TXX-6102, Forwards Comanche Peak Response Team (Cprt) Responses to Case 850814 Comments on Welding,As Requested in NRC 850930 Evaluation of Cprt Program Plan.Emphasizes That Reinsp Efforts by Third Party Are Not Insps of Record
| ML20214T905 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 12/02/1986 |
| From: | Counsil W, Keeley G TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC) |
| To: | Noonan V NRC - COMANCHE PEAK PROJECT (TECHNICAL REVIEW TEAM) |
| References | |
| TXX-6102, NUDOCS 8612090050 | |
| Download: ML20214T905 (11) | |
Text
__
t Log # TXX-6102 File # 10068 clo 10010 TEXAS UTILITIES GENERATING COMPANY SKYWAY TOWER. des NORTH OLIVE STREET. E.B. 88. DALLAS. TEXAS 753e1 December 2, 1986 EYl^".hff.UE Director of Nuclear Reactor Regulation Attn: Mr. Vincent S. Noonan, Director Comanche Peak Project Division of Licensing V. S. Nuclear Regulatory Commission Washington, D. C. 20555
Subject:
COMANCHE PEAK STEAM ELECTRIC STATION (CPSES)
DOCKET NOS. 50-445 AND 50-446 RESPONSE TO COMMENTS ON WELDING CONTAINED IN " CASE's FIRST CRITIQUE OF APPLICANTS' COMANCHE PEAK RESPONSE I
TEAM (CPRT) PLAN" DATED AUGUST 14, 1985
Dear Mr. Noonan:
Enclosed herewith are the CPRT responses to the CASE comments on welding contained in their August 14, 1985, letter. These responses were requested in the NRC Staff Evaluation of the CPRT Program Plan dated September 30, 1985, paragraph 3.3 " Quality of Construction and QA/QC Adequacy Program Plan" (Appendix B), Item 3 which reads as follows:
" CASE's first critique of applicants' Comanche Peak Response Team (CPRT)
Plan submitted to the Atomic Safety and Licensing Board on August 14, 1985, contains numerous concerns on the appropriateness of welding and welded construction in pipe supports. The applicant should address the CASE comments on welding and justify those items which the applicant I
concludes are not applicable."
In general response to CASE's comments on welding, we wish to emphasize that the reinspection efforts by Third Party associated with the CPRT program are I
not inspections of record.
These inspections are con?ucted subsequent to the final acceptance inspection.
Reinspections subsequent to the acceptance inspection are intended to evaluate and verify that previous inspections were l
adequate and that the design criteria were not compromised. They are not intended to upgrade or downgrade the level of workmanship. All welds which are included in the CPRT program of reinspection have previously been l
completed, inspected, repaired or reworked if necessary, and then again reinspected and accepted. The CPRT visual reinspection program utilizes a statistically based sampling plan to determine the group of welds to be reinspected from the population of completed welds.
8612090050 861202 DR ADOCK 0500 5
h 41,,nsrov or rnw im,mes tr rcruic an,nw
TXX-6102 December 2, 1986 Page 2 For the reinspections, the CPRT uses inspection checklists which satisfy AWS D1.1 criteria for stru.:tural welding or ASME III NF for ASME component support welds. The Visual Weld Acceptance Criteria (VWAC) is used as the acceptance criteria for AWS welds. Tin VWAC document was developed by the Nuclear Construction Issues Group (NCIG) for non-ASME structural welding. This document, formally NCIG-01, Rev. 2, dated May 7,1985, was carefully researched by a consortium which included prominent Architect-Engineering firms and utilities and, after detailed review was accepted by the NRC on June 26, 1985.
Very truly yours, k'.C3unsil W. G By:
G. S. Keeley
- Manager,Nuclearhhensing JDS/mlh Attachment cc: CPRT File
Attachment to TXX-6102 December 2, 1986 Page 1 of 9 RESPONSE TO CASE ALLEGATIONS MKEILN Applicant plans "... to rely upon visual inspection only as versus the use of liquid penetrant, magnetic particle radiography, destructive testing, etc."
(Paragraph E, Page 8)
RESPONSE
l For the original and final inspection of record, the Applicant inspected in accordance with the appropriate code requirements. Welds installed in accordance with AWS Dl.1 do not require non-destructive tests for their inspection. A visual weld inspection is the only Code requirement, and that was accomplished under the applicant's original inspection program.
For ASME-NF component support welds, Non-Destructive Examination (NDE) inspection is only required on specified class I linear, plate and shell support welds.
Class 2, Class MC and Class 3 support welds are required to receive a visual examination, and that was accomplished in the applicant's final acceptance inspection.
During the subsequent visual reinspection under the CPRT program, all reinspections will also be or were in accordance with the applicable code criteria.
For the welds where an NDE was specified as a hold point on the weld data cards, a check is being or was made to assure that the test was performed and that the inspector who signed for the NDE was certified at the time of inspection.
In doing this, assurance is gained that the proper NDE was performed by a qualified inspector.
If this check cannot verify that the appropriate NDE was performed by a certified inspector, it will be reported and resolved in accordance with CPRT procedures and the TUGC0 QA program.
CONCERN l
"The original inspection plan of the Applicants also relied exclusively (in most cases) on visual examination. However, this was a compromise in that they used an inspection process which was more economical than the most preferred methods.
For example, welds were not x-rayed or tested to destruction."
(Paragraph E, Page 8)
RESPONSE
There was no compromise in selection of the inspection processes by the applicant.
The original inspection plan relied on Code and regulatory requirements, supplemented with accepted industry standards. As a minimum, the inspection plan met these requirements. There is no requirement or f
justification for X-ray or destructive testing of the production welds involved herein. When the Code deems that a weld, due to its criticality or configuration requires a volumetric inspection, that inspection was performed under the original inspection program. The CPRT program checks the documentation and inspector certifications for such a volumetric inspection.
1 a
Attachment to TXX-6102 December 2, 1986 Page 2 of 9-There is no one universally preferred method of inspection for locating all types of defects in all types of welds.
If one simply chooses the method by which the greatest degree of inspection could be achieved, then obviously a volumetric type of inspection would be employed. However, the degree of inspection needed or required is based on such factors as service conditions, joint criticality, types of materials involved and types of joints.
Volumetric inspection of fillet welds, the majority of welds involved herein, would be neither practical nor useful in the CPRT reinspection effort.
Except -
in very few cases, AWS D1.1 only requires visual examination. The CASE report-cites ASME III NF. requirements (see page 11-CASE) and it should be noted that the vast majority of ASME III NF welds only require a visual examination, not liquid penetrant (PT) or magnetic particle (MT) examination.
It should also be noted that later editions of NF dropped the radiography (RT) requirements for Class I linear type support welds in recognition that RT inspection of fillet welds is a futile exercise.
4 CONCERN "The proposed reinspection program relies upon a depleted list of attributes to provide all information necessary to made-a safety decision."
(Paragraph F,'Page 8)
RESPONSE
Inspection of AWS DI.1 welds will be carried out in accordance with NCIG-01,
" Visual Weld Acceptance Criteria for Structural Welding at Nuclear Power 3
Plants," 05/07/85, Rev. 2.
This document was approved by the NRC on June 26, 1985, and also approved by Gibbs & Hill for use at Comanche Peak. This document is the most current evaluation of weld inspection criteria and the l_
engineering basis which is the foundation for such acceptance criteria.
Reinspection of welds installed in accordance with ASME-NF will be or was i
accomplished using the appropriate ASME visual weld acceptance criteria. No relevant inspection attribute is omitted from the reinspection checklist.
CONCERN-l Other attributes not addressed by Applicant in their letter of 8/1/85 - list i
of attributes.
(Paragraph F, Page 10) i
RESPONSE
l The intent of the CPRT reinspection program is to address all the safety significant attributes of the governing code for final weld inspection.
Therefore, the following attributes from CASE's list are being considered either by a visual reinspection or a documentation review:
" incomplete penetration", " corrosion", " fit-up defects", " wall thickness", " material as per drawing", " welding as per symbol on drawing", " weld cleanliness",
~
" compliance with bill of materials", " welder symbol", " heat number", "WFML i
(Weld Filler Material Log) number", " Weld Procedure Specification (WPS)",
"related NCR number."
i
-we w--.-,-
y-,-,-%,-
,,w,..-,,,,------m.-,,,,,-+,,,-m
,,,,%,..-w--,nm,v,_,,,,-,r.-,,,,,,, -, -
i..,,i.,,,,-,,-,-,.,,,,,v.,-
Attachment to TXX-6102 December 2, 1986 Page 3 of 9 The following attributes are not being considered in relation to reinspection of welds since they are not directly related to welding.
They will be utilized as applicable during reinspection of items other than specific reinspection of welds:
" lack of bond", " support level and plumb", " torque previously verified", " dimensional evaluation", " CMC, including revision numbers", "IR closed", " support drawing number and revision number".
CONCERN
" Applicants are evidently planning to ignore cracking as a potential problem at Comanche Peak."
(Paragraph F, Page 11)
RESPONSE
On the contrary, cracking and lack of fusion have been specifically addressed in the front-end evaluation which was accomplished prior to finalization of the reinspection program.
. An analysis of the history of weld-related problems at Comanche Peak, for the purpose of determining whether cracking or lack of fusion had been a problem, was performed. Any instances of cracking and lack of fusion were required to be reported on Non-Conformance Reports (NCRs). A statistically based sampling plan was used to evaluate applicable NCR files for evidence of these deficiencies for welded supports. The results indicated that neither cracking nor lack of fusion was a significant problem.
Furthermore, inspections by the NRC indicated that there was no evidence of these deficiencies being a problem at Comanche Peak.
Seventy-seven percent (77%) of the pipe supports inspected by the NRC Technical Review Team (TRT) were unpainted. Neither cracking nor lack of fusion was reported. The TRT inspected fifty-nine (59) electrical raceway supports. Again neither cracking nor lack of fusion was reported.
As part of the formulated reinspection program an evaluation using 60 randomly selected welded connections is being performed to obtain supplementary information regarding inspection of non-pressure boundary welds through paint.
This evaluation is taking place in two steps.
Initially the welded connection is inspected with the paint on the weld. The paint is then carefully removed and the weld is reinspected.
The results of both inspections will be compared, and any discrepancies between the two inspections will be evaluated.
The results of these inspections will provide additional information regarding the feasibility and limitations of inspecting welds through paint.
In addition, metallurgical analysis supports the fact that the involved materials are readily weldable and therefore quite " forgiving" in terms of tolerating a wide range of parametric variables. The single most important consideration for these materials, excluding unusual service conditions, i.e.,
corrosive environment, fatigue, etc., is overly restrained welded members.
Cracking associated with this situation results in cracks of a magnitude that would make paint masking an extremely unlikely event.
i Attachment to TXX-6102 December 2, 1986 Page 4 of 9 CONCERN "Also from AWS (Page 326), one can conclude that cracking can take place such that it would not be observable by visual examination alone (with or without paint)."
(Paragraph F, Page 12)
RESPONSE
As has been previously stated, both the AWS DI.1 and ASME-NF codes stipulate that a visual inspection of the completed weld is the final acceptance inspection required on the vast majority of structural and component support welds. This requirement was based on several factors: weld service.
conditions, joint criticality, types of materials involved, types of joints, etc. Additionally, properly trained welders, using controlled materials and -
working to approved welding procedures will produce sound welds which are acceptable to the design conditions. The reinspection program will verify that' qualified welders did the welding, that the correct materials were used and that approved welding procedures were utilized.
CONCERN "You only see the indication of the surface conditions' with visual inspection (TR.12159/5-8)."
(Paragraph F, Page 13)
RESPONSE
This observation is correct. However, as previously stated the codes applicable for the welds in question for the most part only require a final visual inspection for acceptance of the weld.
Consequently, CASE's observation appears to be a criticism of the codes in question and not the CPRT reinspection program.
It is also true that both MT and PT non-destructive examinations are surface inspection techniques. PT is limited to the detection of discontinuities open to the surface. Certain MT techniques are capable of detecting surface and near-surface discontinuities. However, MT is not considered a volumetric examination method, since it is not a reliable method for the detection of subsurface discontinuities.
CONCERN "In Applicants' 8/1/85 letter to the NRC staff, they state that undercut is actually emphasized by the existence of paint..."
(Paragraph F, Page 13)
Attachment to TXX-6102 December 2, 1986 Page 5 of 9
RESPONSE
Experience by Stone and Webster Engineering Corporation on other projects has revealed this to be true. Due to a shadow effect of the paint / epoxy, especially with an enamel (glossy) type paint, undercut appears to be larger than is the actual case.
Stone and Webster's field experience has shown that inspection of painted welds revealed numerous examples of undercut.
- However, when the paint was removed the undercut was either not present at all, or was present to a lesser degree than originally observed. Consequently, CASE's concern is not substantiated with field observations of the phenomena cited.
CONCERN "Then they state (without offering any justification) that fatigue loading is not a factor for these welds."
(Paragraph F, Page 13)
RESPONSE
The statement made was based upon a design engineer's functional knowledge and application of FSAR requirements, applicable codes and standards, the materials used in the supports along with the material properties and the support loadings as applied in support design.
Based on this knowledge and first hand experience with support design, fatigue is not a factor for these welds.
CONCERN
" CASE recognizes that the current QA/QC team has not addressed this matter of visual acuity."
(Paragraph F, Page 14)
RESPONSE
CASE's assertion is not correct.
" Visual acuity" is a recognized factor and is addressed in the CPRT Program through the qualification of inspectors used in the CPRT Program.
The reinspections are conducted by experienced personnel who are required to have an annual vision examination to verify that their natural or corrected near distance acuity is such that they are capable of j
reading a minimum of Jaeger No.1 on a standard Jaeger Test Chart or equivalent.
In addition, these personnel are familiar with the lighting requirements addressed by ASME and NCIG-01 and are provided with tools such as mirrors and flashlights to ensure visual reinspections can be adequately conducted.
Attachment to TXX-6102 December 2, 1986 Page 6 of 9 CONCERN "It is obvious that any organization capable of developing a procedure so patently inadequate and deliberately designed to hamper good QC inspection cannot be relied upon to develop the criteria and procedures to be utilized in the upcoming reinspection effort."
(Paragraph F, Page 15)
RESPONSE
As previously stated, the CASE concerns were addressed by the CPRT prior to receipt of these concerns.
In addition, the QA/QC reinspection procedures were developed by experienced, degreed engineers based on their experience in applying codes and standards requirements to the design and construction of systems and components. Also, the procedures were reviewed by personnel experienced in conducting QA/QC inspections and related activities.
Consequently, this judgement of CPRT adequacy by CASE is not valid.
CONCERN "Many of the listed attributes should have been totally cleaned up according to the procedures in the original QA/QC plan making them no longer observable."
(Paragraph F, Page 16)
RESPONSE
The analysis of the history of weld-related problems at Comanche Peak referred to earlier indicates that, of all the attributes, the only ones which exhibited significant deficiencies were weld size and length and these are readily observable through paint. None of the evidence from either the Applicants' or the NRC investigation has shown any of the other listed attributes to be a problem. This will be verified by the reinspection program.
CONCERN
"... the Applicants appear to downgrade these attributes when considered with respect to the codes and standards."
(Paragraph F, Page 16)
RESPONSE
Attributes have not been downgraded when considered with respect to the codes and standards. ASME-NF welds will be inspected in accordance with ASME Code requirements. AWS welds will be inspected in accordance with AWS Dl.1 as implemented by the NRC-approved NCIG-01 VWAC criteria, an industry developed and approved standard.
1
Attachment to TXX-6102 December 2, 1986 Page'7 of 9 CONCERN
" Spatter... is q tite likely indicative of improper welding technique and if the likelihood or other associated faults."
(Paragraph F, Page 16).
"The surfaces to be, inted shall be cleaned of spatter..."
(Paragraph F, Page 17)
RESPONSE
Some spatter is a normal condition associated with most arc welding processes 4
and according to NCIG-01, spatter remaining after the cleaning operation is acceptable. The basic purpose for removing weld spatter is to clean the area so it can be properly inspected and painted. Tightly adhering weld spatter is not harmful to the weld or base metal.
j CONCERN
" Arc strikes... can initiate failure in bending or cyclic loading. They can create a hard and brittle condition in alloy steels, and are inadvisable even on mild steel when... normal fatigue stresses may be encountered."
(Paragraph F, Page 17) 4
RESPONSE
The majority of welds which are inspected through paint are not alloy steels nor are they subject to fatigue loading. The alloy steel welds which are to be reinspected are piping welds. The areas adjacent to piping welds are inspected for arc strikes.
CONCERN
... Slag shall be cleaned from all welds."
{
RESPONSE
I ASME-NF does not address slag in its visual inspection criteria. According to NCIG-01, it is not considered a necessity to remove tightly adherent areas of slag since it has no effect on the weld. The amount of slag permitted will not mask any significant condition of concern.
CONCERN i.
... the very idea of conducting an investigation of porosity and craters with paint present stretches the imagination."
(Paragraph F, Page 17)
~ O Attachment to TXX-6102
~
December 2, 1986 Page 8 of 9
RESPONSE
The ASME-NF visual weld acceptance criteria do not consider porosity important i
enough to even address.
Rejectable porosity (1/16 inch or greater in diameter) will not be masked by paint without at least showing a dimpling effect. This is particularly true for porosity with any appreciable depth.
In addition, the VWAC training document specifically states that craters can be evaluated through paint. The CPRT program to conduct reinspections before and after removal of paint which was described earlier will provide additional information regarding the adequacy of inspections through paint. The QC inspectors for the reinspection effort have been instructed to have paint removed if they observe a questionable indication, such as dimpling of the surface.
CONCERN "How Applicants determined that the paint was no thicker than 10 mils or whatever in the presence of craters is unknown and illogical."
(Paragraph F, Page 17) r
RESPONSE
1 l
A paint sampling program was conducted to determine the average paint thickness on fillet welds associated with supports. One hundred and eighteen (118) supports were examined at 1,132 locations to determine the paint thickness. The conclusion reached was that the average paint film thickness is approximately 10 mils. There is no doubt that paint will likely partially i
fill a crater.
It is highly unlikely, however, that paint will hide a rejectable defect. The CPRT program to conduct reinspections before and after i
removal of paint will provide additional information regarding the adequacy of inspections through paint.
CONCERN
"... if there is a lack of adhesion, visual inspection may not detect even gross discontinuities."
(Paragraph F, Page 18)
I "In addition, the uniformity of the paint cannot be counted on due to the following factors... This was especially true in the coatings area when painters were made ' instant' QC paint inspectors."
(Paragraph F, Page 18) 1 I
i 3
+
I
Attachment to TXX-6102 December 2, 1986 Page 9 of 9 i
RESPONSE
,The lack of paint adherence and paint uniformity have not been found to be a proble.. during the CPRT reinspection program.
Furthermore, the inspectors are instructed to have paint removed anytime there is a question as to his ability
-to perform an adequate inspection of the weld. There have been, however, only a few occasions where it has been necessary to remove paint due to excessive paint build-up, lack of adherence and paint runs.
CONCERN i
"... the codes require that inspection take place prior-to the application of paint."
(Parcgrhoh F, Page 19)
RESPONSE
True. This has been done. Final acceptance inspections on AWS and ASME-NF component support welds were accomplished prior to the application of paint.
The CPRT reinspection effort is not a final acceptance inspection.
It is a sample reinspection of the accepted welds.
4 CONCERN
" Industry codes and standards appear to suggest that relying upon any subset of attributes is inadequate and that the cumulative results of investigations of all attributes is much more significant than the sum of its parts... CASE is not aware of any plan by Applicants to consider cumulative effects."
(Paragraph F, Pages 19-21) b
RESPONSE
1 Each reinspection attribute is verified independently of the others and j
deviation reports are written per attribute. When the deviation reports for a particular sample or for an adverse trend of deviations associated with an attribute are evaluated, the cumulative effects of all deviations are considered whether they are welding related or not.
+
l 1
.