ML20054E997
| ML20054E997 | |
| Person / Time | |
|---|---|
| Site: | Beaver Valley |
| Issue date: | 05/26/1982 |
| From: | Woolever E DUQUESNE LIGHT CO. |
| To: | Starostecki R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| Shared Package | |
| ML20054E994 | List: |
| References | |
| NUDOCS 8206150115 | |
| Download: ML20054E997 (9) | |
Text
.- *.
= en
'A P (412) 787 - 5141 e
(412)923 - 1960 Telecopy (412) 787-2629 Nuclear Construction Divis6on Robinson Plaza, Building 2. Suite 210 MAY 26,!982 Pittsburgh, PA 15205 United States Nuclear Regulatory Commission Region I 631 Park Avenue King of Prussia, PA 19406 ATTENTION:
Mr. "ichard W. Starostecki Division of Project and Resident Programs
SUBJECT:
Beaver Valley Power Station Unit No. 2 Docket No. 50-41?
USNRC 1E Inspection Report No. 50-412/82-02 Gentlemen:
)
1 This is in response to the items of infraction cited in Inspection Report No. 50-412/82-02 and listed in Appendix A (Notice of Violation) attached to your letter to Mr. E. J. Woolever dated April 21, 1982.
NRC VIOLATION (82-02-01) 10CFR50, Appendix B, Criterion V and ANSI N45.2(6) require that quantitative criteria, such as dimension and tolerances, shall be specified, as appropriate, for determining satisfactory work performance and quality compliance.
Contrary to the above, as of April 1, 1982 drawing No.
12241-BZ-STD-PS-2A-2 for installation of pipe anchor PSA-027 shown on draw-ing No. 103310'-0E failed to include sufficient dimensions to assure adequate weld thickness.
This is a severity Level IV violation (Supplement II)".
Further the report stated:
"The inspector observed that the Stone & Webster Engineering Corporation standard procedure STD-PS-2A-2 specified the weld configurations and sizes. The weld dimensions as described in the standard are such that a weld could be made meeting all the require-ment s without having an adequate throat or weld metal deposit foi structural integrity. The inspector further observed that once a weld was made the effective throat or weld metal deposit could not easily be determined. The failure of the standard to specify quantitative criteria to preclude inadequate weld size is a violation of 10CFR50, Appendix B, Criterion V.
(S2-02-01)".
8206150115 820610
{DRADOCK 05000412 PDR i
United States Nuclear Regulatory Commission page 2
RESPONSE
Standard BZ Drawing No. PS-2A is used to describe the weld joint details for connecting trunnions to run pipe. The original issue, PS-2A-1 dated January 9, 1979, contained both an angular range requirement of 60 to 100 degrees between the pipe and trunnion surfaces to be welded and a weld "the size of the weld shall equal or exceed leg length requirement that, Typical cross-the specified size, w, at all points in the connection.
sectional views were provided showing the weld joint angle, chamfer width, weld leg length, and weld surface profile.
Infractior. 82-02-01 cites BZ Drawing No. PS-2A-2, dated January 11, 1980.
In response to 82-02-01, Drawing No. PS-2A-2 will be voided, work per-formed to its requirements reviewed, and new instructions provided for The potential consequenses of work performed future fabrications.
to PS-2A-2 include the use of joint angles in excess of 100 deg. and under-This latter sized weld leg lengths on the run pipe side of the joint.
concern is based on a misunderstanding that this side of the weld is However, the use of a flush weld controlled by achieving a flush weld.
criterion is only appropriate when the groove angle is nominally equal tc To provide clarification for future work drawing Note 2 will be A 2 100 90 deg.
revised to limit the maximum angle of chamfer to 100 deg.(i.e. 60 g and a reference to Note 3, describing the weld leg length, will be shown on both sides of the joint in each typical cross-section provided in The potential use of excessive joint angles and inadequate weld leg lengths will be prevented in the future by these changes to the PS-2A PS-2A.
standard.
NRC VIOLATION (82-02-02) require 10CFR50, Appendix B, Criterion X and ANSI N45.2 and N45.2.5 (5.5) fit-up prior to start of welding on struc-inprocess inspection of jointN45.2 further requires that where a sample is used to verify tural steel.
acceptability of a group of items, the sampling procedure shall be based on recognized standard practices and shall provide adequate justification for the sample size and selection process.
Contrary to the above, as of April 1, 1982 inspection of joint fit-ups was not being performed in accordance with a recognized sampling procedure.
This is a Severity Level IV Violation (Supplement II).
Further the report stated:
"The inspector expressed concern that the inspection of welding preparation and fit-up were not performed on all safety-related welds. In response the inspections for to the inspector's concern the licensec stated that fit-ups and welding-prep were carried out on a random sampling basis, and was considered adequate by the licensee to assure the integrity of welds.
The inspector, however, observed that the random inspections performed for weld-preps and fit-ups were not based on any established sampling plan, recognized standard practice, and/or a predetermined frequency; also, the licensee could not provide adequate justification for the sample size or selection process.
.m
United States Nuclear-Regulatory' Commission
~
,Page 3 q
The inspector further' observed that the weld #2-SIS-87-F500,in the safety -
Injection' System was' inspected but had?not been'accep'ted because the final weld size was not verifiable by QC due to-the -lack'-of weld-prep. b and fit-up data, and the dimension of land area in the.jointi_
The method used bylthe licensee to' perform the fit-up. inspections does not meet the intent of 10CFR50, Appendix B, C;iterion X^for the following reasons:
~
~
a.
The itema selected.for fit-up inspection are= highlighted with an asterisk on the weld data sheet. This provides the construction, personnel advance notice of those that will~be inspected and those that will not.
This could result in 'such welds being. a non-representative sample. Added attention to quality would bespossible for those welds so identified for inspection including assignment of the best fitter (s) to work on such welds.
b.
The method used by Quality Control personnel to select the fit-ups that will be inspected is not based on any approved sampling program.
It appears to be based on the decision of that particular person marking the weld data sheet. This method could vary from person to person and does not provide adequate justification for the different variables involved such as different crafts and welding.
processes involved.
The failure to perform fit-up inspections by recognized standard practices is a violation of 10CFR50, Appendix B,. Criterion X.
(82-02-02). "
RESPONSE
Because of the extent of the subject affected by this. item of Infraction 82-02-02, Beaver Valley Power Station #2 (BVPS #2) requests that this response be considered as an initial response'. Further interim or a final response will be transmitted thirty days from the date of this letter.
UNEDIATE ACTION A stop work order was issued against the welding of trunnions to pipe by memorandum #2BVSE-60355-CLB/N, dated April 8,1982. This action was taken for two basic reasons:
the insufficient criteria reported by Infraction
- 82-02-01 and the trunnion weld subject for inspection has not been selected for fit-up inspection.
SHORT TERM ACTION It is considered necessary to place on record the degree of fit-up verification on the typec of weld configurations installed, performed in addition by Quality Control (QC) and Welding Supervision-Staff (WSS) to the verification performed by the Craft Foreman and Craft.
In some instances QC and WSS both inspect the same fit up.
In other instances, i
either one or the other verify the attribute.
T.
s t '
7 a
.~
)
N 'i
% f y
0
United States Nuclect R gulctory Commission
?, %
Page 4 Q~'r..
i s
\\
,.g g,
a As the ASME Section III Pipe Support Attachment Welds (ll Unnion Type)
/h I were the subject of the original concern, they are deschibed in more I
detail as follows:
I,l
-y 3
s,,
(a. b A review indicates that 84 supports were installed. Some'oftl$ti supports have double trunnion welds resulting-in a tota), of 10J Weld Data Sheets (WDS). The fit-up verification assignments were as k 7 follows:
h QC 42 40%
98% Combined
's\\
WSS 97 93%
1 a,
(
\\ l i
Craft only 2
2%
s s
s q
\\15 COMMENTS r
5 BVPS #2 considers the number of fit-up inspection assignments to jo mo\\ \\
re N
than adequate. The problem is related to the need for more definitive criteria (see 82-02-01) and instructions to QC, WSS and Craft person.hl, for this type of weld configuration. These items wilf$A resolved prioi to recommencement of the trunnion to pipe welding activity.
g
/
s In order to establish the actual weld conditions an UltrasoMc t.ximination l'
program (Procedure DLC/SQC UT-13) has been ic.itiated,
>,- } ' ',
The results of the examination and engineering analysis W of M,iy,>,&
E7, 1982, are as follows:
\\
).
Welds Examined 57
(
i '
,=
,., /
Welds Analyzed 29 l
g
.),
'3 Welds Accepted to Original Engineering Criteria 22 ij
%.g
(
Welds Subject to Further Evaluation 7
/
g The examination is continuing and the results of the examinat. ion a'rd analysis will be made available as required.
3 1
N ASME SECTION III PIPE BUTT WELDS (Clast. 1, 2 & 3 )
1 q
i Class 1:
Requires 100% QC fit-up verification and 100% Radiography' sJ
,l Test (RT). Therefore, no review was performed.
r i
Class 2:
Requires random fit-up verification by QC and 100% RT.
'Ihe i
UDS of the 675 welds shown on the RT Weld Analysis Report 3/30/82 were reviewed.
Assigned QC 671 -
99%
o9% Combined Assigned WSS 671 -
99%
Craft Only 4
1%
i
3
-l United States'Nucicar Regulatory Commission Page.5 Inspected QC 675 100%
Inspected WSS 675 Craft Only 0
0 l
The four welds not assigned for fit-up verification were actually inspected and so recorded.
Class 3:
Requires random fit-up verification by QC with no R/T. The method of. selection of the WDS for review was as follows:
61 Isometric Drawings (ISO) were selected from the total E 198 in the drawing log, covering the period July 1976 through' February 1982. The ISO's selected contained 4
450 welds, the 450 WDS obtained had fit-up assignments maic by 10 out of the 12 personnel employed for this function by' C and.USS during the period covered.
Assigned-QC 428 95%
100% Combined
.Assighed WSS 444 98%
1 i
Craft only 0
j COMMENTS 1
j A11'1125 pipe welds reviewed had the fit-up attribute verified by QC/WSS. The 4 welds that were not originally assigned this attribute were ' required to be examined by RT.
I ASME SECTI,0N III PIPE SUPPORT ATTACHMENT WELDS (Excluding Trunnion Type)
Class 1:
None of these welds have been fit-up or completed.
l Class-2:
A similar selection process, by ISO and drawing log, as that l
used by Class 3 pipe weld selection,was used. This resulted in 67 (32.8%) IS0's being selected from the total of 204. The I
IS0's selected contained 90 welds. The results obtained for fit-up assignment are as Tollows:
QC 6
6.57.
87% Combine'd WSS' 74 82.0%
Craft only 12 13.0%
12 welds of the 190 WDS reviewed, completed.
Class 3:
The same IS0's used for Class 3 pipe weld selection were used.
The 61 ISO's contained 106 welds. The results obtained for
- fit-up are as follows:
4
\\
s
7 i
t
, United Stetes Nuclect Ragulatory Commiscica Page 6 QC 1
1%
96% Combined WSS 102 96%
Craft only 4
4%
16 welds, of the 106 WDS reviewed, completed.
COMMENTS This type of weld configuration.can be considered as simplistic and in some instances the fit-up gap can be verified visually to some degree at the completed weld stage.
SUMMARY
OF SHORT TERM ACTION I
The foregoing illustrates the degree of QC/WSS verification applied to the fit-up attribute under ASME III for welds that have been considered to have a structural application and require Weld Data Sheets and Hold Point Assignment. The WDS review has indicated the following:
a)
In certain pipe classes, the QC verification could be reduced, particularly where RT examination is involved; b) QC has over-emphasized the " simplicity" aspect of the Pipe Attachment Welds (not Trunnion Type);
c) The degree of fit-up verification assigned to QC (80.8%) and the verification assigned to QC and WSS (98.7%) combined nullifies:
- 1) Any adverse effect that may be caused by the lack of "any recognized sampling procedure",
- 2) The contention that an " unrepresentative sample" was obtained by giving " advance notice";
- 3) The contention that.a variance occurred in the assignment of this caused by the " method varying from person to person";
d)
The summary relating to the Trunnion to Pipe type weld is given in the ASME Section III Pipe Support Attachment Welds (Trunnion Type) discussion on page 4 of this report.
SHORT TERM CORRECTIVE ACTION a) The QC (fit-up attribute) assignment for the Pipe Support Attachment Welds (Non-Trunnion) Class 1, 2 and 3 will be increased i
with particular emphasis on those configurations not allowing any fit-up verification at completion; I
,s Unitzd Stctoo Nuclect R:guictory Connaiscicn
- Pigs 7
+
b) On the receipt of revised criteria and appropriate' instructions, the QC fit-up attribute assignment on Trunnion / Pipe Welds will be increased to 100% on the basis.on a new activity with some complexity until credibility is obtained; c) The actions in a) and b) may cause a decrease in QC verification of fit-up for ASME III Class 2' Pipe Welds.
LONG TERM CORRECTIVE ACTION A more defiStive frequency plan for QC attribute assignment will be formalized, possibly based on:
a) The simplicity or complexity of the configuration; b) The degree of visibility of the fit-up condition at completion; c) The application of volumetric Non-Destructive Examination (NDE) at final condition; d) The relationship between QC assignment and the assignment of fit-up inspections attribute by WSS; l
l e) The feedback of unsatisfactury conditions.
This approach is subject to change. The program, when confirmed, will be submitted in a future response.
SAFETY IMPLICATION BVPS #2 considers there is no safety implication directly related to the degree of weld fit-up verification assignment. However, the results of the Trunnion / Pipe attachment weld UT examination may indicate l
such an implication when evaluation is complete. If.this is the situation, the matter will be reported in accordance with the requirements of 10CFR50.55 (e).
l GENERAL BVPS #2 is proceeding with reviews of the fit-up verification of welds related to Pipe Supports, Elcetrical Cable Tray and Conduit Supports, i.
HVAC Supports and AWS welding. Unless otherwise instructed, the l
results of the reviews and programs for these activities will be submitted to the Resident-Inspector when completed.
Sincerely yours, e
E.
. Woolever, Vice President Nuclear Construction Division i
. _. ~., _. _ _.. -.
United States Nuclear Regulatory Comission Page 8 SDil/ge cc: Messrs:
Mr. G. Walton, By-2 NRC Resident Inspector Ms. E. Doolittle, NRC Project Manager l
l 1
l
E kJnitedStatesNuclearRegulatoryCommission
. Hay 25, 1982 Page 9' C0:e!ONWEALTH OF PENNSYLVANIA COUNTY OF ALLEGHENY On this 26th day of May,1982, before me, A LA A>
/8. 8AuAI,a Notary Public in and for said Commonwealth and County personally appeared E.
J. Woolever who, being duly sworn, disposed, and said that (1) he is Vice
~
President of Duquesne Light Company, (2) he is duly authorized to execute and file the foregoing Report on behalf af said Company, and (3) the statements set forth in this Report are true and correct to the best of his knowledge, information, and belief.
WITNESS my hand and seal the day and year first above written.
se< (Z-*AV ALAN B. EA1 AS. NOIARY IU3ll0 h
ECBIN00N TW.. AL;WIW CSUNTY c1 CC'G155 c;g m:p;$ pt IL 1916 behe.Pematvan a Anotianca cf Notvies i