive review by the project team, Bechtet and CP&L staff technical specialists, Bechtel and CP&L staff welding specialists, the Technicci Advisory Committee (TAC) (with NRC -
~
participation), and significant on site discussions with NRC representatives. The selective chipping as determir'ed by the welding inspector, while nutigating the need for extensive paint and slag removal, has been clearly outlined in the program, in addition, this approach I
meets the program objectives, and the slag removal exception to strict VWAC adoption is important and valid.
B.
Corrective Actions; Steos Taken. and Rerytts Achieved Enhancemot of WDP-002 has continued as a result of ongoing walkdown experience, j
technical interfaces, irnerfaces with the TAL, and interaction with the NRC. As indicated in
- CP&L's response to issue 3, item B,' Appendix A was revised to permit selected discretionary slag or paint removal to strengthen the weld verification prccess. While this revision it. not completely compatible with all VWAC requirements it provides a balanced approach which properly address the ALARA and contamination concerns. WDP-002 revision 3 also added Appendix C which more clearly described tt.e specific guideline to be used for evaluation and disposition of welding irregularities.
El12
C.
CarLective Actions to Avoid Further Violations The enhancements made in revision 3 of WDP-002 should eliminate future problems. The prccedure neither accepts a lack of fusion for partial penetration groove welds, nor is the constrained removal of paint and slag a significant issue impacting verification of the welding evaluations.
ro validate the effectiveness of the program evaluation quality and address NRC concerns, a selective sampling and reverification program was establishoo alth the addition r'8 Appendix C. The sampling program was intended to confirm the in-place welding verificatio1 process reflected through the procedure by assessing its potential impact on some of the most heavily loaded members. This also confirms whether the process in any way compromises the accurate weld verification on some of the most important members.
The VWAC sampling program was used as a guide to select G4 welds for slag removal and further ovr'uation. For additional conservatism and, as discussed with NRC personnel, the sample seinaon was based on the most heavily loaded connections using analysis data from the piatform sampling portion of the program. While the specific purpose of the sample was to verify whether the assessment process was valid, the secondary purr.ose was to pmvide a high level of confidence
'a he program that the weld evaluation prx sss, for the most important welds, is nct ccmpromited by the presence of surface slag.
The validation process for the sample welds was as follows:
- 1) The 64 most highly loaded welds were identified and ranked by load magnitude from the sample drywell analysis computer output.
1 M The weld locations were identified. At Elevation 17'-101 24 were located; and at Elevation 38'-5" 40 were lonated. As a back up, the next 38 most highly loaded welds were included in the populat'on.
- 3) At Elevation 17'-10", a reinspectic7 of the sample weld:: identified that no slag was present. Since several welds had been brushed clean in preparation for other work, the original photographs were re-reviewed to identify the presence of slag. No slag was identified. The conclusions reached was t. no slag was evident on the 24 most heavily loaded welds at Elevation 17'-10".
- 4) At Elevation 38'-5", a reinspection of the 40 remaining welds indicated no slag was 4
present.
5! To provide additional evaluation data, the 38 next most highly loaded welds (the ones selected as back-ups) were reviewed using photographs. This adoitional assessment concluded thtt no slag was evident or the additional 38 welds.
The validation process concluded that the conservative sample taken of most heavily loaded welds, from the platform sampling prograrr Mcated that no slag is evide'it. While the VWAC sample program could not be confirmed due in the !!mited amount of slag, the process verified that the presence of limited slag is not a deterrent to the MSVP weld verification process and the determination of weld quality is not compromised. The most E1-13
I highly loaded 102 samples have no evidence of slag. A general overview of the balance of the packages at Elevation 17'-10" and 38'-5" showed that approximately 4% of the welds ha:) slag indications, but this amount is not considered significant.
Documentation of the selec !ve sampling and reverification program is to be included in the "MSVP Report For Restart.'
O.
Date of Full Comnliance The program contents and procedures in place are in compliance with the NRC requirements.
m El-14
ENCLOSURE 2 BRUNSWICK STEAM ELECTRIC PLANT, UNITS 1 AND 2 NRC DOCKET NOS. 50-325 & 50-324 OPERATING LICENSE NOS. DPR-71 & DPR-62 REPLY TO THE REQUEST FOR ADDITIONAL INFORMATION
- l. COLDSIDE AND HOTSIDE WAI.KDOWNS j
'l On April 21,1992 both units were taken off line in response to issues dealing with the seismic qualification of the Diesel Generator Building walls, it was determined at that time that a general walkdown of the Unit 2 areas, which were normally inaccessible during operation, would be conducted to assess the material condition (this was not intended to be a design verification walkdown). These series of walkdowns were referred to as the "Hotside" walkdowns, At this time, Unit 2 was scheduled for a 6/15/92 startup and Unit 1 for 5/25/92 and walkdowns were initially limited to Unit 2.
Guidance from existing procedure Al-96 "Drywell Inspections", input from tha Nuclear Engineering Department (NED) Lead Civil Engineer, and a Turkey Point procedure / check!!st were used by Technical Support management to develop a briefing p?ckage for the walkdowns. Personnel from Technical Suoport Training and NED were selected to function as inspection teams in the performance of the walkdowns. At the time of the initial Unit 2 walkdown briefing (4/27/92),
startup was anticipated as 7 weeks away and plant senior management set an expectation that theso walkdowns should capture about 80% of the existing material condition discrepancies.
During the Unit 2 walkdowns, the Unit 1 startup date was rescheduled to 9/1/92, and therefore "Hotside" walkdowns were initiated on Unit 1, with a briefing held on 5/8/92.
Prior to each walkdown, the areas were divided into " Zones" and personael from Technical Support training or NED were assigned. The inspectors were responsible for collecting the walkdown data and providing this data to a data coordinator designated by the Manager assigned to perform the walkdowns. Trouble tags were used to mark discrepancies found in the field. Where needed, pictures of the discrepancies were taken to supplement the data sheet describing the discrepancies.
The data sheets were forwarded to the Mainteaance planners for the development and planning uf '
Work RequestuJob Orders (WR/JOs).
The rmmber of discrepancies noted.during the "Hotside" walkdown led to a decision to perf:rm a "Coldside" walkdown In addition, Technical Support conducted a walkdown of the Diesel Generator Building.--The data from the "Coldside" walkdowns was gathered by the Zone Leaders and entered into various databases. The databases were merged for conveniencelt.to a single database.
E2-1
= _.
The current project manager tasked with the walkdowns has been required to track the discrepancies noted during the walkdc wns from " cradle to grave". This information has been uploaded to the database called the IBIR (Integrated Backlog item Report). The walkdown data is ranked by the System tingineer upon his review of the IBIR per the methodoiogy described in CP&L's July 23,1992 letter to Mr. Ebneter (NLS-92160). The ranking process is still ongoing. In parallel to this, WR/JO's and Site Memoranda have been initiated to address the walkdown discrepancies. Of the over 2400 items ider'tified duritto the walkdowns,19 were inadequ9te!y idsntified, such ?nat the discrepancies could not be relocated. Future walkdowns per "Special Plant Walkdown Procedure" (OSP 92-076) are planned and willinclude the areas containing these unidentified discrepancies.
T An NRC follow-up inspection was performed, at which time an area that had been included in the "Hotside" walkdown was revisited. Additional discrepancies were d scovered by the NRC inspector that were previously undocumented. The NRC further expressed concern that neither the drywell or torus had been included in the walkdown. Subsequently, additional walkdowns were performed to assess operability concerns found on one unit aga!nst the opposite unit and to capture the dryw61 and torus. Note however, that the drywell would have been walked dovan prior to startup per Al-96 and inspections per OPT-20.5.1 will cover the torus on its i igularly s'heduled basis.
RESPONSE TO FOUR SPECIFIC ITEMS:
u %Ilowing is a summary of the 4 specific NRC concerns addressed in IER 92-27 associated vuth te, u lkdown inspection program and the associat9d actions BNP has taken or plans to take to addion those concerns. The intent of these actions is to improve the overall quality.,f the hotside/ccidside inspection procees.
ITEM 1 Hotside/coldside ir sections were perfunned withou* Jeteiled written paocedural guidance.
RESPONSE
Although a formal procedure was not used in performance of the original walkdowns, guidance from existing procedure Al-96 "Drywell inspections," input form the NED Civil Lead Engineer, and a Turkey Point procedure / checklist were used by Technical Support management to develop a briefing package for the walkdowns. This method was used in lieu of a procedure, because at the time, star up was scheduled for approximately 7 week from the 4/21/92 shutdown. Tralniao in the form of a projob bnefing was conducted with the walkdown team members, prior to starting the walkdowns. These briefings were conducted on 4/27/92 (Unit 2) and 5/8/92 (Unit 1) by Technical Support management. N Plant General Manager, and the Manager of OM&M provided manageme.u eds and expectations uuring the Unit 2 brief;ng.
l Subsequent to the original series of walkdowns, procedure OSP 92-076 was approved on 10/19/92 for the performance of additional plant walkdowns as deemed necess9ry..
This procedure incorporates the recommendations of the INPO Good Practice for Pirat l
Inspection Prograrns. Additional walkdowns are being scheduled which will be performed per the requirements of this procedure.
E2-2
?
4 ITEM 2 Failure to perform walkdowns of the drywell, torus, ar.d portions c'.he control room.
Other safety-related areas may not have been inspected, but the insp 'etors could not determine which onc6, if any, without furthar review.
RESPONSE
The drywell was not included in the original scope of the hotside walkdowns as a o
general walkdown of the drywellis required prior to startup per procedure Al 96.
Additionally, drywell structural concerns were omitted from the original scope of this effort since a civil / structural walkdown, performed by Bechtel, we planned for the drywell. It was decided that these inspections would identify the major deficiencies and; the radiological dose that would be incurred to identify the remaining minor deficiencies was nct wnrranted.
1-Additional walkdowns aru being scheduled which. Nil' include the orer, identified above as well as repeating portions of the original walkdown scope as appropriate. The expansion of the hotside/coldside walkdowns b to include the torus and control *oom.
ITEM 3 Use of inspection personnel v!ith varying quOifications and experience levels to perform the walkdown.s. The level of experience and expertise varied significantly from one inspection team to ano;her, which affected the results of the walkdowns.
RESPONSE
Personnel used for the hotside/coldside walkdowns were selected based their knowledge of the plant and experience. Althot,gh personnel came from various plant organizations, previous plant field experience was a prerequisite.
Procedure OSP-92-076, now requ.tes thn walkdowns be performed by a multi-disciplinary team of inspectors. The wna inspection teams are to be cc,mprised uf, as a -
minimum, three member 1 l' cl6ctrical/l&C,1 Mechanical,1 Civil /Structurai).
Additionally, this procedure requires that these personnel have the appropriate skills and knowledge in their discipline (Mechanics, I&C Technicians, Engineering Technicians,.
Engineers, Quality Control, Maintenance Planners).
l l
E2-3 1
ITEM 4 The extent of the inspection effort and the attention to detail veried significantly from one inspection team to another. Interviews with several walkdow') team members disclosed significant differences in the bme spent performing the walkdowns and the inspection methodclogy they emp'cyed during thc, walkdowns.
RESPONSE
As identiPed above, multi-disece.ary inspection teams will be used to perform additior.at plant walkdowns. Procedure OSP-92-076 requires that a pre-job brMing be conducted prior to performing any inspections and that the length of time spent inspecting each zone take as long as the zone team menibers feel is necessary to ensure a complete and thorough walkdown. Additionally, these plant walkdowns are intended to be performed with a limited number of qualified personnel. The intent of the preceding is to ensure that:
- 1) Each zone be inspected for deficiencies by a team of personnel Qualified in various plant disciplines, c
- 2) All personnel invodd in the walkdowns are aware of their duties and responsibilities.
(
- 3) Thore is consistency in the inspection methodology,
- 4) Adequate time is Opent identifying defit.iencies in a given area.
W E2-4
~
II, APPROPRIATE DOCUMENTED lhSlBUCTIGN3 OR PROCEDURES The NRC's ' spection Report 92-27 ndicated that a weakness was identified regarding the use of a draft procedure to perform calculat;ons for evaluation of irret"larities The following provides the resuhs of the investigation conducteo bi, CP&! which included interviews with Bechtel personnel, to address the NRC's concein in this regard.
Bect.tel Engineering Department Procedure (EDP) 4.37 r ovidos the OA requirements for the preparation, checking review, approval, and control of design calculations. Bechtel Project Engineering Procedures Manual (PEPM). which contained this approved EDP, was initially issued oa July 2,1992. The PEPM provio J tho procedural basis for performing engineering activities. It was recognized by CP&L and Bechtel that certain Engineering Department Project Instructions (EDPIs) would be developed to address CP&L plant specific requiremems. As a part of the first group of plant specific EDPIs, EDPI 4.37-01 was issucd witn Revision 1 of the PEPM on August 24,1992.
This EDPIincluded the plant unique formats and a requirement for review and approval of
)
calculations by CP&L. At the time of the NRC inspection, Revision 6 of this EDPI was in effect and was being imp!amented for preparation of design calculation.
The calculations reviewed by thc NRC during their Site visit of September 14-18, 1992 listed EDPI 4.90-02 as a reference. At that time EDPl 4.90-02 had been internally reviewed by Bechtel and was pencing finat approval by CP&L prior to its issuance. The pnmary purpose of preparing EDPl 4.90 02 was to provide an overall description of the Misce':aneous Steel Verificatico Program (VSVP) and to address i.s major functions and corresponding ocverning p ocedures. During the development of this EDPI, Bechtel in coordination with CP&l, decided to include in this EDPl a matiix of Updated Final Safety Assessment Report (UFSAR) criteria and clarifications as it applied to the miscellaneous steei. The purpose was to provide uniformity in the use of applicable criteria from the UFSAR, ;ocluding its clarificaticns and other assumptions. This EDPI, was pending approval by CP&L and was issued for use with Revision 3 of the PEPM on Saotember 17,1992.
EDPI 4.90 02 does not provide OA requirements for prepar tion, review and control of calculations.
Rathe it provider, an overview of the MSVP and provides criteria available in the UFSAR and clarification and assumptions which could have been otherwise individually included in the calculations. As indicated +o the NRC at the tims of inspection, all cak:ulations were in-process. It had been recognized by CP&L and Bechtel that some work had to be performed in parallel but, this was consistent with this type of EDPl. As indicated to the NRC at the time of the, inspection, the final reviews and approvals of calculations were pending i*suance o' 1e Erspi, !ri order that the calculat;ons could be revi.wt,d to insure consistency and conformarn e. Bechtel has assured us that no work activitits have been or will be performed and issued outside the approved QA Program and procedures. It is CP&L's policy not to allow safety related work tr
- a pe formed and issued outside of the approved OA Program and procedures. CP&L has the fine;,
ul of GA program, procedurcs, and work product (calculations, modification sketches, design criteria, etc.).
I 5
E2-5 I
4
_