ML20086R710

Final Rept on Byron Inspector Reinsp Program

ML20086R710
Person / Time
Site:	Byron
Issue date:	02/29/1984
From:	COMMONWEALTH EDISON CO.
To:
Shared Package
ML20086R708	List: ML20086R705 ML20086R710
References
NUDOCS 8403010087
Download: ML20086R710 (223)

Text

- - - -

e O

Commonwealth Edison 9

Report on the Byron QC Inspector Reinspection Program Docket Nos. 50-454 and 50-455 l

l i

l f

February,1984

!$0So00Eo!$0$$$4 G PDR

g g,-

Z

,.e'_ +

i REPORT ON THE BYRON QC INSPECTOR l REINSPECTION PROGRAM l

t .

EXECUTIVE

SUMMARY

(:'

i

- FEBRUARY 1984 i

(. '

t

[37 h

L t :;

L (O

jil

I ES-1 EXECUTIVE

SUMMARY

A. OVERVIEW The Reinspection Program at Byron Station has been completed. Construction inspections performed by quality control (QC) inspectors prior to September 1932, were shown to be adequate. On this basis, Commonwealth Edison Company concludes that the adequacy of inspector activities at the Byron Station has been confirmed. Furthermore, the Reinspection Program validated the previous Commonwealth Edison conclusion that Byron construction quality is good. It is for these reasons that we reassert our belief that reasonable assurance exists that the Byron Station meets regulatory requirements and can be operated safely.

B. BYRO_N REINSPECTION PROGRAM OBJECTIVE Commonwealth Edison agreed with the Nuclear Regulatory Commission (NRC) to initiate and carry out a Reinspection Program to verify the effectiveness of former certification practices and QC inspector qualification programs that had been developed to address ANSI N45.2.6. The Reinspection Program was intended to resolve concerns about the reliability of inspections previously performed, by demonstrating that the performance of previously certified inspectors could be reproduced at an appropriate rate through reinspections performed by inspectors whose qualification and certification met current standards. Our objective was to confirm the reliability of all inspectors previously certified through a demonstration of the effectiveness of these former certification practices and qualification programs (see Figure ES-1).

i ES-2 l

C. BYRON REINSPECTION PROGRAM VALIDITY AND ADEQUACY l

The Reinspection Program was developed and implemented by Commonwealth l Edison with oversight by the Commonwealth Edison Quality Assurance (QA) organization. That oversight role included regular interaction with Commonwealth Edison's construction management and contractors' ,

management to ensure that all Program requirements were being implemented. Furthermore, Commonwealth Edison QA conducted periodic audits of the reinspection activities and, through an independent testing i

agency, performed overinspections of the reinspectors to verify tce adequacy of their work.

The Reinspection Program itself was structured in such a way that we assessed l past inspection activities conservatively (see Figure ES-2). The Program had three basic parts:

1. Sampling - Inspectors to be reinspected were chosen over the full period of construction activities, and their work to be reinspected was selected from the period during which each inspector had the least on-the-job experience. The NRC also added inspectors to our sample of inspectors.

2. Acceptance Criteria - Criteria were established for the repeatability of I inspection activities which would provide appropriate levels of assurance that the original inspections were reliable. For objective inspection attributes, the acceptance criterion was 95%; for subjective inspection attributes, the criterion was 90%.

3. Expansion Criteria - Expansion of the sampling both for the inspectors and the quantity of work to be reinspected appropriately focused on areas where deficiencies had been identified. San ple expansion to 100% of the inspectors and 100% of the questionable work could result from failure of one contractor in a single area of qualification.

ES-3 1

! i 1

Of 356 incpectors certified prior to September 1982, 110 were sampled. The r

Byron Reinspection Program sampling plan, in terms of inspector sample size, meets or exceeds that defined in commonly used statistical standards (Military Standard 105D). Approximately $36 inspector-months were reinspected, which represent 15% of the total inspection months for the contractors reviewed in the Program. The eight contractors involved will perform about 93% of the safety-related work at Byron, as can be seen on Figure ES-3. When the HVAC contractor is considered, for whom 100% reinspection of work is being performed outside of this Reinspection Program, approximately 95% of the safety-related work at the Byron site will have been subjected to at least a sample reinspection.

Conservative bias was intentionally introduced into the Program in order to provide added confidence in our results. This conservative bias took many forms. For example, the initial random sample of inspectors was expanded by the NRC staff to include additional individuals, whose qualifications the NRC staff considered suspect. Further, the work first subjected to reinspection was chosen from that performed within the first 3 months after an inspector's certification, when his on-the-job experience would be at a minimum. All

' discrepancies observed in the reinspection were counted against the inspector, whether or not the actual condition recorded by the criginal inspector was within' design tolerance. Finally, no credit was given in assessing inspect 0r qualification if a discrepancy had no design significance.

Inasmuch as all the areas of inspector qualification which could be reinspected were included in the Program, and because these reinspected areas provide a very broad basis for review of construction activities from the start of safety-related construction through September 1982, the data produced from the

< Program can be used to assess the quality of construction work at the Byron Station.-

ES-4 D. BYRON REINSPECTION PROGRAM RESULTS The results of the Reinspection Program are summarized by contractor in Table ES-1. This table also delineates the number of reinspections performed as part of the Program.

Table ES-i Reinspection Program Summary Total Objective Subjective Objective No.of Inspection No.of Inspection and Objective - Results Subjective Results Subjective

. Contractor Inspections

• Acceptable 1 Inspections *

• Acceptable 2,3 Inspections Blount Brothers 2,390 98.2% NA NA 2,390 Johnson Controls 7,812 99.4%" ~ 1,459 95.5%" s,271 Hunter 69,624 99.0 % 3,725 97.0 % 73,349 NISCo 2,792 99.6% 229 100.0 % 3,021

-Hatfield Electric 60,245 96.5% 26,660 92.0% 86,905 Powet s-Azco-Pope 8,047 96.3%" 6,607 86.2%" 14,654 Pittsburgh Testing 6,016 98.9 % 6,137 85.3%" 12,153 Peabody Testing 0 NA 163 75.5% 5 163 TOTAL 156,926 44,980 201,906

• From Appendix D, Table D-1.

• From Appendix C, Table C-1.

Notes for Table ES-1: ,

1. Program acceptance criterion is 95%.

2. Program acceptance criterion is 90%.

3. Includes concurrence by third-party inspector.

4. l00% of inspectors sampled; 100% of accessible work for inspectors not meeting acceptance criterion reinspected.

5. 100% of inspectors sampled; 100% of accessible work reinspected.

' As can be seen from Table ES-1, over 200,000 reinspections were performed as part of the Byron Reinspection Program. All seven contractors performing objective inspections exceeded the acceptance criterion. Four of seven c.

I'

i ES-5 l

contractors performing subjective inspections had results that exceed the -

acceptance criterion. Although the cumulative performance for each of the other ,

three contractors was somewhat below the acceptance criterion, this disparity has been rectified by expanded sampiing to ensure that discrepant conditions have been identified. Furthermore, it was determined that all observed discrepancies had nn design significance (see Figure ES-4).

It is worth noting that the reliability of the sampled inspector population for objective attributes is very high. These objective attributes are not strongly affected by human factors arid are, therefore, mere amenable to quantitative assessmen: (e.g., an inspection performed with calibrated instruments or inspection of a material heat number).

Although the results for the subjective attributes are somewhat lower, this was expected. That expectation was based on recognition of the strong effect of human hctors on the inspection of these attributes, which are limited to visual weld inspections. Unlike most nondestructive examination techniques, the product of which is a quantifiable record, visual weld inspections rely on subjective - interpretation. Therefore, the repeatability of such inspections is expected to be somewhat lower than for the more quantitative objective inspections.

E. QUALITY OF WORK The quality of construction work at the Byron Station was determined to be good. This is based on the Commonwealth Edison management approach to ensure quality of construction and was confirmed by the engineering and statistical evaluation of the Reinspection Program data.

1. Management Approach To Ensure Quality of Construction Commonwealth Edison has implemented a comprehensive quality program to ensure that the Byron Station is constructed properly and is of good quality. The many layers of inspections, overinspections, audits,

t ES-6 surveillances, and evaluations implemented as part of this quality program further assure us that the results of the Reinspection Program are representative of the overall plant quality.

2. Plant Quality Inferred From Inspector Qualification The Reinspection Program validates the adequacy of the inspector training

(

and certification programs in use prior to September 1982 for six out of eight contractors reviewed. These contractors are responsible for 88% of the total work at Byron. This ensures that all work performed by these contractors was adequately inspected, from which it can be inferred that the contractors' construction work is of good quality.

I

3. Plant Quality Inferred From the Reinspection Program For tne objective inspections, a total of 156,926 items were reinspected, and 3,247 discrepancies were noted. For the subjective inspections, a total of 44,980 items were reinspected, and 4,132 discrepancies were noted. The evaluation of these subjective and objective discrepancies showed that many of the discrepancies are insignificant or do not affect the design (e.g., chipped paint, documentation, measured dimensions different than those of the original inspector but still within design tolerance, etc.). The remaining discrepancies which had potential for affecting the design were evaluated further. This engineering evaluation showed that these discrepancies had no design significance. This provides direct evidence of the quality work at the Byron Station.

. F. CONCLUSIONS

1. The Byron Reinspection Program has been completed in accordance with the agreemer;t reached with the NRC staff.

2. The Program verified that the vast majority of inspectors whose work was reinspected passed the established acceptance criteria and were qualified (see Table ES-7).

~ .- -. .

ES-7

3. On the basis of the very high rate at which their inspectors were shown to pass the Program acceptance criteria, it was verified that the three major site contractors--Blount Brothers (structural), Hunter (mechanical) and Hatfield Electric (electrical), as well as Johnson Controls, NISCo, and Pittsburgh Testing Laboratories-had effective certification practices prior to September 1982, as shown on Table ES-2 below.

Table ES-2 Effectiveness of Contractor Certification Programs Inspectors Inspectors Contractors Performing Performing Program for Objective Subjective Qualifying Contractor Inspections Inspections Inspectors Blount Brothers 100% passed NA Program ef fective Johnson Corporation - 100% passed 100% passed Program effective Hunter Corporation 100% passed 100% passed Program effective NISCo 100% passed 100% passed Program effective Hatfield Electric 100% passed 100% passed Program effective Pittsburgh Testing 100% passed 92% passed Program effective h Powers-Azco-Pope 88% passed 59% passed Not uniformly effective

4. Because those contractors discussed in Item 3 above had effective QC inspector certification practices, all inspectors certified under those same practices are considered to have been properly certified.

5. For Powers-Azco-Pope, the program for certification of QC inspectors was not uniformly effective. One hundred percent of this contractor's

{- inspectors were sampled under the Program, and 100% of the

.reinspectable work was reinspected for the inspectors who did not pass the Program acceptance criteria. In this way, the adequacy of the work performed by Powers-Azco-Pope was confirmed.

?  !

ES-8

6. Peabody Testing preceded Pittsburgh Testing as the independent testing agency at the Byron Station. Peabody Testing was employed for 20 months at the inception of construction activities. For Peabody Testing,100% of their inspectors were reviewed, and 100% of the reinspectable work was reinspected. However, insufficient data was gathered (163 reinspections).

On the basis that this contractor's work was primarily overinspection of other contractors, some of whose work was reviewed in the Reinspection Program (e.g., Blount Brothers), and because the first-line inspections actually performed by Peabody were monitored by Commonwealth Edison, the uninspected work for this contractor is judged to be acceptable.

7. Because the quantity of work inspected and shown to be acceptable under this Program is extensive, covering most types of construction activity frem the start of safety-related construction at the Byron Station, reasonable assuunce exists that the overall quality of work at Byron is acceptable (see Chepter VI).

8. The results of the QC/QA checks and audits that comprise the Commonwealth Edison Construction QA Program, as well as other controlled overinspection activities which have been completed, all vet'fy the good quality of construction at Byron (see Chapter VII Section B, and Appendix E).

9. Moreover, because extensive engingering evaluation and statistical assess-ments have been made for specific work attributes witich cover most construction activities, and because the contractors whose work has been g

evaluated and assessed have performed 93% of the safety-related work on g the Byron site, a high degree of confidence exists that no major defects of design significance remain undetected at the Byron Station (see Chapter VII, Sections C and D).

I

ES-9 It is for these reasons that Commonwealth Edison concludes that the adequacy of inspector activities at the Byron Station has been confirmed. Furthermore, the

. conclusion previously reached by Commonwealth Edison that Byron construction

, quality is good has been validated. Therefore, we reassert our belief that reasonable assurance exists that the Byron Station meets regulatory requirements L and can be operated safely.

I (J

l g

(

(?

ts

(<

4

- Program Development Figure ES-1 N

aj M

mai

'll <

jn 1974  % I 1979/80 1982 Qualification of OC Oualification NRG Region 111 -

inspectom kgan + programs were -* CA) inspection  ! l I", I' , reassesset with NRC d qualificatiore programt l E

• E" " !ql s d s s k~~ U s

'l ]

l i

1982 / -

.l' Qualification "

programs were ,

reassessed with NRC participation k s i

'f g

\ E Are f-

• Syron site %d contractor 1 '

QC inspectors y

qualified? 5

-r ,

.# K  !

Before 1982 After 1982 -

• 5

~

1983/84

' CECO reinspection program ;m-implemented _ g

< C \ M u

• y g S N b

N

-' NRC and CECO ,

Qualification program

, concur that quellfica- is offective-tion program is demonstration through g reinspectior; program _g

-' s l adequate _g r

x s s d g 1 q ws v 4

.= -

i ll U

l g  ;.-;

^

4041-15 4 02 248

.* Yq

, . - E t - , . 4

. -- _ _ _ _. _ _ _ _ ...._ _ _ _ _ 4

Process for Determining the Effectiveness Figure ES-2 of a Contractor's inspector Qualification Program

y h

,y First 3 months of " '

inspections for an inspector are ..

reinspected A N Reinspections meet Yes applicable criterion f  :

(90% or 95%)

-4 i r No An additional 1 3 months of 4 inspections reinspected s s 1I Reinspections meet \ Yes applicable criterlon I (90% or 95%)

1r No N ' ~ ~

For the arsa(s) that criterion was not met:

100% reinspection forinspector plus 50% Increase of number of Inspectors for contractor i s Have all No hspectors .-

~

for the contractor been selected?

i T

, , y,, Contractorinspector qualification program

% la determined to be 100% of contractor's inspectors '"'*I"* " '" C '

the selected inspec-reinspected ,

lors have been reviewed t x A N 4041-17 02 248 A

Percent of Safety Relateu Site Work Figure ES-3 L;,

Performed by Contractors Evaluated by .g the Heinspection Program +%g 31;p 3M--

y[e;'.k

?

[ R2 G:*iA Y$d2

.t. , o, yfi ,;

s.

$~d ['

3 .. } l g Hatfield Q' #

R .?i (21%) '

~1h

. g. f [[

,', Hunter

-lvg - (43'/e) .

-Of m ,.

5'

, >W '

Slount1 , ')!

, (21%) ' ,,

w u J

f >:& --

~

. j' s yQ- ..

,c

w _

t' -,.

Site work performed by

=

remaining contractors 4.3%

, oj' e Powers-Azco-Pope 5%

_ Reliable Sheet Metal -

$ (HVAC) 2%

NISCo 0.5%

g Johnson Controls 1.0%

4 Pittsburgh Testing 2.0%

2 Peabody Testing 0.2%

=

y $ 3.7%

b z n i Site work performitd by 5 contractors evaluated by

_- the reinspection program (93.7%)

i a

2 E

=

+3 .

b 4041-18 9 02-84 -248 J

,3

_ _ - . . , a

w--

Design Significanca cf Figure ES-4 Reinspection Program Results 200,000 a--

'yj[

q .

150,000 g

s

[Cfc~

. ~ . s

i: '

100,000 - ,

kbN

gc.

~ .l 50,000 if'.- ,

0 m 201,906 7079 zNo Total + ObserWd + Number of inspections discrepa ncies discrepancies with design significance 4041-19 02 248 l

1

i CONTENTS P_a_gte I. - SYNOPSIS I-l II. BACKGROUND II-I Exhibit 11-1: Reinspection Program Definitions Ill. DESCRIPTION OF THE REINSPECTION PROGRAM III-l A. Background III-I B. Program Description III-2

1. First Element - Selection of Contractors  !!I-2

2. Second Element - Selection of Inspectors III-3

3. Third Elertient - Selection of Inspector's Work III-6

4. Fourth Element - Establishment of Acceptance Criteria III-8 C. Reinspection Program Implementation III-10 D. Summary III-11 IV.- QA COVERAGE OF REINSPECTION PROGRAM IV-1 A. Overview of QA Coverage IV-1 B. First Audit of Reinspection Program Implementation IV-2 C. Overinspection of the Reinspection Program IV-4 D. Second Audit - Hatfield Electric IV-5 E. Third Audit -Implementation of Reinspection Program IV-6 F. Conclusions IV-9 V.. RESULTS OF THE REINSPECTION PROGRAM V-1 A. - Conclusions V-1 B. ~ Reinspection Program Results V-2 Exhibit V-1: Reinspection Program Results - Objective Inspcctions

/'

V . .

ii Page Exhibit V-2: Reinspection Program Results - Subjective Inspections VI. ENGINEERING EVALUATION OF DISCREPANCIES VI-l A. Overall Description VI-1 B.- Categorization of Discrepancy Evaluations VI-l

1. Categorization VI-l

?. Subjective Evaluations VI-2

3. Objective Evaluations VI-4 C. - Results of Engineering Evaluation VI-5

l. Subjective Discrepancy Evaluation VI-5

2. Objective Discrepancies VI-6 D. Conclusion VI-9 VII. ' QUALITY OF THE WORK VII-l A. Introduction VII-l B. Management Approach To Ensure Quality of Construction VII-2 C. Work Quality Based on Inspector Qualifications VII-4

1. Sampling Adequacy VII-5

2. Validation of Inspectors Certified Prior to September 1982 Vil-8 D. Inference of Work Quality from the Reinspection Program VII-9

1. . Applicability of Reinspection Program Data Tc Plant Quality Inferences VII-9

2. Evaluation of Reliabilities VII-10 E. Conclusions VII-12 Exhibit VII-1: Calculated Reliabilities for Work by Contractor

L.

l.. . ..

iii APPENDIXES A. Reinspection Results, by Type of Inspection and by Inspection Area of Certification B. Reinspection Results by Inspector C. Engineering Evaluation of Subjective Discrepancies Exhibit C-1: Summary of Subjective Discrepancy Evaluation by Contractor Exhibit C-2: Engineering Evaluation of AWS Weld Discrepancies Exhibit C-3: Engineering Evaluation of ASME and ANSI B31.1 Weld Discrepancies D. Engineering Evaluation of Objective Discrepancies Exhibit D-1: Summary of Objective Discrepancy Evaluation by Contractor E. Management Approach to Ensure Quality of Construction Exhibit E-1: Blount Brothers Corporation Exhibit E-2: Johnson Controls incorporated

. Exhibit E-3: Hunter Corporation Exhibit E-4: Nuclear Installation Services Company Exhibit E-5: Hatfield Electric Company Exhibit E-6: Powers-Azco-Pope

~

Exhibit E-7: Reliable Sheet Metal Company Exhibit E-8: Pittsburgh Testing Laboratory Exhibit E-9: Peabody Testing Services F. Answers to NRC Staff Questions

4 I-1 I. SYNDPSIS -

This report reviews the program of reinspection (hereinafter referred to as the Reinspection Program or the Program) performed at the Byron Station to demonstrate the effectiveness of quality control (QC) inspector qualification programs employed prior to 53ptember 1982. The report supersedes the interim " Report on Reinspection Conducted as a Result of Noncompliance Item 50-454/82-05-19 and 50-455/82-04-19,"

submitted to the Regional Administrator of the Nuclear Regulatory Commission (NRC)-Region 111, on January 12,1984, and it constitutes the final report on this subject.

Chapter 11 first gives a chronological listing of the events of importance to this Program up to the present. It then defines terms used both in the Program and in this report.

Chapter Ill provides a detailed description of the Rein;pection Program, including the scope of work reinspected.

Chapter IV contains a discussion of the Commonwealth Edison Quality Assurance (QA) controls over the Reinspection Program, including the generai principles employed and the audits performed.

Chapter V gives the results of the Reinspection Program.

Chapter VI discusses the engineering evaluation of the items identified during the Reinspection Program, including the cor.clusions reached from this evaluation.

Chapter VII addresses generally and specifically the quality of the work at the Byron Station, including a discussion of the methods used to assess the plant's overall quality. In this chapter, we review the program used to ensure high quality, wnich was inplemented from the start of ccnstruction activities. In this chapter we aho provide an analytical assessment of data developed through the Reinspection Program, including confirmatory statistical inferences from which overall conclusions on the quality of ccnstruction can be drawn.

1-2 Also included in this document are six appendixes. Appendixes A, B, C, and D, which were contained in our original report, have been significantly changed and clarified for this revised -report. Appendix E discusses the Commonwealth Edison management approach to ensure quality of construction. Appendix F provides responses to formal

questions from the NRC staff.

t - - -

t 11 - 1 II. BACKGROUND Construction of safety-related structures began at Byron in January 1976, with the issuance of the construction permits for the Byron Station, Units 1 and 2. Contractors working on the site were required by Commonwealth Edison, among other things, to develop programs and procedures for the purpose of certifying the qualification of quality control (QC) inspectors. The certification programs were based on the 1973 version of ANSI standard N45.2.6. This standard set forth general guidelines for the type and amount of training and experience necessary to certify QC inspectors. These guidelines permitted wide latitude and discretion in the development of the certification process. All on-site contractors developed certification programs that qualified their QC inspectors in accordance with the ANSI standard.

In 1978, the ANSI standard was revised to provide more specific guidance with respect to the type and amount of training and experience needed to properly certify QC inspectors. Though the certification processes were developed against the 1973 version of ANSI standard N45.2.6, Commonwealth Edison concluded that the on-site contractor programs were, after re-evaluation, still adequate to satisfy the 1978 s

revision. Nevertheless, these certification programs for QC inspectors were revised to reflect several concerns expressed by the NRC as a result of an inspection conducted in 1979. Commonwealth Edison conformed to the NRC staff's interpre-tation, effected the necessary revisions, and thereby resolved the deficiency noted by the NRC in its 1979 inspection report.

A special NRC inspection was conducted at Byron during the spring of 1982 by a Construction Assessment Team (CAT). The CAT findings were published in IE Report Nos. 50-454/82-05 and 50-455/82-04. One finding questioned the adequacy of the on-site contractors' programs for qualifying, and thereby certifying, QC inspectors.

Specifically, tne NRC inspectors found deficiencies in (1) the contractors' evaluations of initial inspector capabilities, (2) the documentation of initial certification, and (3)

• American National Stancards Iratitute (ANSI) Standard N45.2.6 Qualifications of h*cection, Examination and Testing Personnel for Nuclear Power Plants.

s II-2 the criteria used to establish inspector qualification. Although there was no finding that these deficiencies had compromised the quality of construction, the NRC adopted the position that the site contractors' QC inspector qualification programs had to be upgraded and that the quality of the inacctions already completed required verification.

The NRC's position was based, in part, on a modified interpretation of the 1973 version of ANSI standard N45.2.6. Commonwealth Edison accepted the NRC's

-interpretation and began revision of the on-site contractors' certification programs for QC inspectors in -June .1982. This revision established documented minimum requirements for education, experience, on-the-job and classroom training, testing, and examinations as criteria for QC. inspector certification.' By April 30,1983, the site contractors had recertified all QC inspectors against the revised certification requirements. Commonwealth Edison's Byron QA organization conducted a 100%

review of -the contractors' certification programs and concluded they had properly implemented the new -' certification requirements. The NRC accepted the QC inspector certification program and concluded that the inspectors certified under the program were qualified. .

' The recertification effort that occurred in 1983 did not alleviate the NRC's concerns with. respect to QC inspections. conducted prior to September 1982. Consequently,

- Commonwealth Edison agreed to develop a Reinspection Program to verify the effectiveness of former certification practices and inspector qualification by re-examining, on a sampling basis, inspections performed by QC inspectors certified prior to September 1982.

. Exhibit 1I-1 provides definitions of the terms used in the Reinspection Program and in this report.

3 4

EXHIBIT !!-l Page 1 of 4 EXHIBIT 11-1 REINSPECTION PROGRAM DEFINITIONS A. INSPECTION The checking of work, against established criteria, by an individual trained and qualified to check such work. The inspection must be conducted by someone other than the individual who did the work.

B. REINSPECTIONS

- The inspection of work af ter an initial inspection has been made. Reinspections must be performed by an individual other than the initial inspector.

C. AUDITS A documented activity performed in accordance with written procedures or checklists to verify, by examination and evaluation of objective evidence, that applicable elements of a QA Program have been developed, documented, and effectively implemented in accordance with specific requirements.

D. SURVEILLANCE An examination of documentation, performance, or products at the location where work is being performed to verify that established requirements are being met. Though similar to an auditing function, surveillance is usually limited -to observing specific work activities, operations, or items against documented requirements such as those included in procedures, specifications, work packages, etc.

l,

EXHIBIT 11-1 Page 2 of 4 i

I E. WORK NOT REINSPECTABLE Certain work activities _were not reinspectable in the Reinspection Program.

That work is divided into two types, inaccessible and not retreatable, both of  !

which are discussed below.

1. Inaccessible: A condition where " extensive dismantling" would be required to gain access for a reinspection. Examples are: piping, steel, or conduit which is embedded in concrete; piping alignment; or cable splices. (Note:

fireproofing, paint, and insulation do not make an item inaccessible.)

2. Not recreatable: A condition where a process or event cannot be recreated for reinspection. Examples are: an item is reworked as a result of a revision; pull force during cable pulling tension; interpass welding temperature during welding; receiving inspection, rigging, etc.

F. ATTRIBUTES AND ELEMENTS

1. Attributes: Broad contractor work items that have inspection activities performed on them (e.g., visual welding, documentation, conduit installation, etc.). Typically, attributes are the broad areas of qualification within which inspectors are certified. '!

l

2. Elements: Sub-parts of attributes. They are the individualitems inspected (e.g., conduit hanger location, work process sheet data /signoffs, etc.).

G. DISCREPANCIES Lack of conformity between an initial inspection result and a reinspection result, t-

EXHIBIT II-l Page 3 of 4 H. QC INSPECTOR REINSPECTIONS

1. Objective: An inspection element that can usually be quantified or measured. Examples are: material, size, shape, traceability, dimensional configuration, etc.

2. Subjective: An inspection that cannot be truly measured and is dependent upon human senses and judgment. An example is a visual weld examination without supporting gauges.

I. ACCEPTANCE CRITERIA The ' minimum requirements necessary to establish acceptability. For the purpose of this Reinspection Program, the following acceptance criteria apply:

1. For objective inspections - 03% agreement rate

2. For subjective inspections - 90% agreement rate

- J. THIRD-PARTY REVIEW An inspection of an observed discrepancy identified during the Reinspection Program to determine true rejectability for a subjective inspection (visual welding only).

K. RELIABILITY, CONFIDENCE LEVEL, SINGLE SAMPLING PLAN, AND MULTIPLE SAMPLE PLAN

1. Reliability: The probability that a component meets a specified acceptance criterion.

EXHIBIT 11-1 Page 4 of 4

2. Confidence level: Tne probability that all components in the plant have reliability equal to, or greater than, that computed based on the results of a sample of components inspected.

3. Single sampling plan: A one-stage sampling plan. The decision for acceptance (or rejection) of the total population is made based on the ,

results of this single-stage evaluation.

I

4. Multiple sampling plan: A multi-stage sampling plan where, based on the cumulative results of the successive sample evaluations, a decision is made at each stage on whether to accept or reject the total pcpulation, or to continue with further sampling.

L. - UNIT CONCEPT INSPECTION A special inspection by the site independent inspection agency under the direction of Commonwealth Edison QA Department. It consists of visually examining field-installed items and vendor-suppiled equipment for compliance with, or deviations from, design documents and drawings.

M. FINDING Violation of, or deficiency with, the requirements of established documents for which Commonwealth Edison requires clarification or corrective action.

N. OBSERVATION An item for discussion but not considered to be a finding.

I.

III-1 III. DESCRIPTION OF THE REINSPECTION PROGRAM

A. BACKGROUND Commonwealth Edison's General Office Manager of QA and representatives of Commonwealth Edison's Project Management and Licensing Departments undertook to structure a Reinspection Program that would resolve the NRC's concerns with respect to the adequacy of the qualifications of the contractors' QC mspectors. The focus of the Reinspection Program was to assess the

- qualifications of the site contractors' QC inspectors who had performed inspections curing - the 1976 to - September 1982 time frame. This was accomplished by using QC inspectors, who were qualified under the certification processes accepted by the NRC in mid-1982, to reinspect the original inspectors' work. The original inspection and reinspection results were then compared and evaluated to assess the original inspector's qualification.

Each contractor used its own QC inspectors, who, as indicated above, were properly _ certified, to conduct the Reinspe . tion Program. No inspector was permitted to reinspect his own work. Rein.;pections were performed using the same requirements that had been used in the initial inspection. This was the case even if design changes or inspection criteria were relaxed subsequent to the initir.! inspection. It was deemed important to recreate the conditions of the original inspection because the objective of the Reinspection Program was to evaluate the quality of the original inspector's performance.

The Reinspection Program was documented in a letter dated February 23,1983, from Commonwealth Edison to the NRC Region III Regional Administrator.

This Program description had been discussed with the NRC staff prior to the formal submittal date. The Program description was accepted by Region III in a letter from Mr. R. L. Spessard to Commonwealth Edison's Mr. C. Reed dated March 22,1983.

_ _ _ _ _ _ - _ .m___ _- _ _ _ _ - _ _ _ _ - _ - _ _ _ _ _ - - _ _ _ - _ _ _ _

III-2 B. ' PROGRAM DESCRIPTION j

1. First Element - Selection of Contractors The first element of the Program was to select the site contractors whose QC inspectors were to be subject to reinspection. Eight of the 19 contractors who had performed or were performing safety-related work at the Byron Station were selected. These site contractors were:

a. Blount Brothers Corporstion (structural contractor)

b. Johnson Controls incorporated (HVAC controls contractor) i

c. Hunter Corporation (mechanical contractor)

d. Nuclear Installation Services Company (NISCo) (NSSS equipment

-erection)

e. Hatfield Electric' Company (electrical contractor)

f. Powers-Azco-Pope (process instrumentation contractor)

g. Pittsburgh Testing Laboratory (independent inspection agency)

h. ' Peabody Testing (independent inspection agency)

The work' performed by these contractors amounted to approximately 93%

cf the safety-related work r.t the Byron Station.

Of the 11 contractors excluded from the Program, three were excluded

, because they were not subject to ANSI N45.2.6-1978 and, hence, the qualification of their QC inspectors was not in question. Three other contractors were already undergoing extensive reinspection of their work, thereby rendering it unnecessary to address the question of their QC inspector qualification.

The remaining five were excluded from the Reinspection Program because their wcrk was neither-- accessible nor recreatable for purposes of reinspection. Work that was not accessible included inspected work embedded in concrete. Work that was not recreatable includeo such activities as rigging operations involving the placement of eculpment prior e

r,,n . - - ,.. . -

[

III-3 to installation. The work performed by these five contractors was -

subjected to varying forms of overinspection and surveillances, thus assuring the adequacy of their work and inspection performance.

2. Second Element - Selection of Inspectors The second elemer.t of the Reinspection Program involved the selection of the inspectors for reinspection. All QC inspectors for two contractors (Powers-Azco-Pope and Johnson Controls) were reinspected to the extent their work included reinspectable items.

The work cf the QC inspectors of the six remaining contractors was reinspected by a sampling technique. To ensure a representative selection of inspectors from the total population, Commonwealth Edison compiled rosters of the six contractors' QC mspectors. The names of the inspectors were listed chronologically by date of certification. The first inspector on '

each roster was selected and every fif th inspector thereafter was included in the Program. Af ter the original sample population was selected, the NRC Senior Resident Inspector reviewed the sample and added two to four names to each contractor's group of inspectors. For example, four names were added to the sample population for Hatfield Electric.

In order to provide a uniform basis for assessing an inspector's per-formance, a minimum number of reinspections was identified. Generally, an inspector had to accomplish a minimum of 50 inspections during the period of interest. In the case of independent testing agency personnel (Pittsburgh Testing and Peabody Testing), 25 inspections were accepted because of the limited population of inspections for the typical inspector.

If an inspector did not have these minimum quantities, the next inspector listed chronologically was substituted. The substitute inspector's work was reinspected and the results were included in the Program data base, In those cases for which reinspection was initiated for the original inspector but a " minimum quantity" was not reinspectable, all reinspections actually performed were included in the Program data base. For one centractor

{

III-4 (Peabody Testing), the substitution feature resulted in all inspectors who -

had reinspectable work being reinspected.

Of 356 QC inspectors employed by the eight contractors prior to September 1982, Commonwealth Edison sampled the work of 110 (31%).

The total number of inspectors, the number of inspectors reinspected, and the percentage of the total number of inspectors included in the Reinspection Program are presented in Table 111-1.

Table III-l Quantity of Inspectors Reinspected Total No.of Percentage of Population of Inspectors Inspectors Contractoc Inspectors Reinspected Reinspected

)

Blount Brothers 28 8 29 %

• Johnson Controls 7 5 71%

Hunter 84 22 26 %

NISCo S 4 50 %

Hatfield Electric 86 23 27 %

• Powers-Azco-Pope 21 19 90 % l t Pittsburgh Testing 85 23 27 %

• Peabody Testing 37 6 16 %

. TOTAL 356 110 31%

• All of the inspector population was reviewed for possible reinspection.

There were no reinspectable items for those inspectors not included.

Ill-5 Table !!!-2 identifies the number of inspectors who had inspections -

physically reinspectable versus the total population of inspectors since safety-related construction started in 1976. This table demonstrates that the inspectors whose work was reinspected span the entire period from the beginning of the work to 1982.

Table 111-2 Distribution of Inspectors Reinspected by Contractor by Year

• 1976 1977 1978 1979 1980 1981 1982 TOTAL Blount Brothers 2/7 2/3 0/2 3/12 0/2 1/2 8/28 Johnson Controls 1/2 3/4 1/1 5/7 Hunter 1/2 2/6 1/6 2/7 4/19 9/3l 3/13 22/84 N15Co 1/3 1/1 1/2 1/2 4/8 Hatfield Electric 1/3 2/4 1/1 2/5 15/60 2/13 23/86 Powers-Azco-Pope 2/2 5/5 9/10 3/4 19/21 Pittsburgh Testing 6/34 6/16 3/18 2/5 3/7 3/5 23/85 Peabody Testing 1/23 5/14 _

6/37 TOTAL 4/32 16/60 10/31 11/40 15/37 40/116 14/40 110/356

• To September 1982.

Note for Table III-2:

In the above table, the numbers shown as x/y indicate: the number of inspectors reinspected versus the total number of inspectors certified.

III-6 s

3. Third Element - Selection of inspector's Work -

The third element of the Reinspection Program involved the selection of the amount of each inspector's work to be reinspected. A random sampling of each selected inspector's work was not judged adequate to indicate the inspector's initial qualification. Rather, the first 3 months of each inspector's work was determined to be a conservative measure of that inspector's _ qualifications. This time frame is particularly significant because deficient work of any inexperienced or ill-trained inspector is most likely to appear during the early months of any jcb performance, before time permits on-the-job experience to correct any original deficiencies.

1 As indicated above, a sampled inspector's first 3 months of inspections were reinspected. If the acceptance criterion was not met for the first 3-month period,- the inspector's certification was considered suspect. In order to validate the trend established after the 3-month reinspection, an expanded reinspection covering the second 3 months of the individual's inspection tenure was reinspected. If the trend developed based on the second 3-month period alone did not meet the acceptance criteria, the inspector was judged unqualified. In this event,100% of the inspections performed by that inspector of the typo found to fail the acceptance criteria wera reinspected. In addition, the original inspector sample popu-lation-was expanded by as much as.50%, depending on the number of inspectors still available for inclusion in the Program. If an inspector had no inspections beyond 3 months and did not meet the Program acceptance criteria, the next inspector listed chronologically was substituted. The

. qualifications of the original inspector in such a case were considered indeterminate, but his results were retained in the Program data base, r.

If expansion was required, Commonwealth Edison's selection of the inspectors to be added to the sample was made from an overall list of inspectors certified in the specific area where the unqualified inspector was found. Thus, the expansion focused specifically on areas where i

l III-7 qualification was suspect. This approach resulted in a very broad sampling of the discrepant area of qualification when a single inspector failed to meet the acceptance criteria.

The inspectors selected for reinspection and their corresponding inspection durations were compared to the c'uration of inspections accumulated by all inspectors. The results of the comparison are presented in Table III-3.

This table shows that 15% of all work inspected during the period of interest (i.e., start of safety-related construction in 1976 up to September 1982) was subject to reinspection as part of the Reinspection Program.

Table III-3 Quantity of Inspector Work Period Reinspected by Corttractor Percent of Reinsoected Inspection Total Accumulated Inspection Months '

Inspection Months Months Reinspected Blount Brothers 424 89 21%

Johnson Controls 60 22 37 %

Hunter 1,107 65 606 NISCo 51 12 24 %

Patfield Electric 628 68 11 %

Powers Azco-Pope 152 145 95%

Pittsburgh Testing 1,015 115 11%

Peabody Testing 181 20 11%

TOTAL 3,618 536 15 %

III-8 r

4. Fourth Element - Establishment of Acceptance Criteria -

The establishment of acceptance criteria constituted the fourth and final element of the Reinspection Program. Acceptance criteria were established that Commonwealth Edison judged would provide reasonable assurance of the adequacy of the inspector's qualifications. Two criteria were established which, if met, provided the basis for concluding that an inspector had been properly certified. For inspections not strongly affected by qualitative interpretation, for example, an inspection performed with calibrated instruments or the inspection of a material heat number, agreement between the reinspection and the original inspection was required to meet or exceed a rate of 95% (objective inspections). This acceptance criteria was considered a reasonably conservative acceptance level, recognizing that unintentional human error prevents c.omplete perfection. Moreover, many objective inspections require some subjective judgment on the part of the inspector, preventing complete agreement between the two individuals.

For inspections known by past experience to be heavily influenced by qualitative interpretation, such as surface quality of fillet welds subject to visual inspection without supporting gauges, agreement between the rein-spection and the original inspection was required to meet or exceed a rate j of 90% (subjective inspections). This acceptance criterion applied only to -

visual welding inspections. Again, the 90% acceptance level represents a reasonable judgment recognizing the margin for disagreement between inspectors with respect to judgmental decision making. In order to further ensure that visual weld inspection results were consistent and accurate, the Reinspection Program accepted by the NRC staff provided for a third-party review of discrepant results. The third-party review was effective in rectifying overly conservative interpretations by the reinspector. This judgment was confirmed by Region ill in IE Reports 50-454/83-39 and 50-455/83-29.

- -m _ . .

L III-9 Items graded as acceptable were defined as those for which the reinspector agreed with the condition recorded by the originalinspector on the original inspection record. Without that agreement, the item was graded as unacceptable. These statistics were compiled and recorded in such a way that correlation to the original inspector could be accomplished. The Frading was executed in this manner regardless of whether or not the installed item was in conformance with design drawing tolerances. For example, if the original inspector recorded a value for a finite dimensional measurement and the reinspector could not obtain the same measured value, we graded the item as unacceptable (and hence an observed discre-pancy), even if the installed product d'.mensions were acceptable to design drawing tolerances: if the original inspector identified the distance between two points as 3 feet 2 inches, but the reinspected value was 3 feet 1-5/16 inches (a difference of 11/16 inch), a discrepancy was recorded even though both measurements meet the requirements of the design drawings, i.e., they are within the design tolerance. In addition, if the original inspection record only required a notation of acceptable or unacceptable, without additional detail, the reinspector graded the item as unacceptable if the installed condition did not meet the quality standard requirements but was recorded by the original inspector as acceptable. An example of an item that is recorded only as acceptable or unacceptable is crimping of a wire termination lug.

It is also important to an understanding of the Reinspection Program data to recognize that the inspector selection process identi;ied an individual.

That individual may have performed objective or subjective inspections.

Some individuals performed both types of inspection. Table Ill-4, which follows, delineates the quantity of inspectors performing each type of inspection. From this table, it can be seen that of the 110 individual inspectors reinspected, 82 (38+44) performed objective inspections and 72 (28+44) performed subjective inspections.

l l

e->

III-10 Table III-4 -

Quantity of Inswtors Performing Subjective or O 2iective Inspections No. of Inspectors No. of Inspectors No. of Inspectors Reinspected for Total Reinspected for Reinspected for Subjective and No.of Only Objective Only Subjective Objective Inspectors Contractor Inspections Inspections Inspections Reinspected Blount 8 NA NA 8 Brothers g Johnson i NA 4 5 Controls Hunter 3 2 15 22 NISCo NA NA 4 h Hatfield 15 6 2 23 Electric j Powers-Azco- NA NA 19 19 I Pope Pittsburgh 9 14 NA 23 Testing Peabody- NA 6 NA 6 Testing TOTAL 38 28 44 110 )

l C. REINSPECTION PROGRAM IMPLEMENTATION The Reinspection Program began in February 1983, when Commonwealth Edison's Project Construction Department met with specific contractors whose inspection work was to be reinspected. At those meetings, the purpose and content of the reinspection activities to be performed and the requirements of the February 23, 1983, letter to the NRC staff which outlined the Program and criteria for reinspection were discussed. Thereaf ter, weekly meetings were held with the participating contractors and Commonwealth Edison's Site QA organization to communicate and resolve questions requiring clarification, to establish methods to be employed in recording results, and to determine action to be taken on discrepancies observed in the course of the reinspection effort.

The primary directives given were that reinspections were to be conducted w _.

III-11 employing the original criteria used at the time of the original inspection and that individuals involved in reinspection of the work could not be the same inspectors who performed the original inspection. Additionally, it was identified that the presence of fireproofing, paint, and insulation were not conditions for classifying an item inaccessible. Questions requiring clarifications were documented in an inquiry / interpretation format. All information was documented and is available.

Before the reinspection eitort was undertaken, Commonwealth Edison recog-nized that, in all p obability, discrepancies would be found. In order to create a data ba:e sufficient to determine whether the discrepancies were either non-critical or critical to the design basis requirements, the contractors were directed to record the reinspection results but not to implement corrective action immediately. This approach was taken so that the "as found" physical conditions could be observed at a later date for possible detailed analysis. In subsequent meetings, the various contractors were directed to incorporate the unacceptable conditions into their non-conformance system in order to implement corrective action.

All of the inspectors used in the Reinspection Program were qualified to current requirements. No inspector used for reinspection looked at his own work. In some cases (i.e., conduit as-built reinspection), the reinspection criteria were made more stringent than the original installation tolerances to ensure that all discrepancies were captured. This resulted in many observed discrepancies that were not valid discrepancies.

D.

SUMMARY

The Byron Reinspection Program sampled work performed by eight site contractors, who together performed over 93% of all safety-related work at Byron. The sampling plan employed in the Program resulted in the inspectors in the initial sample (110 mspectors out of 356 certified prior to September 1982) being subjected to reinspection. The sampling plan employed has been

L v -

i III-12

.. validated as discussed in Chapter Vil, Section C, and leads os to conclude that

- the -results are representative of ',the full scope of safety-related work performed by the contractors subject to reinspection. .

-s L I 4

1 I

b. g s

t

.(

i i

hus un gumiiigWuin u iiug

IV-1 IV. QA COVERAGE OF REINSPECTION PROGRAM A. OVERVIEW OF QA COVERAGE In mid-1982, prior to the Reinspection Program, Commonwealth Edison required that the site contractors' QC inspectors be recertified to ANSI N45.2.6-1978 in accordance with guidelines established by Commonwealth Edison to meet the latest NRC interpretatica of this standard. Commonwealth l

= Edison Site QA personnel were assigned to work full time with the contractors to achieve these recertifications, which were completed by April 30,1983.

8 In February 1983, the Reinspection Program, described in Chapter III, was y developed to reinspect the work of past QC inspectors. The Manager of QA j was directly involved in the development of the concepts and approach to be

followed in the Reinspection Program. Commonwealth Edison Site QA then f actively followed the reinspectica activities and also performed specific audits j and surveillances to ensure the Reinspection Program was properly 8 implemented and was in compliance with the established Suidelines. Other j objectives of the Commonwealth Edison audits were: to ensure that the

=

= inspectors chosen to be reinspected were not selected to provide a bias in favor 1 of their company, to ensure that all discrepancies identified during the d Reinspection Program were processed for evaluation in accordance.with the s appropriate discrepancy reporting system, and to ensure that all the inspectors 5 -

performing the reinspection were properly qualified.

x

= A summary of the QA coverage of the QC inspAtor recertification and the

-2 Reinspection Program is as follows:

1 c3

1. For the QC inspector recertifications, site QA reviewed and accepted all q contractor recertification programs prior to their implementation. This enabled QA to provide comments on the contractor QC inspector g certification prograni and pra.edures.

-a M --

'T E =

M m

~

i'2R_ ____. . . .

d

,[ IV-2

2. All contractor QC inspector recertification packages were reviewed by site QA. This included a 100% review of certifications of all inspectors performing reinspection activities in the Reinspection Program.

3. Site QA participated in the weekly Reinspection Program meetings held at Byron in which the methodol9 gy and interpretation of the Program were established. Site QA ' personnel were assigned specifically 'o the Reinspection Program.

4. Site QA directed the independent testing agency (Pittsburgh Testing) to conduct a special inspection of the site contractor inspection personnel engaged in the Reinspection Program. The purpose of the review was to a u rtain ~the acceptability of the reinspections performed by the

- contractors' reinspection personnel. The approach fallowed in this special review was to independently inspect work which had been reinspected and accepted by the contractors' QC personnel and compare the results of the two reinspections.

5. Site audits and surveillances were performed to verify that the site contractors properly recertified their QC inspectors and were in compliance with the Reinspection Program. Twenty-two audits and 64 surveillances have been conducted since June 1982.

These activities demonstrate QA's involvement and coverage cf the QC inspector recertification process and the Reinspection Program to ensure compliance with the Program criteria as the Program was progressing and i1,

$l y

to verify the integrity of the end results. The QA activities began when .

j the programs were initiated and continued until the programs were _

completed. . ,j B. FIRS f AUDIT OF REINSPECTION PROGRAM IMPLEMENTATION When the Reinspection Program was established in February 1983, and reinspecticns began in late March, the Site QA Department performed audits . .

and surveillances to monitor the reinspection activity. The first audit was .?

,y..- , . . ._- - .... +. - p. ,. . ;,

.. . , ,.n -

s w. ., .: . . .

. .. q. , a w.L;. 8. m.y A; ~+y :y . . .; . .e, a,

.;; : .; yc ; c?. .w . . :.n: Q_ s - s.7 ;+. m as

~i

LA.

y ..[

u.p

. y_F IV-3 }.[..

i. , n

<; . . i. ,7. t

• { * : ; .

.g . ;

performed June 21 through July 6,1983. The audit (No. 6-83-66) was conducted f.'g .3..

j, ' . *

.4- by a six man team and covered the activities of the following seven site 'ik "

C contractors: Blount Brothers, Johnson Controls, Hunter, N!SCo, Hatfield 7/ J,..

Electric, Powers- Azco-Pope, and Pittsburgh Testing Laboratory. @

... .: -.i.. .-

.-r..

The purpose of the audit was (1) to verify that the Reinspection Program was 3.! .

..11: being implemented in accordance with the program description outlined in .- i. ': ..s'

u.. .

Commonwealth Edison's ietter dated February 23,1983, to the NRC staff, and }.7) .

f..

(2) to determine if the contractors clearly understood the guidelines. hh i.4 3;.

1l .5 The scope of the audit included the following: ...f.1 p.

-; .:s

}p e Reinspection sample size

'I.. .

e Application of inaccessibility criterion f<.0,: _4.

} e Third-pirty overinspection review p y ,

/d.,

e Disposition of discrepancies l [

y;j e Documentation of inspection results 1 k.. .

[- e Qualifications of reinspection personnel L- l

~

j

i. : . Ye

. y, y- _

Overali, the audit team concluded that the contractors were implementing the .

[f

.' ' ? Reinspection Program; however, it was also concluded that a better program g.g..

- #n; q2.y .

definition was required in order to clarify how the reinspection guidelines were to be interpreted. This later conclusion was based on the observation that y h

.{ ; $ there was some misunderstanding among the contractor s regarding how to F$,h

,n,

..{ handle special problems. For example, in one case, a component had been .f. s r . ..q .

{7 inspacted in January 1980, and again in May 1982. The contractor incorrectly q,g

(;3 classified the first inspection as " inaccessible" without evaluating whether or N

[-LI. .

not the second inspection completely superseded tne earlier inspection. In k, another case, a contractor had a QC inspector who had been certified in several

-w .d.

. s h. f.

?,

different disciplines within the first 3 months of his employment, but the ,f s t

~.? contractor had only reinspected the first discipline in which he was certified. M.

s . -1 i "h In total, eight observations were identified which addressed issues such as these M*2 4 '; that needed clarification. As of October 26, 1983, all of the eight audit %y.;

. + -

'd .- observations were properly corrected and closed by Commonwealth Edison p";.g. .

h QA. In addition, it was established that al!" interpretations," which are needed 'k 7.V .P-t.

I V ., .

. :., y , f,,, y

..e -, ..

b

. .:.. f:..;,,,..s.,Q. . y;'_ ;r

. y *.p

.',.- . y ;. . _ _ . ' . . . .s,

.;.' l ;A.lp'. +?

~.b .*

.,F.

,, i

,,,%3 %. ; Y. .- > .;it. .,. W ,m.-

5. , . %,,. >,J%.-R,.

. 'k.v ..t. _.*

l {' '

..f , :. ,~. g ...

IV-4 to clarify the Program, be put in writing. These-interpretations have been' reviewed and found acceptable by Commonwealth Edison QA.

One tinoing was identified in the audit. It was found that certain contractors had not yet initiated the documentation necessary to correct or dispositicn the discrepancies identifieo in the reinspection process in accordance with the requiremen:s of the specific contractor's QA program. The discrepancies were identified, documented, and controlled, but no action t.ad yet been taken to write a formal discrepancy or nonconformance report to correct the discre-pancy in the field. Commonwealth Edison QA concluded that failure to write the discrepency or nonconformance report was a findmg. This finding, however, did not have any direct relationship to the validity of the Reinspection Program results. Furthermore, when discrepancy and nonconformance reports were issued, the documentation requirements of the contractor's QA program were satisfied and the finding was closed. The four contractors were Hunter, Hatfield Electric, Pittsburgh Testing, and Blount Brothers.

C. OVERINSPECTION OF THE REINSPECTION PROGRAM In order to ensure that QC inspectors performing reinspections were fully

.. qualified and performing acceptably, site QA directed Pittsburgh, Testing to.

perform a special unit concept inspection (UCI) to determine if the Pittsburgh Testing inspectors could independently arrive at the same inspection results as the contractor's QC inspectors who were performing the reinspections. This overinspection was performed during the period between . August 1 and September 19,1983. l l

The UCI inspectors rechecked the work of seventeen reinspectors who were employed by Hunter, Blount Brothers, NISCo, Johnson Controls, Powers-Azco-Pope and Hatfield Electric. ' Work which these contractor reinspectors found to be acceptable was rechecked by the Pittsbwgh Testing inspectors. Of about 1,185 items checked by overinspection, only nine (involving six inspectors) were deemed to be discrepant af ter independent third party review.

IV-5 This UCI overinspectim by Pittsburgh Testing, which compared the Pittsburgh Testing inspectors to the contractor QC reinspector, also served another purpose. The results provided evidence that the QC inspectors did not provide a bias in favor of their company. The Pittsburgh Testing inspectors were totally independent from the contractors being reviewed, and they inspected the contractors' wuk to the original acceptance criteria. The results from this overinspection demonstrate with a high level of confidence the integrity of the contractors' QC inspectors who were involved in conducting the Reinspection Program.

The aforementioned UCi was performed by Pittsburgh Testing (tt.e Byron independent testing agency) personnel who were certified in accordance with the requirements of ANSI N45.2.6 (1978).

D. SECOND AUDIT - HATFIELD ELECTRIC Commonwealth Edison had previously identified documentation problems associated with the manner in which Hatfield Electric generated and maintained various inspection records during the early years of construction.

Although this matter was being rectified by Hatfield Electric, a concern arcse as to whether Hatfie!d Electric was adequately dealing with the prob!em under the Reinspection Prograr.% Adequate documentation is necessary to permit the traceability of the inspections to the original inspector. The attribution of inspections to the wrong inspector would distort the results of the Prcram.

Hence, Commonwealth Edison Audit No. 6-83-124 was initiated, which was conducted by a three-man team during the period between August 24 and September 1,1983.

It was concluded in the audit that Ha: field Electric had instituted special precautions to maintain the integrity of the Reinspection Program. For example, during the audit, the auditors examined the manner in which Hatfield selected the work to be reinspected for each inspector. It was found that where Hatfield Electric could not determine which inspector originally performed an inspection on a particular weld, the weld was not included in the reinspection population. Bis was in accordance with the interpretations

IV-6 l

applied to the Reinspection Prograrn and it helped achieve accuracy because it eliminated using an> original inspection data unless it was positively identified to a specific inspector.

To help improve the record selection and correlation process, a computerized data base management system, which was in the process of being finalized, was used to reconcile some weld travellers to hangers. This reconciliation between inspectors and har:gers ensured that the initial hanger inspections assigned to ea h inspector were correct, and it provided an accurate means for identifying those hangers whi.ch did not have cenplete inspection records.

In two situations, it was found that it could not be determined which original-inspection records were associated with a specific component in the field. The

- first situation involved the absence of information on weld traveler cards regarding the precise location of the hanger, and whether weld inspections had been performed originally. The second situation involved comparable documentation problems for Hatfield Electric and Reliable Sheet Metal hangers. Both of these situations were identified as an audit finding. New inspections were initiated for these hangers. Since these original inspections have nc documentations which would tie those inspections to original inspectors, the results of the new inspections were not input to the ,

Reinspection Program data.- f l

Hatfield Electric has completed the reconciliation of hanger and weid inspections, which are documented on the weld travellers. For hangers that have weld traveler cards with incomplete data, new inspections are being performed. These new inspections are in addition to, and outside the scope of, the Reinspection Program. These inspections are expected to be complete in March 1984. Audit No. 6-83-124 remains open pending completion of these

' inspections.

E. THIRD AUDIT -IMFLEMENTATION OF REINSPECTION PROGRAM Subsequent to the issuance of the " Preliminary Report on the Reinspection Program" (October 23, 1983), an add.tional audit was performed. Audit No.

1

I IV-7 6-83-93 was conducted between November 14 and November 17,1983, to ensure-that the data in that Preliminary Report were valid and reliable.

The scope of the audit covered the following areas:

o Correction of discrepancies identified as a result of the Reinspection Program o Expansion of sample size e Independence of reinspection personnel e Accuracy of Reinspection Program results as reported to the NRC in the Interim Report e The design basis for the engineering evaluation of visual weld inspection discrepancies, as described in the Interim Report e Qualifications and documentation of the third-party inspectors e Basis for Project Construction Department " interpretations" with regard to the Reinspection Program The following four areas were reviewed for each of the seven contractors that were on site and which were involved in the Reinspection Program:

1. Correction of discrepancies All contractors were found to have documented the discrepancies identified in the Reinspection Program on the appropriate discrepancy or nonconformance report. For Pittsburgh Testing and Hatfield Electric, welding discrepancies were being documented on inspection reports and weld maps. These reports and maps were being accumulated for submittal to engineering for evaluation and disposition under a Commonwealth Edison nonconformance report.

V IV-8

2. Eynsion of an inspector's reinspection sample size and the number of-inspectors to be reinspected upon failure as defined by Program Requirements All contractors were found to have expanded the sample size in accordance with the requirements. Also, the results given in the Interim Report were properly presented.

3. Independence of the reinspection personnel A 100% review of the QC inspection personnel of all contractors involved in the Byron Reinspection Program verified that none had been involved in the reinspecticn of work that they had originally inspected or had reviewed and accepted.

4.- Accuracy of results reported in the interim report During this aedit, all ccatractors, with the exception of Pittsburgh Testing

, and Johnson Controls, were found to have properly documented and reported the results of the reinspections. For Pittsburgh Testing, it was found that some welds were first rejected by 1:te reinspector. Third-party concurrence was obtained; and, then, because of clarified acceptance criteria, Pittsburgh Testing reinspected and accepted the weld. This final acceptance by Pittsburgh Testing had inadvertently not received third-party concurrence; and, therefore, QA identified this as a finding. Fot all .

welds processed in this manner, third-party review was ulumately

obtained, and the audit finding was closed.

In the case of Johnson Controls, two minor differences existed between the number of subjective rejects reported in the Interim Report and the third-

. party inspection report. This difference was determined to be a record-keeping problem and when the records were corrected the audit item was clesed. .

Also, the fM. lowing areas were reviewed to assess Commonwealth Edhon Project Construction Department's implementation of the Program:

IV-9 e The engineering evaluation for disposition of the visual weld discrepancies performed by Sargent & Lundy were reviewed against the design requirements. Calculations tha; supported the evaluation were performed in accordance with appropriate structural design standards and approved guidelines which outlined the assumptions to be followed in performing the calculations.

e Those individuals who performed the third-party review of subjective deliciencies were properly qualified for the task. Additionally, adequate documentation exists for these reviews.

e Last, those interpretations offered by the Project Cocstruction Department during the Reinspection Program have adequate basis and fall within the guidelines of the Program.

F. CONCLUSIONS On the basis of these audits, the QA Department has concluded that:

e The Reinspection Program was properly implemented in accordance with the Program requirements.

e The pet =or'nel performitig the reinspactici. were properly qualified, a The reinspection results were properly processed and evaluated.

a The corrective actions for the deficiencies identified in the Commonwealth Edison QA audits were appropriate and adcquate to resolve the audit concerns.

As a result of the above, it can be concluded that the results of the Reinspection Program provided reliable information.

V-1 V. RESULTS OF THE REINSPECTION PROGRAM A. CONCLUSIONS The Reinspecticn Program at the Byron Station has been completed. That Program involved extensive reinspection of werk performed by QC inspectors prior to September 1982. It is on the basis of that reinspection data and our subsequent review of it that we conclude the following:

1. The Program verified .that most of the inspectors whose work war rein-spected passed the established acceptance criteria and were qualified. In cases where an individual did not pass those criteria, 100% of his accessible and recreatable work on the attribute in question was reinspected.

2. On the basis of the very high rat,e at which their inspectors were shown to pass the Program acceptance criteria, it was verified that the three major site contractors-Blount (structural), Hunter (mechanical), and Hatfield Electric (electrical), as well as three other contractors, Johnson Controls (HVAC instrumentation), NISCo (equipment setting), and Pittsburgh Testing (independent testing agency)-had effective QC inspector certification programs prior to September 1982.

3. Inasmuch as the contractor programs for certification of QC inspectors were verified to be effective for the six contractors discussed in Item 2 above, all inspectors certified under those same practices are considered to have been properly certified.

4. ' Powers-Azco-Pope's inspector certification program was not uniformly effective. That contractor's scope of safety-related work annunts to less than 5% of the Byron total, of which about 1% had been completed by September 1982. One hundred percent of the QC inspectors certified prior

m. . _ . _ _

V-2 I

to September 1982 were reviewed. Nineteen of 21 inspectors were sampled, and 100% of the accessible inspections on the attribute in l question for inspectors found to be unqualified, were reinspected and that work has been shown to be acceptable. The two other inspectors performed receiving inspections, which were subject to Commonwealth Edison QA overinspection and are judged to have been performed acceptably (see Appendix E, Exhibit E-6).

5. The effectiveness of Peabody Testing's inspector certification program is indeterminate. That contractor's scope of safety-related work amounts to 0.2% of the facility total. Peabody Testing was employed at the Byron Station for about 20 months after construction began in 1976. One hundred percent of the accessible inspections were reinspected. However, due to

- the very small quantities of recreatable work, no conclusion could be drawn as to inspector certification practices. The wcrk reinspected was shown to be acceptable. This contractor was subjected to extensive review and field surveillance by Commonwealth Edison due to the nature of his scope of work. Those overinspections did not identify unacceptable performance (see Appendix E, Exhibit E-9).

B. REINSPECTION PROGRAM RESULTS The results of the Reinspection Program are best explained in tabular form.

Table V-1 provides the number of inspectors who meet the acceptance criteria.

l i ".

l V-3 Table V-1 Number of Inspectors Meeting Acceptance Criteria Inspectors Inspectors Performing Peforming

, Objective Subjective Contractor Inspections Inspections Blount Brothers 100% passed Not applicable Johnson Corporat.on 100% passed 100% passed Hunter 100% passed 100% passed NISCo 100% passed 100% passed Hatfield Electric 100% passed 100% passed Pittsburgh Testing 100% passed 92% passed Pcwers-Azco-Pope 88% passed 59% passed Peabody Testing is not displayed in the table because insufficient quantities were reinspected.

A detailed review of the Reinspection Program results is presented by contractor in Tables V-2 (Blount Brothers), V-3 (Johnson Controls), V-4 I

(Hunter), V-5 (NISCo), V-6 (Hatfield Electric), V-7 (Powers-Azco-Pope), V-8 (Pittsburgh Testing), and V-9 (Peabody Testing). The results presented in Tables V-2 to V-9 have also been retabulated, by contractor, for objective and subjective inspections (see Exhibits V-1 and V-2 respectively). In addition, the results for each of the contractors by type of inspection (objective or subjective) and area of qualification (attribute) is provided in Appendix A. The results for each inspector tose work was reinspected is provided by attribute in Appendix B.

s V-4 ,

Table V-2 Results of the Reinspection Program Blount Brothers Inspection Status of Type Reinspection Condition Objective Complete All 8 inspectors acceptable at end of first 3-month period.

Subjective NA All inspections included in objective reinspection population.

Table V-3 Results of the Reinspection Program Johnson Controls

-Inspection Status of Type Reinspection Condition Objective Complete Four inspectors acceptable at cad of first 3-month period. One inspector did not have minimum quantity in first 3-months or in total inspections and failed to meet the acceptance cri-teria. All of his accessible and recreatable work was reinspected.

Subjective Complete All 4 inspectors acceptable at end of first 3-month period.

V-5 Table V-4 Results of the Reinspection Program Hunter inspection Status of Type Reinspection Condition Objective Complete Nineteen inspectors acceptable at end of first 3-month period.

One inspector passed who did not have minimum quantity in first 3 months or in total inspections. All of his accessible and racreatable work was reinspected. ,

Subjective Complete Sixteen inspectors acceptable at end of first 3-month period.

One inspector failed to meet the acceptance criterion at end of first 3 month period and had no more rein-spectable work. An additional inspec-tor was substituted. Reinspectior, of the additional inspector's first 3-months work was reinspected and was found acceptable. He is included in the above sixteen inspectors.

Table V-5 itesults of the Reinspection Program NISCo Inspection Status of Type Reinspection Condition Objective Complete All 4 inspectors acceptable at end of first 3-month period.

Subjective . Complete All 4 inspectors acceptable at end of first 3-month period.

l V-6 Table V-6 Results of the Reins *ction Program Hatfield E,ectric Inspection Status of Type Reinspegtion Condition Objective Complete All seventeen ir.cpectors acceptable at end of first 3-month period.

Sulsjective Complete Seven inspector:, acceptable at end of first 3-month period.

One inspector failed to meet the ac-ceptance criterion at end of first 3-month period and had no more rein-spectable work. An additionalinspec-tcr was substituted. The additional inspector's first 3 months work was reinspected and his work was found acceptable. He is included in the above seven inspectors.

Table V-7 Results of the Reinspection Program Powers-Azco-Pope Inspection Status of Type Reinspection Condition Objective Complete Eleven inspectors acceptable at end of first 3-month period.

One inspector passed who did not have minimum quantity in first 3 months or

-in total inspecticns. All of his work was reinspected. One inspector did not have minimum quantity in first 3 months or in total inspections and failed to achieve the necessary threcnold. All of his accessibie and recreatable work was reinspected.

)

V-7 Table V-7 (Cont.) -

Inspection Status of Type Reinspection Condition Two inspectors failed to meet the acceptance criterion at end of first 3-month period and had no more rein-spectable work. All of their acces-sible and recreatable work was reinspected.

Two insp:ctors acceptable at end of second 3-iionth period.

Two inspectors unacceptable at end of period. All of their accessibl2 and recreatable work was reinspected.

Subjective Complete Eight inspectors acceptable at end of period.

One inspector acceptable at end of second 3-month period.

One inspector passed who did not have minimum quantity in first 3 months or in total inspections. All of his work was reinspected. One inspector did not have minimum quantity in first 3 months er in total inspections and failed to meet the acceptance crite-rion. All of his accessible and recreatable work was reinspected.

One inspector failed to meet the acceptance criterion at end of first 3-month period and had no more rein-spectable work. All of his accessible and recreatable work was reinspected.

7 inspectors unacceptable at end of period. All of their accessible and recreatable work was reinspected.

s Note for Table V-7:

The results for Powers-Azco-Pope are presented for the first 3-month and second 3-month periods, respectively. The interim report dated January 12, l934, presented these results on a cumulative 6-month basis only.

V-8 Table V-8 -

Results of the Reinspection Program Pittsburgh Testing Inspection Status of Type Reinspection Condition Objective Complete Nine inspectors acceptable at and of first 3-month period.

Subjective Complete Seven inspectore acceptable at end of first 3-month period.

Three inspectors passed who did not have minimum quantity in first 3 months or in total inspections. All of their accessible and recreatable work was reinspected.

One inspector accqtable at end of second 3-month period.

Two inspectors failed to meet the acceptance criterion at end of fitst 3-month period and had no more rein-spectable work. Additional inspectors were substituted as described below.

One inspector unacceptable at end of both first and second 3-month periods.

All of his work was reinspected. Be-cause this inspector's work was not found satisfactory e.f ter reinspection of his second 3-month period, the pro-gram required expansion of the sample by six inspectors. This resulted in the inclusion of all inspectors qualified to perform visual weld inspection. Only fote of the remaining population had accessible work, so the first 3-months work of all four was reinspected. The inspectors added to the reinspection sample were acceptable at the end of first 3-month period, and are included in the seven above.

i I

y , , , , . . , . . . . .. . .

-t V-9 Table V-9 Results of the Reinspection Program Peabody Testing Inspection Status of Type __

Reinspection Condition Objective NA All reinspection population classified as subjective.

e Subjective Complete Two inspectors passed who did not have minimum quantity in .first 3 months or in total inspections. All of their accessible and recreatable work was reinspected.

One inspector did not have minimum quantity in first 3 months or in total inspections and failed to achieve the necessary threshold. All of his accessible and recreatable work was reinspected.

Three inspectors failed to meet the acceptance criterion at end of first 3-month period and had no more reinspectable work. All of their accessible and recreatable work was reinspected.

Very few of the Peabody Testing QC inspectors had reinspectable work. All of

- the reinspections were visual inspections of welding. Because the quantities -

were small, the work of all six (100%) reinspectable inspectors was reviewed in

- the initial sample.

_ _ _ _ _ .___________________m______.,______,,_.____m_________ -

l EXHIBIT V-1 Page 1 of 2 EXHIBIT V-1 REINSPECTION PROGRAM RESULTS OBJECTIVE INSPECTIONS Table VE-1, which is a summary of the information contained in Tables V-2 to V-9, provides the Program results for inspectors performing objective inspections. It should be noted that 82 of the 110 inspectors performed inspections which wera classified as objective.

Table V"_-1 Program Result: Hor Inspectors Porforming Objective Inspections QC Inspectors Passing QC Total No. of Acceptance Criteria (1) Inspectors Inspectors At At Did Not Pats Qualification Reinspected for 3 Mo. 6 Mo. Total Threshold Indeterminate (2) Objective Inspections Blount Brothers 8 -

8 -- -

8 Johnson Controls 4 --

4 -

1 5 Hunter 20 --

20 -- -

20 NISCo 4 --

4 -- --

4 Hatfield Electric 17 --

17 - --

17 Powers-Azco-Pope 12 2 14 2 3 19 Pittsburgh Testing 9 --

9 -- --

9 Pebody Testing N/A N/A N/A N/A N/A 0 TOTAL 74 2 76 2 4 82 Notes for Table VE-1:

(1) The criterion established by the Program for demonstrating acceptaoility was 95% agreement rate in repeat instections for objective inspections.

(?) If an inspector had no insrections beyond 3 months and did not meet the Program acceptance criteria, a substitution was made as described in Chapter III, Section B.2. The qualifications of the original inspector in such a case were considered indeterminate.

(3) The percent passing rate shown for e contractor in Table V-1 was calculated by dividing the number of QC inspectors passing the acceptance criteria by the

EXHIBIT V-1 Page ? ;f 2 -

total number of total QC inspectors minus those whose qualification was indeterminate. For example, the porcent passing for Powers-Azco-Pope for objective insnections shown as 88% in Table V-1 was taken as 14 divided by 16 (i.e.,19-3).

g 4

O 4'

s 89

/

o f

EXHIBIT V-2 Page 1 of 2 EXHIBIT V-?

REINSPECTION PROGRAM RESULTS SUBJECTIVE INSPECTIONS Table VE-2, which is also a summary of the information contained in Tables V-2 to V-9, provides the Program results for inspectors performing subjective inspections. It should be noted that 72 of the 110 selected inspectors performed inspections which were classified as subjective.

Table VLi-2 Program Results for Inspectors Performing Subjective Inspections QC Inspectors Passing QC Total No. of Accepterze Criteria (1) Inspectors Inspector.

At At Did Not Pass Qualification Reinspected for 3 Mo. 6 Mo. Total Threshold Indeterminate (2) Subjective Inspections

~ Blount Brothers - -- -- -- -- -

Johnson Controls 4 --

4 -- -- 4 Hunter 16 -

16 --

1 17 NISCo 4 -

4 - --

4 H tfield Electric / --

7 --

1 8 Pc,wers-Azco-Pope 9 1 10 7 2 !9 Pittsburgh Testing 10 1 11 1 2 14 Prabody Testing 2 --

2 - 4 6 TOTAL 32 2 54 8 10 72 Notes For Table VE-2:

(1) The criterion established by the Program for demonstrating acceptability was 90% agreement rate in repeat inspections for subjective inspections.

(?) If an inspector had no inspections beyond 3 months and did not meet the

~

Program acceptance criteria, a substitution was made as described ir. Chapter 111, Section B.2. The qualifications of the original inspector in such a case were considered indeterminate.

(3) The percent passing rate shown for a contractor in Table V-1 was calculated by j dividing the total number of QC inspectors passing the acceptance criterion by 4

i

,e hd

~f EXHIBIT V-2 Page 2 of 2

' the tout number of- QC inspectors minus those whose qualification was indeterminate. For example, the percent passing fer Powers-Azco-Pope for

- subjective inspections shovin as 59% in Table V-1 was taken as 10 divided by 17

. (i.e.,19-2). I i

e s

a J

V d -

l

- f .: +

g s'.

f.

VI-l VI. ENGINEERING EVALUATION OF DISCREPANCIES -

A. OVERALL DESCRIPTION The Reinspection Program focused on an assessment of individual inspector 4-qualifications and contractor certification Fractices; however, a significant amount of work quality data was accumulated. In order to apply the data from the Program to an assessment of the quality of construction work at Byron, an evaluation was made to determine whether or not the observed discrepancies had any design significance.

s The engineering evaluation of discrepancies is divided into subjective (visual weld quality inspection) and objective discrepancy evaluations. All observed discrepancies, both subjective and objective, have been evaluated either by

. detailed calculations, engineering judgments, or engineering comparisons of the discrepancy to the existing design margin.

The result of these engineering evaluations for all observed discrepancies is that none have design significance. In other words, design margins have been maintained, and code design criteria and equipment functional requirements have been met. Based on these evaluations, we conclude that the Byron construction activity prior to September 1982 properly implemented the design and, therefore, meets applicable regulatory requirements.

B.- CATEGORIZATION OF DISCREPANCY EVAI.UATIOt;S

l. _ Categorization An engineering evaluation has been performed for each observed discrep-ancy. The evaluation methods used are divided into the following three categories:

l Lj~

VI-2 Category X - Evaluation by comparison with current design parameters and tolerances, and including the elimination of cosmetic items and resolution of minor documentation errors.

Category Y - Evaluation by engineering judgment based on a comparison of the discrepancy with design margins.

Category Z - Evaluation by engineering calculations.

2. Subjective Evaluations

% AWS Wcld Evaluation The three evaluation methods discussed above can be related to the acceptance criteria for AWS visual weld inspection. The AWS acceptance criteria consist of inspecting welds for arc strike, spatter, convexity, crater, incamplete fusion, overlap, porosity, undercut, underrun, and cracks. The presence of these weld inspection items are considered as weld discrepancies. These weld discreparcies vary in degree as to their effect on weld capacity.

The Category X evaluation method is based on' the fact that arc strike, spatter, and convexity do not reduce the weld capacity. Arc strikes and spatter are cosmetic indications that relate to appearance only. Convexity relates to weld metal on the face of a weld in excess of the weld metal necessary for the required weld size (see Exhibit C-2 for a detailed discussion).

The Category Y evaluation method considered crater, incomplete fusion, overlap, porosity, undercut, and underrun. Each of these elements are treated the same in the evaluation of weld capacity.

' That is, portions of the weld with these discrepancies are considered ineffective and weld capacity is based on a reduced weld length. This

p VI-3 is a conservative approach because the portions of the weld with these types of discrepancies do exhibit significant load carrying capability.

Engineering judgment is used to evaluate a weld discrepancy in this class. That judgment considers the available design margin in the weld and the reduced weld length after accounting for the assumed ineffective portien resulting from the discrepancy.

The Category Z evaluation method is also based on reducing the weld length for the presence of weld discrepancies as given for Category Y. Cracks were also included in this category. There were only two cracks among all observed weld discrepancies in the Program. The total weld length with a crack was considered in the evaluation to be ineffective. This method of evaluation of weld discrepancies required engineering calculations because the magnitude end types of dixrepancies could not be judged as adequate without a detailed calculation.

A detailed discussion of the engineering evaluation of AWS weld discrepancies is contained in Appendix C, Exhibit C-2.

b. ASME and ANSI B31.1 Weld Evaluation An engineering evaluation was performed to determine the design significance of ASME and ANSI B31.1 code welding discrepancies. The visual we!d reinspection discrepancies were first evaluated for compliance v'ith the code design criteria. All discrepancies were determined to meet ASME and ANSI B31.1 code design criteria. This evaluation was divided into three basic categories as previously discussed.

Category X involved discrepancies that are within current design or installation parameters and tolerances. Discrepancies in Category X are not considered valid discrepancies.

l l

VI-4 Category Y involved discrepancies which were determined to be minor in nature and could be accepted based on engineering judgment.

Category Z involved discrepancies which required a weld assessment calculation to be performed. Weld assessment calculations were performed assuming no weld material existed where a relevant discrepancy was located. In all cases, it was determined the weld joint with the discrepancies met or exceeded code design criteria.

3. Objective Evaluations The objective evaluations followed the same format as the evaluations already outlined for subjective discrepancies. The evaluation of discrepancies in Category X is based on a comparison of the installation to the current design requirements. Actual component locations were compared to corresponding design locations with applicable installation tolerances; documentation errors were compared to the actual installation; and actual component aimensions were compared to the corresponding design dimensions with applicable tolerances applied.

Discrepancies in Category Y were evaluated by judgment and were compared to current design analysis, calculations, or component ratings.

Discrepancies were judged to be acceptable when the component design function was unaffected by the discrepancy. .

Evaluation of discrepancies in Category Z that required engineering calculations resulted in the revision of existing design documents to incorporate the discrepant condition or in the preparation of a specific calculation to address the impact of the discrepancy on the design.

i

[ C E

E g VI-5 E C. RESULTS OF ENGINEERING EVALUATION _

f

1. Subjective Discrepancy Evaluation

~~

The resiilts of the subjective discrepancy evaluations for each contractor f are summarized in Table V!-l.

=

^

Table VI-l

= Summary of Suojective Discrepancy Evaluation Red ._

m L

w

= No.cf Category X Category Y Category Z No. uth

[ Discrepancy No. Within No. Acceptable No. Acceptable Design significance

{ Contractor Evaluations Parameters by Judgment by Calculation ,

I t- Blount Brothers' 0 N/A N/A N/A N/A L Johnson Controls 65 15 12 38 0 5 Hunter 109 25 23 61 0 g

NISCo 0 N/A N/A N/A N/A Hatfield b lectric 2,117 11 2,064 42 0

{ ,

A Powers-Azco-Pope 914 201 77 636 v Pittsburgh Testing 905 l 887 17 0 f

Peabody Testing 22 0 11 11 0 q ..

e

l

~

g TOTAL 4,132 253 3,074 305 0 W

3  ! <

j ' Inspection of Blount Brothers was performed by Pittsburgh Testing. Inspection results are reported "nder Pittsburgh Testing.

=

'l j Table VI-l shows that 253 of the discrepancies (6%) identified in the Reinspection Program are not " valid" discrepancies and represent work J that is within current design parameters. The Category X discrepancies 3 _.[

result primarily from design parameters that have been expanded since the yg time of the original inspection. Therefore, the observed discrepancies are e-

_ actually within current design limi s.

E m .

5

[" .

VI-6 The Category Y evaluation covered 3,074 of :he weld discrepancies (74%) -

wherein weld capacity was reduced by approximately 10% af ter accounting for the weld discrepancy. h all cases the design margin remained within the specified design limits.

The Category Z evaluation covered 805 (.I the weld discrepancies (19%).

The reduction in weld capacity varied after accounting for the weld discrepancy. However, in all cases the design margin remained within the specified design limits.

The engineering evaluation of subjective discrepancies has shown that none have design significance.

A detailed presentation of subjective discrepancy evaluation is contained in Appendix C.

2. Objective Discrepancy Evaluation The results of the objective discrepancy evaluations for each contractor i are summarized in Table VI-2. _.

Y l N; l, c, ,'T y: .' sct 2%W [AY

' _- h ;'h'. .y.-

fQ ,,

,. .; g.

y_{.<~u

, .1 "*; -f , :

f$ . .-l 9, nq.

. n.

t*/i ?- . ' '.Pl T, ' k '7 .. h

.+.e

.. . ) 5. u .'

T< .

h i(E-(

m . ,.. a yy.

3 . . y.-

VI-7 Table VI-2 Summary of Objective Discrepancy Evaluation Results No.of Category X Category Y Categocy Z No. with Discrepancy No. Within No. Acceptable No. Acceptable Design Contrac tor Evaluations Parameters by Judgment by Calculation Significance Blount Brothers 28 10 8 10 0 Johnson Controis 47 l5 19 13 0 Hunter 684 614 52 IS 0 NISCo [2 0 12 0 0 Hatfield Electric 1,675 1,243 74 358 0 Powers-Azco-Pope 295 232 5 58 0 Pittsburgh Testing 66 1 9 56 0 Peabody Testing

• N/A N/A N/A N/A N/A _

TOTALS 2,807** 2,i i 5 179 51: 0

• Reinspection of Peabody Te . ting involved only subjective inspections.

• In some cases, more than one discrepancy was associated with a component. This results in the number of discrepancy evaluattons (2,8;7) being di'ferent than the number of ob .erved discrepancies (3,247) shoxn in Appendix D, Table D-1.

Table VI-2 shows that 75% cf the observed objective discrepancies ,

identified in the Reinspection Program are not considered valid discrepancies and represent work that is within current design parameters. The discrepancies in Category X that are not considered valid discrepancies result primarily from minor documentation errors that do not ;ji c7 affect the quality of the installed work; tolerances that have been N 9~

expanded since the time of the original inspection so that the observed discrepancy is actually within c Jrrent design limits; and as-built reinspection to' erances that were .ntentionally chosen to be more x

restrictive than actual installation tolerances in order to assure that all actual discreparcies were captured by the Reinspection Program.

VI-8 The discrepancies in Categories Y and Z represent either an installation that differs from the original design bat also meets the design requirements (e.g., the actual location of a conduit support is different trcm the support location shown on the installe. tion drawing but the actual

} location also meets the design requirements) or design features that are not required for normal or emergency operation (e.g., the diesel generator cylinder temperature indication at the local control panel which does not serve either an operating or alarm fonction but is connected because it is part of the package prv ~ded by the sendor).

The discrepancy evaluations of these types were typically uncomplicated in

~

that the criteria against which significance was measured were clear and the comparison to those criteria was straightforward. This is demon-strated through the examples presented in the notes accompanying the tabler in Appendix D, Exhibit D-1. The engineering evaluation of cbjective discrepancies has shown that the majority of the observed discrepancies are not valid discrepancies.

Only one validated discrepancy (see Note 4, Table DE-5), which involved a cable termination error af fecting an auxiliary building exhaust damper, presented any question as to its significance. However, as explained in Note 4, the discrepancy was determined to have had no design significance because the function of the dam?er was not precluded. All other validated objective discrepancies were also shown to have no design significance.

Therefore, all of the observed discrepancies are not significant to the overall design of the plant.

A!! ASME and ANSI B31.1 objective discrepancies were evaluated in detail and determined to have no significance to the overall design of the

]( affected component. All ASME and ANSI B3l.1 discrepancies were also (fhk

.e.c 4 reviewed to their applicable code design criteria and determined to be

.3 3 acceptable.

fk '

A detailed presentation of objective discrepancy evaluation is contained in Qp/ Appendix D.

VI-9

. D. CONCLUSION The engineering evaluation of the data obtained from the Reinspection Program a!!ows us to conclude that:

o None of the ob' served discrepancies had design significance and, in all cases, the design margin is within specified design limits.

o. The work performed by all contractors is of good quality.

Y '

u

l VII-1 VII. QUALITY OF THE WORK A. INTRODUCTION The work at tha Byron plant is of high quality and meets the plant design requirements. This conclusion is supported by both the general effectiveness of the comprehensive quality program implemented by Commonwealth Edison at u Byron and the results of the Reinspection Program reported and evaluated in this report. The three main points of the analysis leading to this conclusion are:

1. The approach to quality adopted by Commonwealth Edison's management provides many independent layers of inspection and review of field installations. This approaches ensures that a high quality of work is consistently maintained during construction and that work items with discrepancies from design requirements are identified and corrective action is taken. This requirement is applicable to the work of all contractors at Byron.

2. The vast majority of inspectors whose work was reinspected under the Byron Reinspection Program passed the Program acceptance criteria. On this basis. the effectiveness of the QC inspector certification programs of six centractors was validated. These six contractors are responsible for 88% of the total site work. The effectiveness of those certification programs ensures that all work performed by those contractors was adequately inspected, from which it can be inferred that the contractor's construction work is of good quality.

The certification programs for the remaining two contractors reviewed under the Reinspection Program could not be validated. However, ti.is fact alone doos not affect our conclusion relative to the quality of work.

Powers-Azco-Pope, whose program prior to Septembe- 1982 was not

VII-2 uniformly effective, performed most of his tork af ter September 1982.

The vast majority of the work performed prior to September 1982 was reinspectable, and 100% of that reinspectable work for all inspectors failing to pass the Program acceptance criteria was reinspected. Because no discrepancy of design significance was identified, the quality of work was shown to be good. Peabody Testing had too little reinspectable work from which conclusions on certification program effectiveness could be drawn. However, this contractor had a very limited scope of work (0.2% of j the site total), most of which was overin=pection of other contractors or inspections overseen by Commonwealth Edison personnel. Inasmuch as 100% of this contractor's reinspectable work was reinspected and no discrepancy with ' design significance was found, the good quality of this contractor's work can be inferred.

3. The Reinspection Program resulted in a total of 156,926 objective inspections and a total of 44,980 subjective inspections being repeated by cut <ently qualified inspectors. These reinspections ranged over a wide variety of plant work items. Engineering evaluation of all observed discrepancies showed that none had design significance. This data supports the inference that the quality of work for all eight contractors in the Reinspection Program was good.

The remainder of this chapter elaborates upon and substantiates these points.

1 B.. MANAGEMENT APPROACH TO ENSURE QUALITY OF CONSTRUCTION  !

Commonwealth Edison has implemented a' comprehensive quality program to assure that the Byron Station is constructed properly and is of high quality.

The program begins prior to award of contracts by requiring that procurement documents include commitments to specific quality requirements and it con-

. tinues throughout the construction phase. The essence of the approach is the provision of many independent layers of inspection and review of field instal-

.lations to assure compliance with requirements and, thereby, to ensure quality construction.

w 1

VII-3 s

The quality of the construction at the Byron Station is reviewed and inspected at varying levels' and in varying detail. The initial inspection of completed

' work is accomplished by each site contractor's QC inspection personnel who, in turn, are reviewed, in part, under the contractor's internal audit program. It should be noted that certain types of destructive and nondstructive testing are <

performed -by the Byron independent ~ testing : agency (Pittsburgh Testing

~ Laboratory) which supplements the contractor's QC activities.

' ~

'Under the direction of Commonwealth Edison Site QA, the independent testing agency' also performs independent "overinspections" of randomly selected portions of the previously accepted work. This additional level of inspection, although not required by specification ~or. regulation, enables Commonwealth Edison to view the quality of. the field work and inspection activities of the

contractors.

This' indepe.ndent testing agency also conducts a separate--

~ . Independent inspection :of . installed work bounded within specific areas or

units.' ;This unit concept inspection (UCI) is an in-depth crerinspection of the electrical,' mechanical, HVAC ~ and structural installations compared ' to the

. latest _ design ocuments. -

Another level of inspcction was performed by a few site contractors who, for

~

various reasons, found it necessary to reinspect portions of previously inspected

, work. 3 Additionally, the Commonwealth Edison Project Construction Depart-1 ment performs scheduled surveillances of site contractor field activities. This

.is yet another level lof control over the quality of workmanship and the site

. contractor's ability to perform quality related work.

3 The overall quality related activities of each site contractor are completely reviewed,'at a minimumiannually by Commonwealth Edison Site QA auditing land Tsurve.illance ' activities : which are . supplemented by - the Commonwealth Edison General . Office QA activities. . Finally, a multitude of ~ independent evaluatio'ns L have been ' conducted = of various site contractors by such

-Lorganizations' :as -INPO; ' ASME; Energy Incorporated, as part of the Commonwealth Edison management audit; and the NRC.

N

^- *

\

'{ ;,

__ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ______mm____.__._ _ _ _ __ - - _ . _ ___-

l VII-4 The aforementioned multi-levels of inspections and reviews have, in some instances, resulted in updating and expanding of the contractor QC program procedures. The site structural and mechanical contractors have very adequately adjusted their QC program, as has the site's independent testing agency. Site QA "stop work" actions and audit requirements have compelled the site electrical contractor to periodically revise its QC Inspection program.

Additionally, a site QA stop work action imposed by Commonwealth Edison QA on the site HVAC contractor, in conjunction with a complete review and re-evaluation of this contractor's QC inspection program, as well as a reinspection of all safety related work performed by this contractor, provides reasonable assurance that the HVAC system is being installed properly.

Based upon the multi-levels of inspections, overinspections, audits, surveillances, and evaluations, Commonwealth Edison is confident that site contractors have properly executed their work in accordance with specifica-tion, code, and design requirements. The contractors have shown the ability to provide corrective actions for identified discrepancies and to maintain their QA program implementation at an acceptable quality and performance level.

An expclided discussion cf this quality program is provided in Appendix E.

C. WORK QUALDY BASED ON INSPECTOR QUALIFICATIONS The results of the Reinspection Program verify the effectiveness of the QC inspector certification programs in existence prior to September 1982 for six of

. eight contractors reviewed, accounting for approximately 88% of the total work at Byron. This gives aaditional confidence that the work of these six contractors at Byron meets a 1.igh standard of quality. This conclusion is valid for both the inspectors and work items specifically reinspected, as well as the remainder of the inspector and work item populations. This generalization can be made for the folicwing reasons:

i 6-t

-- --__mm___ ____.-- _mm . - _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _

VII 5

?

e lhe sampling of the inspector population for all eight contractors was adequate and, in many cases, more exten ive than required by common statistical standards. the only biases introduced in the Reinspection Program were more conservative than generally present in the population

-(i.e., the first 3 months of each inspector's work was reinspected because deficient work of any inexperienced or ill-trained inspector is most likely to appear during the early months of his job performance, before time permits on-the-job experience to correct his original deficiencies).

e To a great extent, the performance of the inspectors reflects the adequacy of the certification programs. As all inspectors of a contrsetor were certiiled using :he same program, the inspectors from any giver' contractor form a homogenous group with similar qualifications and thus similar performance. - Thus, results from an adequate sample of inspectors can be used to draw inferences for the inspector population in general.

Successful completion of the Reinspection Prograrr. has validated the

~

-e-adequacy of six of the eight contractors' QC inspector certification and training programs in existence prior to September 1987 The demonstrated adequacy of the certification programs for these six con:ractors assures the quality of all work whether specifically reinspected or not.

The following paragraphs address these points in more detail:

1 Sampling Adequacy Sampling inspection for making inferences about a larger population is a valid approach provided the larger population is homogenous, i.e., any one

_ item can be expected to be similar to any other, with variation:; occurring in a random, unsystematic way; the sample is of sufficient size to capture both the " average" behavior of the larger population as well as a full

- spectrum of variations about that average. Any sampling biases should be

. conservative relative to inferences made for the larger population.

2 - - _ - _ - - - _ _ - - - - - - _ - - - _ - - - - - - - - - - _ . _ . _ _ . _ _ _ _ _ - - _ _ _ . - - - . . _ - - _

VII-6 If the inspectors for any given contractor are considered the " larger population," then the homogeneity of this population can be argued from the fact that all individual inspectors for that contractor were subjected to a common certification program. These individuals performed their inspections from a similar base of technique, knowledge, and capacity.

Variations in inspection performance are thus exoected to occur randomly about this base because they are the result of human errors.

.jy W. The Reinspection Program sampling scheme as described in Chapter Ill was

_i

._,in[g.g

+g f not specifically designed on a formal statistical basis. Rather, it was a 7M result of engineering judgment augmented by NRC-directed additions of II individuals. However, the plan generally adheres to the fundamental

%d'

b. g stati:tical principles of adequate sample size and representation of the

.O.h- larger population. The selection of every fifth individual on a list

.m.

chrunologically ordered by initial certification date assures a wide ran;ing

@p;f. .e .,

{..g ,.

representation of inspection activities over the time scale of the 4,'.g contractor's participation in the plant construction, while the requirement

.i. f. , ,%, to reinspect each inspector's first three months' work introduces a f f, t systematic conservative bias because deficient work of any inexperienced or til trained inspector is most likely to appear during the early months of his job performance before time permits on-the-job experience to correct his original deficiencies.

The adequacy ot the sample sites can be judged by comparisoa to those specified by Military Standard 105D*. Military Standard 105D is a standard ~ ANSI document containing sampling plans for performing inspection by attributes. The standard specifies sample size as a function of population size. Table VII-l lists the total population of inspectors, number of inspectors who were reinspected, and the number of inspectors

• Military Standard 105D, " Sampling Procedure and Tables for inspection by Attributes," Washington, D.C.; U.S. Government Printing Office,1963. Also appears as ANSI /ASQC Zl.4-1931.

k-r:

VII-7 l

required to be sampled for each contractor based on Military Standard  !

105D single sampling plan and a normal inspection level.

Table VII-l Sample Sizes Used in Reinspection Program vs. Those Required in Military Standard 105D No. of No. of Inspectors Total Population Inspectors to be Reinspected per Contractor of Inspectors Reinspected Military Standard 105D Blount Brothers 28 8 8

• Johnson Controls 7 5 2

. Hunter - 84 22 13

. NISCo 8 4 2 Hatfield 86 23 13

• Powers-Azco-Pope 21 19 5 Pittsburgh Testing 85 23 13

• Peabody Testing 37 6 8

• All of the inspector populatiors was reviewed for pessible reinspec_ tion.

There were no reinspectable items for those inspectors not included.

L The above table shows that the number of inspectors reinspected as part of

-the_ Reinspection Program is equal to or larger than that required in the Military - Standard .'105D, except for Peabody Testing. The smaller inspector sample size for Peabody is due to. the inaccessible and not recreatable nature of the work they inspected; however,100% of Peabody's accessible and recreatable work was reinspected. Therefore, the Reinspection Program _ sample sizes were, to the extent practical, in accordance with or. exceeded that required in Military Standard 105D.

Based on the above, it is concludeo that sampling was appropriate and the inspector sample sizes were large enough so that the results of the Reinspection Program are representative of each contractor's total inspector population.

=--

VII-3

2. Validation of Inspectors Certified Prior to September 1982 The basis of contractor QC inspector certification programs in effect at various times since construction began at Byron is described in Chapter II.

It should be noted that all site contractors developed ccrtification pro-grams that qualified their QC inspectors in accordance with the ANSI standard. The Reinspection Program was implemented to verify the ef-

' fectiveness of former certificaticn practices and inspector qualification by reexamining inspections performed by the QC inspector, to alleviate NRC concerns with respect to QC inspector certification prior to September 1982. The successful completion of the Reinspection Program validates the qmlifications of the inspectors certified prior to September 1982 for the six of the eight contractors who will perform approximately 88% of the total site work at Byron. Qualifications of the inspectors certified after September 1982 have been accepted and meet current industry and NRC requirements. This leads us ta conclude that site work of these contactors at Byron is being (and has been) inspected by qualified inspectors which ensures quality construction.

In the above discussion, we have demonstrated that the sampling methods applied during the Reinspection Program form an adequate basis for drawing inferential conclusions on the entir population of inspectors. Thus the

. generally good agreement between original inspections and reinspections verifies the adequacy and effectiveness of inspector certification programs in use prior to September 1982 for the six contractors. Moreover, because the contractors' QC programs were applied to all work performed, the larger infer-ence can be drawn that the quality of work is good for these six contractors.

It should be noted that for the .two remaining contractors, accounting for approximately 5% of the total work, the Reinspection Program did not validate the adequacy of their pre-September 1982 certification program. However, 100% of the accessible and recreatable work performed prior to September 1982 was reinspected for all inspectors not shown to meet the Program

C VII-9 acceptance criteria, and no observed ciiscrept .termined to have design significance. The quality of their wc:k . rom the reliability calculation presented in Section D below.

D. INFERENCE OF WORK QUALITY FROM THE REIMSPECTION PROGRAM in this section, the detailed reinspection data given in Appendix B and the engineering evaluation data given in Appendixes C and D are combined to obtain reliability estimates for each of the eight contractors' work. These reliabilities are intended to address the quality levels for work th,t was not reinspected.

1. Applicability of Reinspection Program Data to Plant Quality Inferences The data from the Reinspection Program provides a reasonable basis for estimating plant quality when samples are adequate in size and scope, and the entire Reinspection Program is of sufficient technical scope.

The question of sample size and adequacy of representation for sampled inspectors in the overall inspector population is discussed in section C above. The sampling of inspector's work was far from random in that it concentrates entirely on each inspector's first three months of work. More generally, it should be noted that the calculation of the reliabilities presented below is based on a formal statistical methodology which is not entirely consistent with the non-statistical (biased) way the samples were selected. However, the biases introduced by the sample selection procedure are conservative and the results of the formal calculation are thus justified in that they are underestimates of the true reliabilities.

The sufficiency of technical scope is an engineering judgment that relates to the relevance of inspected attributes to work quality. Based on the QC procedures and their associated checklists used in the Reinspection Prcgram, it is our conclusion that the Reinspection Program did have sufficient technical scope to yield information on construction quality.

1 L

VII 10 2 Evaluation of Reliabilities For objective inspections, a total of 156,926 items were reinspected and -

3,247 discrepancies were noted (Appendix D). For the subjective in-spections, a total of 44,980 items were reinspected, and 4,132 discrep-ancies were noted (Appendix C). De discrepancies which had potential for I affecting the design were evaluated. His engineering evaluation showed that none 01 these discrepancies had design significance. Bis demonstrates the good quality of the work performed by the contractors reviewed at the 1)yron Station.

Tables VIIE 1 through V"lE 8 of Exhibit VII l list the number of inspected items, the number of discrepancies of design significance, and calculated reliabilities for each of the eight' contractors. Objective and subjective attributes are listed separately.

De reliability for each attribute'can be defined as the proportion of work ite'ms in the total population of work for that attribute which have no discrepancies with design significance. Statistical estimates of the reliability can be made from inspections and engineering analyses of random samples from the population. The precision of these estimates, of course, increases with the sample size.

A generally accepted statistical method for calculating such reliabilities is to compute reliabilities at 95% confidence level from the sampled data.

Such a reliability represents a conservative estimate of the true reliability.

It is conservative in the sense that there is a 95% chance that the true 1

reliability is greater than the estimate. In the case where no discrepant I items are abserved in a random sample from a large popuiation, the reliabil- l ity at 95% confidence level can be calculated from the formula *

• Reference 1: Miller, I., and Freund, J.E., " Probability and Sta'istics for Engineers,"

Prentice Hall, Inc.,1977, Chapter 9

9 VII-l1 R = 1 2M9" n 1:q. VII-I where R = Reliability at 95% confidence level n = number of inspections in the random sample In Exhibit VII-l and Tables VIIE-1 (Blount), Vile-2 (Jchnson Controls),

VIIE-3 (Hunter), VIIE-4 (NISCo), VIIE-5 (Hatfield), VIIE-6 (Powers-Azco-Pope), Vile-7 (Pittsburgh Testing), VIIE-8 (Peabody Testing), which follow, the reliabilities presented are based on Eq. VII-1. i.e.. they represent reliabilities at 95% confidence level based on samples which contained no discrepancies of design significance. It should be emphasized that, when a sample size is small, the true reliability is likely to ue much greater than indicated.

Tables Vile-1 through VIIE-8 show better than 95% reliability for 25 of the 30 attributes reinspected. In 21 of these cases, the calculated reliabilities are better than 99%. For four cases, the reliabilities are computed in the 96% to 99% range. For the remaining five cases, in these tables no reliability estimate at 95% confidence level is projected because Equation VII ! requires at least 60 observaticns to provide 95% reliability. For these five cases, the sample sizes were too small to obtain meaningful reliabilities based on Equation VII-1. This does not prevent us from concluding. on the basis of calculated reliabilities, that all contractors performed good work. This conclusion remains valid because all inpectors within a contractor organization were qualified under the same program and good reliability demonstrated in one objective attribute provides a valid basis for inferring the reliability in another objective attribute where s'.mpling was limited. In Table VIIE-5, five out of eight objective

VII-12 attributes, and in Table VIIE-6, three out of five attributes have better than 95% reliability, therefore it ic inferred that the reliabilities not listed in the tables would also be better than 95%. Note that for attributes where the number of items reinspected is large (>300), the computed reliabilities are better than 99%. This is indicative of good quality work.

The above discussion of reliabilities indicates that better than 95% ,

reliability is expected for the work of all eight contractors. This component level reliability is considered to be high enough to conclude that work quality is good.

E. CONCLUSIONS The evaluation of Commonwealth Edison management approach to ensure quality of construction and the successful completion of the Reinspection Program leads us to conclude that:

1. The good quahty of construction at Byron is ensured because of the comprehensive quality program implemented by Commonwealth Edison management. The many layers of inspections, overinspections, audits, surveillances, and evaluations implemented as part of this quality program further assure us that the results of the Reinspection Program are representative of the overall plant quality.

2. With limited exceptions, the Reinspection Program verified the effectiveness of QC inspector certification programs prior to September 1982. This ensures that work performed by the contractors whose programs were effective was adequately inspected, from which it can be inferred that the contractor's construction work is of good quality. The quality of work for the contractors whose QC inspector programs were not verified has been confirmed through additional reinspection and evaluaticn.

VII-13

3. The adequacy of construction quality for all eight contractors is supported by the Reinspection Program results and by inferences drawn from these results.

Exhibit Vil-1 Page 1of5 EXHIBIT Vil-1 CALCULATED RELIABILITIES FOR WORK BY CONTRACTOR Table Vile _1 Calculated Reliabilities for Work of Blount Brothers No. of Reliability No.of Discrepancies  % at 95%

Inspected with Design Confidence items Significance Level A. Objective Attributes 1,858 0 99.8

1. Weld detail 0 97.9

2. Concrete 145 0 99.0

3. Masonry 310

4. Concrete 0 96.1 expansion anchor 77 B. Subjective Attributes None Table VIIE-2 Calculated Reliabilities for Work of Johnson Controls No. of .

Reliability No. of Discrepancies  % at 95%

Inspected with Design Confidence items Significance Level A. Objective Attributes 7,812 0 >99.9

1. Mechanical B. Subjective Attributes 1,459 0 99.8

1. Visual welding

)

Exhibit VII-1 Page 2 of 5

)

Table VIIE-3 Calculated Reliabilities for Work of

. Hunter No.of Reliability No. of Discrepancies  % at 95%

Inspected with Design A. Objective Attributes Contidence Items Significance Level

1. Documentation 36,632 0 >99.9

2. Hardware 32,992 0 >99.9 B. Subjective Attributes

1. Visual weld 3,725 0 >99.9 Table VIIE-4 Calculated Reliabilities for Work of NISCo No.of Reliability No. of Discrepancies  % at 95%

inspected with Design Confidence A. Objective Attributes Items Significance Level

1. Mechanical 2,792 0 99.9 B. Subjectiva Attributes

1. Visual welding 229 0 98.6

s Exhibit VII-1 Page 3 of 5 Table VIIE-5 Calculated Reliabilities for Work of Hatfield No.of Reliability No. of Discrepancies  % at 95%

Inspected with Design Confidence items Significance Level A. Objective Attributes

1. Conduit 2,793 0 99.9

2. Terminations 7,784 0 >99.9

3. Equipment setting

• 0 0 -

4. A325 ';olting* 8 0 -

5. Equipment modification

• 27 0 ,

6. Conduits as-built 44,777 0 >99.9

7. Pan hangers 4,776 0 >99.9

8. Pan 80 0 96.3 B. Subjective Attributes

1. Visual welds 26,660 0 >99.9

• The samples related to Attributes 3, 4 and 5 are too small to permit meaningful reliability calculation based on Equation Vil-1. Further sampling as discussed in Appendix D, Exhibit D-1, is being conducted for Attributes 3,4 and 5. Note that all objective attribute inspections require similar inspector training. This contractor did exceptionally well on other objective attributes; thereitre, the effectiveness of his QC inspections are not in question. The additional work is being performed to complete the cata base. This additional data is not considered essential to our conclusions as to QC inspector qualification or the quality of the work done.

i Exhibit VII-l Page 4 02 5 r

Table Vile-6 Calculated Reliabilities for Work of Powers-Azco-Pope No.of Reliability No.of Discrepancies  % at 95%

Inspected- with Design Confidence items Significance Level A. Objective Attributes

1. Pipe material 2,286 0 99.9 verification

. 2. Hanger / mount material 5,389 0 >99.9 verification

3. Final hanger 312 0 99.0

4. Flexible hose

• 18 0 -

5. Pipe bend

• 42 0 -

B.. Subjective Attributes

1. Pipe weld visual 4,016 0 >99.9

2. Hanger weld visual 2,591 0 99.9

• The sample size is too small to compute meaningful reliability based on Equation Vil-1. However, it should be noted again that 100% of the work recreatable and accessible for reinspection done by inspectors whose work was judged ineffective -by the Reinspection Program has been - re -

inspected. For this contractor, 99% of the work was accessible for reinspection. For this reason, the likelihood of a valid discrepancy left undetected within this contracter's scope of work is extremely small; therefore, no additional sampling is required.

s k

Exhibit VII-1 f~ Page 5 of 5 e,

Table VIIE-7

-Calculated Reliabilitie.s for Work of Pittsburgh Testing No.of Reliablitty No. of Discrepancies  % at 95%

Inspected with Design Confidence items Sigrificance Level A. Objective Attributes

1. Concrete expansion anchor 6,016 0 >99.9 B. Subjective Attributes

1. Visual welding 6,137 0 >99.9 Table VIIE-3 Calculated Reliabilities for Work of Peabody No.of Reliabilty No.of Discrepancies  % at 95%

Inspected with Design Confidence A. Objective Attributes Items Significance Level None B. Subjective Attributes

1. Visual welding

• 163 0 98.2

• 100% of accessible and recreatable work was reinspected.

1

Appendixes Commonwealth Edison

9 .-

Report on the Byron "

QC Inspector Reinspection Program Docket Nos. 50-454 and 50-455 e

a February,1984 6 , , ,, .. . . . .

g;_ '

A-1 APPENDIX A REINSPECTION RESULTS, BY TYPE OF INSPECTION AND BY INSPECTION AREA OF CERTIFICATION fA tabulation of inspection results by inspection type (objective and subjective) and by

. inspection ~ area of certification (attribute) is presented for each contracter in .

Tables A-1.(Blount Brothers), A-2 (Johnson Controls), A-3 (Hunter), A-4 (NISCo), A-5 (Hatfield Electric), A-6 (Powers-Azco-Pope), A-7 (Pittsburgh Testing), and A-8 (Peabody Testing). Section A of each table presents the overall agreement rate at the Level It inspection step and third-party review step (for visual weld inspection only).

Section B 'of each table presents the number of inspector reinspected in each area

'(attribute) in the initial 5 month sample period and the associated overall agreement rate, as well as the number of inspectors reinspected in expansion sample periods, where applicable, and the associated overall agreement rate. The data in these tables provide a qualitative assessinent of the-work performed by each of the contractors

~ reviewed, as well as the individual work attributes reinspected.

a

l A-2 Table A-1 Reinspection Results ]

Blount Brothers A. Results by Inspection Type .

Reinspection Results Type Level II Reinspection Third-Party Review Subjective (2)- (1)

Objective 98.8 % (2)

B. Results by inspection Attribute Initial Sample Period Expansion Sample Period '

No. of No. of People Final % People Final %

Attribute Reinspected Acceptable Reinspected Acceptable

1. Weld detail ~

(Objective) 4 98.8 % (1) (1)

2. Concrete-(Objective) 2 100 % (1) (1)

3. Masonry (Objective) - 2 99.4 % (1) (1)

4. Concrete expansion 2 96.1 % (1) (1) atchor (Objective)

Y Notes for Table A-1:

(1) Not required (2) -Not applicable

A-3 Table A-2 Reinspection Results Johnson Controls A. Results by Inspection Type -

Reinspection Results (Acceptable / Total)

Type Level II Reinspection Third-Party Review Subjective 94.9 % 95.5% *

(l336/1459) (1394/1459)

Objective 99.4 % (2)

B. Results by Inspection Attribute Initial Sample Period Expansion Samje Period No.of No. of People Final % People Final %

Attribute Reinspected Acceptable Reinspected Acceptable

1. Visual (Subjective) 4 95.5% (2) (2)

2. Mechanical (Objective) 4 99.4 % (2) (2)

Notes for Table A-2:

• Results are cumulative. 73 observed discrepancies were reinspected by third-party inspectors.

(1) Not required (2) Not applicable

A-4 Table A-3 Reinspection Results Hunter A. Results by Inspection Type Reinspection Results (Acceptable / Total)

Typa Level 11 Reinspe-tion Third-Party Review Subjective 96.8 % 97.0 % -

(3604/3725) (3616/3725)

Objective 99.0 % (2)

B. Results by Inspection Attribute Initia! Sample Period Expansion Sample Period No.of No.of People Final % People Final %

Attribute Reinspected Acceptable Reinspec,tg Acceptable

1. Visual welding 17 97.0 % (1) (1)

(Subjective)

2. Documentation (Objective) 20 98.9 % (1) (1)

3. Hardware (Objective) 17 99.3% (1) (1)

Notes for Table A-3:

• Results are cumt.lative. 121 observed discrepancies were reinspected by third-party inspectors.

(1) Not regtilred (2) Not a,r plicable l 1

'6'&h O IMAGE EVALUATION I5

///// / g.f #,4

+%( *[%I/ TEST TARGET (MT-3) #4 a

l.0 Ei2 UE 5 9 EE i

I.I b' E5 1.8 i.25 __

i.4 i i.6 i

< 150mm >

< 6" >

$  % //

• <&'>//p ~

/A 4 ohff 4)g<gf i,

++

1. sk IMAGE EVALUATION p4

~ gpyg %e +%k@/ TEs11AnoEr (mT-a)- ((//

sgx p ' g, 3

%?'>+!'

f "%4$

I 1.0 d2 2 5 9 DE I.I  !. - EE j l.8 "1.25 1.4 I.6

< 150mm

• 4 6" #

I k' .

g a . y,y7/ .

I c h ,,,

h.'A

///4

%o e+e%x p

k.

t .

l 4

A-5 Table A-4 Reinspection Results NiSCo A. - Results by Inspection Type Reinspection Results Type Level 11 Reinspection Third-Party Review Subjective 100.0 % (1)

Objective 99.6 % (2)

B. Results by Inspection Attribute Initial Sample Period Expansion Sample Period No.of Nceof People Final

• People Final %

Attribute Reinspected Accer 3 Reinspected Acceptable

1. ' Visual welding 4 100.0 % (1) (1)

(Subjective)-

2. Mechanical (Objective) 4 99.6 % (1) (1)

Notes for Table A-4:

(1) Not required (2) Not applicable

. , . 7, - . , __g y

A-6 Table A-5 l Reinspection Results Hatfie:d Electri:

A. Results by Inspection Type l

Reinspetion Results (Acceptable / Total) l Type Level II Reinspection Third-Party Review Subjective 89.9% 92.1%

D3,978/26,660) (24,543/26,660)

Obj3ctive 96.5 % (2)

B Results by Inspection Attribute Initial Sample Period Expansion Sample Period No. of Peop;e Final % No. oiPeople Final %

Attribute Reinspected Acceptable Reinspected Acceptable

1. Visual weld (Subjective) 8 92.1% (1) (1) 2.- Conduit 6 97.6 % (1) (1)

3. Terminations (Objective) 5 99.9% . (1) (1)

4. Equipment setting 0 0% (1) (1)

(Objective)

5. A325 colting 1 100.0 % (1) (1)

(Objectivo)

6. Equipment modification 3 100.0 % (1) (1) t (Objective)

7. Conduit as-built 8 95.9 % (!) (1)

(Objective)

8. Cable Pan -

hangers 2 95.5 % (1) (1)

(Objectiw)

9. Cabla Pan 1 100.0 % (1) (1)

(Objective)

Notes for Table A-5:

• Results are cumulstive. 2,682 observed discrepancies were reinspected by

'hird-party inspectors.

(1) Not required (2) Not applicable

A-7 Table A-6 Reinspection Results Powers-Azco-Pope A. Results by Inspection Type Reinspection Results (Acceptable / Total) f- Type Level 11 Reinspection Third-Party Review Subjective 85.4 % 86.2%* (3)

(3,689/6,607) (5,693/6,607)

Objective 96.3% (3) (2)

B. Results by Inspection Attribute Initial Sample Perivd Expansion Sample Period No. of No.of People Final % People Final %

Attribute Reinspected Acceptable Reinspected Acceptable

1. Pipe material verification 19 ?7.8% 12 97.89; (Objective)

2. Pipe weld visual 19 93.7 % 12 92.9 %

(Subjective)

3. Hanger / mount material 19 96.5% 12 97.6 %

veriCcation (Objective)

4. Hanger weld visual 19 74.5 % 12 75.9%

( (Subjective)

5. Final hanger 10 67.1% 7 82.8 %

(Objective)

6. - Fiex hose 7 85.7% 4 100 %

(Objective)

7. Pipe bend 9 90.5 % 5 100 %

(Objective)

Notes for Tcble A-6

• Results are cumulative. 918 observed riiscrepancies were reinspected by third-party inspectces.

(1) Not required (2) Not applicable (3) For inspectors who did not meet the program acceptance criteria,100% of accessible, recreatable work was inspected.

J

f A-8 l

Table A-7 Reinspection Results Pittsburgh Testing A. Results by Inspection Type Reinspection Results (Acceptable / Total)

Type Level II Reinspection Third-Party Review Subjective 83.7 % 85.3%" (3)

(5,138/6,137) (5,232/6,137)

,a Objective 98.9 % (2)

B. Results by Inspection Attribute Initial Sample Period Expansion Sample Period No. of No.of People Final % People Final %

Attribute Reinspected Acceptable Reinspected Acceptable

1. Visual welding l4 86.0 2 77.0 (3)

(Subjective)

2. Concrete expansion 9 98.9 (i) (1) anchor (Objective) l Notes for Table A-7:

• Results are cumulative. 999 observed discrepancies were reinspected by third-party inspectors.

(1) Not required (2) Not applicable

(3) 100% of the work was ' inspected for the two inspectors in the expansion sample period. Discrepancies had no design significance.

\

l

'I' A-9 Table A-8 Reinspection Results  !

Peabody Testing A. Results by Inspection Type Reins *ction Results (Acceptable / Total)

Type Level 11 Reinspection Third-Party Review Suojective 71.8 % 75.5%*(2)

(117/163) (.823/163)

Objective (1) (1)

B. Results by inspection Attribute Initial Sample Period Expansion Sample Perici No. of No. of People Final % People Final %

Attribute Reinspected Acceptable Reinspected Acceptable

1. Visual welding 6 75.5% (2) (1) (1)

(Subjective)

Notes for Table A-8:

• Reelts are cumulative. 46 observed discrepancies were reinspected by third-party inspectors.

(1) Not applicable (2) All inspectors who had accessible, recreatable work were reinspected.

Y l

l B-1 APPENDIX B -

REINSPECTION RESULTS BY INSPECTOR A ^abulation of the detailed inspection results by inspector and by type of inspection i<. presented for each contractor in Tables B-1 (Blount), B-2 (Johnson Controls), B-3 (Hunter), B-4 (NISCo), B-5 (Hatfield), B-6 (Powers-Azco-Pope),

B-7 (Pittsburgh Testing), and B-8 (Peabody). This appandix shows actual quantities of acceptable and rejectable results by attribute. Expansion (exp.)

sample period results indicate results for the reinspection of an inspector's second 3 months of work. Results shown are complete and have had third-party review for the visual weld inspection (subjective) attributes.

B-2 Table B-1 -

Detailed Inspector Results Blount Brothers Attributes (Acceptable / Total)

Inspector No. i No.2 No.3 No.'4 i A 636/640 - - -

B -

70/70 - -

C 437/440 - - ,

D 427/443 - - -

'E - -

253/255 49/51 F -

75/75 - -

G 335/335 - -

25/26 H - -

55/55 -

TOTAL 1835/1858 145/145 308/310 74/77 Notes for Table B-1:

No expanded sampling was required.

Attribute 1 - Weld detail Attribute 2 - Concrete Attribute 3 - Masonry Attribute 4 - Concrete expansion anchors 1

l i

I i

3

-j+-

s-4

-Table B-2 Detailed Inspector Results Johnson _Controis i

Attributes Inspector No. I No. 2

'A -

25/28 B 222/230 3170/3178

-C - o28/660- 2761/2781 D 80/84 499/499 E 464/485 1310/1326 TOTAL' 1394/1459 7765/7812

. Notes for Table B-2:

No expanded sampling was requirec!.

- Attribute 1 - Visual welding

, Attribute 2 - Mechanical y g- - - , - - m

B-4 Table B-3 Detailed Inspector Results Hunter Attributes Inspector No.1 N_o.2 No.3 A 47/48 - -

B 14/14 134/138 -

C 34/34 1181/1186 -

D 33/33 101/102 -

E 283/301 2088/2144 61/64 F 208/214 40/41 258/265 G 116/129 161/161 21/21 H 49/55 19/19 12/12 I 315/319 47/47 129/133 3 -

2195/2269 7836/7893 K 334/344 280/284 186/190 L 273/273 366/366 204/206 1 M -

126/130 331/339 N -

289/294 903/921 O -

416/442 1246/1253

-P 249/263 8141/8214 925/935 Q 383/392 6315/6381 5355/5372-R 232/237 8503/8520 81/81 S 181/181 32?l331 949/952 T -

1789/1804 6248/6323 U 803/822 3671/3739 8004/8032 V 62/66 - -

TOTAL 3616/3725 36191/36632 32749/32992 Notes for Table B-3:

No expanded sampling was required; a substitution (V) was made for (H) because (H) failed the first 3-month period but had no further inspet.tions to reinspect.

Attribute 1 - Visual welding l Attribute 2 - Documentation Attribute 3 - Hardware 4

--- . - - . J

r~_ _

B-5 l

Table B-4 Detailed insgtor Results N1SCo 1

Attributes

- Inspector No.1 No.2 A 103/103 930/930 B 69/69 1567/1579 C 32/32 165/165 D. 23/25 113/118 TOTAL 229/229 2780/2792 Notes for Table B-4:

No expanded sampling was required.

(

Attribute 1 - Visual welding

-Attribute 2 - Mechanical e

3

, - -w -~ ,,-- ~ < ,-e-,,,,,, < . r , ~

w w~ ----ang,- , , - - - - . ~ ,

h i

B-6 Table B-5 Detailed InsWtor Results Hatfield Electric Attributy

. Inspector No.1 No.2' No.3 No.4 No.5 No.6 No.7 No.8 No.9

'A 328/865 - - - - - - - -

B .- - - - - -

4795/4974 - -

C 608/698 - - - - - - - -

D -

80/80 638/638 (1) 8/8 - - - -

E 10221/11312 187/188 45/48 - - - - - -

F -

178/179 72/72 - -

2/2 - - -

G- 724/771 386/401 544/546 - -

1/1 - - -

H - - - - - -

3985/4112 - -

I 4233/4489 - - - - - - - -

J' -

639/661 - - - - - - -

K -

1256/1284 - - - - - - -

.L- - - - - - - -

705/742 -

M - - - - - -

109'2/11457 - -

N 3309/3404 - - - - - - - -

O 31/51 - - - - - - - -

P - - - - - -

2001/2081 - -

Q - - - - - -

4818/5055 - -

-R - - - - - -

11734/12205 - -

S - - - - - -

2753/2879 - -

T - - - - - -

1917/2014 - -

U - -

6473/6480 (2) -

24/24(2) - - -

V - - - - - - -

3854/4034 80/80 W 4569/507C(3) - - - - - - - -

TOTAL 24543/26660 2726/275.4 7775/7734 -

Tlf 27/27 42955/44777 4559/4776 80/80 Notes for Table B-5:

No expanded sampling was required; a subr.titution (W) was made for (C) in

' Attribute No. I because (C) failed the first 3-month period but had no further

inspections to reinspect.

Attribute 1 - Visual weld Attribute 2 - Conduit Attribute 3 - Terminations Attribute 4 - Equipment setting Attribute 5 - A325 bolting Attribute 6 - Equipment modification Attribute 7 - Conduit as-built Attribute 8 - Pan hangers Attribute 9 - Pan (Notes for Table B-5: contimed on the following page)

B-7 Notes for Table B-5: -

1. Upon review of reinspection report for equipment setting for Inspector D, it was found that the reinspection had been performed on an installation which has been reworked since the time of the original inspection, thereby making reinspection of the original inspector "not recreatable." As a result of this, the results for Attribute 4 reported in the January 12,1984, report have been removed in accordance with Program requirements.

2. Upon review of reinspection reports for equipment modification, it was found that the summary tabulation for Inspector "U" had been entered into equipment setting rather than equipment modification tabulation. As a result cf this, the results repcrted for Attribute 4 in the January 12, 1984,. report have been removed and located appropriately in Attribute 6.

3. Upon completion of the initial accumulation of data, Inspector (W) failed to achieve the subjective acceptance criterion. Upon further review of reports rejected for "not per detail" and " arc-strikes"it was found that some reports had been improperly graded; for example, the "not per detail" was a condition where excess weld was present and

" arc-strike" reported as a visual v. eld discrepancy was not present on the weld itself. After correcticq of these items, the results were accumulated as tabulated above.

B-8 Table B-6 Detailed Inspector Results Powers-Azco-Pope Attributes Inspector- No. I No.2 No.3 No.4 No. 5 No.6 No.7 A 23/23 11/11 6/10 4/7 - - -

A(exp.) - - - - - - -

B 54/54 34/35 ?5/33 11/11 - - -  ;

B(up.1) - - - - - - -

C 43/43 64/70 26/29 19/33 - - -

C(exp.1) 9/9 39/42 11/19 33/47 - -

C(exp. 2) ' -

86/93 -

33/35 - - -

D 1/1 16/18 193/195 50/52 - - -

D(exp.1) - - - - - - -

E 30/31 123/151 33/33 36/52 - - -

E(exp.1) 19/19 138/151 63/66 51/65 -

1/1 -

E(exp. 2) -

19/21 -

5/5 - - -

F 45/47 61/92 18/20 27/39 - - -

F(exp.1) ' 111/112 155/181 76/76 13/23 - - -

F(exp. 2) -

60/72 -

23/29 - - -

G 40/41 56/64 20/20 0/1 - - -

G(exp.1) 33/33 287/298 42/42 37/48 3/3 2/2 -

H 10/10 24/24 68/68 43/55 - - -

H(exp.1) - - - - - - -

I 10/11 114/116 61/62 2/13 4/8 1/1 -

! I(exp.1) 27/27 141/166 186/187- 39/68 7/14 -

4/4 I(exp. 2) - - - - - - -

J 47/59 163/178 139/142 97/112 -

1/1 -

3(exp.1) - - - - - - -

K 53/53 -19/19 89/89 4/34 - - -

K(exp.1) 55/55 80/82 173/173 64/81 7/16 -

2/2 K(exp. 2) -

3/5 - - -

L 137/139 83/83 215/221 36/40 12/18 -

5/5 L(exp.1) - - - - - - -

, M. 115/116 433/442 687/742 56/134 13/38 2/2 5/5 M(exp.1) 181/187 103/104 164/170 26/47 3/3 -

5/5 M(exp. 2) - -

555/561 208/246 18/22 - -

N 222/228 318/329 306/308 78/103 - -

1/1 N(exp.1) 56/56 147/152 195/197 94/115 - - -

-0 244/248 204/227 477/490 104/141 - -

2/4 O(exp.1) 116/123 168/I74 311/329 91/128 2/3 -

5/5 O(exp. 2) - - -

5/7 - -

1/1 P. 11/11 11/12 8/3 4/4 - -

P(exp.1)- - - - - - - -

Q- 196/196 265/275 412/419 288/342 1/1 7/9 1/1

' Q(exp.1) 119/119 116/119 299/306 188/247 -

1/1 -

.Q(exp. 2) - - - - - - -

R 9/9 55/55 120/120 52/38 8/14 -

1/1 R(exp.1) 32/37 42/47 $9/63 33/49 18/20 - -

R(exp. 2) - - -

5 72/72 105/109 109/111 54/64 6C/76 1/1 4/4 S(exp.1) 116/117 4/4 80/89 38/5i 72/76 -

4/4 TOTAL 2236/2286 3744/4016 5226/5389 1949/2591 234/312 l6/18 40/42 (Notes for Table B-6: continued on the following page)

4 B-9 Notes for Taide B-6:

The."exp.". designation represents the expansion of an inspector's sample period when the

. acceptable threshold was not met.

. Attribute 1 - Pipe material verification

. Attribute 2 - Pipe weld visual-Attribute 3 - Hanger / mount inaterial verification Attribute 4 - Hanger weld visual Attribute'5 - Final hanger Attribute 6 - Flex hose Attribute 7 - Pipe bend

<.w<-- .-s . , . _ , , .. c . ,_... 3,.p.m.,, ,y. _ ....m,_ . , _ . _ _ _ _ , _ , _ , y__ _

B-10 Table B-7 Detailed Inspector Results )

Pittsburgh Testing Attributes Inspector No.1 No.2 A -

1759/2125 B -

442/487 ,,

c- -

35/68 C(exp) -

27/28 D -

18/18 E 522/524 -

F -

506/616 G -

11/12 H -

7/7

-I -

517/558 3 -

749/929 J(exp) -

377/497 K 299/300 -

L 377/381 -

M 1057/1058 -

N 859/874 -

O 975/1008 -

P 933/935 -

Q 883/890- -

R 46/46 -

S -

125/131 l T -

68/69 U --

482/482 V -

78/79 W -

31/31 TOTAL 5951/6016 5232/6137 Notes for Table B-7:

• Expanded sampling was required. T, U, V, and W were added in Attribute 2 due to failure of J.

The "exp " designation represents the expansion of an inspector's sample period when the acceptable threshold was not met.

Attribute l - Concrete expansion anchors Attribute 2 - Visual welding i

I

B-ll 4

Table B-8  ;

Detailed inspector Results Peabody Attribute Inspector No.1 A 14/26 B 13/14 C. 16/24 D 15/16 EL 28/41 F 37/42 TOTAL 123/163' Notes for Table B-8:

100% of accessible work was reinspected.

Attribute 1 - Visual welding 85

~

C-1 APPENDIX C ENGINEERING EVALUATION OF SUBJECTIVE DISCREPANCIES A. INTRODUCTION This appendix has been reformatted from the Appendix C submitted with the January 12, 1984, Interim Report. A sample of the subjective (weld) dis-crepancies was evaluated for the Interim Report. All weld discrepancies have been evaluated and tabulated for this report. The tables which form part of this appendix tabulate discrepancies by type and by method of engineering evaluation.

B. - QUANTITY OF SUBJECTIVE (WELD) INSPECTION DISCREPANCIES The Reinspection Program -identified 4,132 weld discrepancies associated with visual weld quality inspection out of 44,980 welds inspected. Table C-1 summarizes the number of welds inspected ai,d weld discrepancies for each contractor.

Table C-1 Summary of Weld Discrepancies by Contractor No. of Welds No. of Weld C_ontractor Inspected Discrepancies Blount Brothers 0* N/A Johnson Controls 1,459 65 Hunter 3,725 109 NISCo 229 0 Hatfield Electric 26,660 2,117 Powers-Azco-Pepe 6,607 914 Pittsburgh Testing 6,137 905 Peabody Testing 163 22**

_ TOTALS 44,980 4,132

• Inspection of Blount Brothers was performed by Pittsburgh Testing. The inspection results are reported under Pittsburgh Testiag.

• • 40 discrepancies were identified; [8 were located in n.vi-safety related structures.

C-2 C. CATEGORIZ ATION OF SUBJECTIVE DISCREPANCIES An engineering evaluation has been performed for each observed subjective l (weld) discrepancy. The evaluation methods used can be divided into three l

categories. These three categories are related to the acceptance criteria for l visual weld ir.2pection. The acceptance criteria consists of inspecting welds for arc strike, spatter, convexity, crater, incompl.ete fusion, overlap, porosity, undercut, underrun,~ and cracks. The presence of these weld inspection items are considered as weld discrepancies. These weld discrepancies vary in degree as to their effect on weld capacity.

Category X - Evaluanon by comparisca with current design parameters and tolerances.

Categcry X ccntains weld discrepancies that do not reduce the weld capacity. Arc strikes and spatter are cosmetic indi-cations that relate only to appearance. Convexity relates to weld metal on the face of a weld in excess of the weld metal necessary for the required weld size. Convexity has no effect on weld capacity (see Exhibit C-2 Section C.1).

~

Category Y - Evaluation based on engineering judgment by comparison of the discrepancy with design margins.

Category - Y contains some of the following weld dis-crepancies: crater, incomplete fusion, overlap, porosity, undercut, or underrun. Portions of the weld with these discrepancies are considered ineffective, and weld capacity is based on a reduced weld length. Engineering judgment is used to evaluate the weld discrepancies based on the available design margin in the weld and the reduced weld length, which accounts for the assumed ineffective portions. Typically, this results in less than a 10% ieductica in weld strength

. properties to account for the weld discrepancies.

C-3 l- Cate2ory Z - - Evaluation by engineering calculations.

Category Z evaluations are based on reducing the weld length to account for the presence of weld discrepancies as given for Category Y. Two . welds with cracks were evaluated in Category Z. The method for evaluating the discrepancies is based on engineering calculations because the magnitude and types of discrepancies cannot be judged as adequate without a .

detailed calculation.

D. RESULTS OF SUBJECTIVE DISCREPANCY EVALUATION Tne results of the subjective discrepancy evaluation for each contractor are summarized in Table C-2.

. A more deta i led breakdown of discrepancy evaluation is shown for each contractor in Exhibit C-1, which contains Tables CE-1 (Johnson Controls), CE-2 (Hunter), CE-3 (Hatfield), CE-4 (Powers-Azco-Pope), CE-5 (Pittsburgh Testing),

and CE-6 (Peabody).

A more detailed description of the engineering evaluations that were performed is presented in Exhibits C-2 (AWS Welding) and C-3 (ASME and ANSI B31.1 Welding).

a s

~p e + , s, . - w. . -

.-e ,

~

m

.. C-4 Table C-2 Summary of Subjective Diacrepancy Evaluation Results l No.of' Category X Category Y Category Z No. with Discrepancy No. Within No. Acceptable No. Acceptable Design Contractor Evaluations E-arameters . by Judgment by Calculation Significance Blount Brothers

• N!A. N/A N/A N/A 0 Johnson Controls 65 15 12 38 0 Hunter - 109 25 23 61 0 NISCo 0 0 0 0 0 Hatfield Electric 2,117 11 2,064 42 0 Powers-Azco-Pope 914 201 77 636 0 Pittst urgh Testing 905 1 887 17 0 Peabody Testing 22 0 11 11 0

! TOTAL 4,132 253 3,074 805 0 ,

• Inspection of Blount Brothers was performed by Pittsburgh Testing. laspection results a e reported under Pittsburgh Testing.

Table C-2 shows that 6% of the discrepancies identified in the Reinspection Program as Category X are not " valid" discrepancies and represent work that is within current design . parameters. The Category X discrepancies result primarily from design parameters that have been expanded since the time of ~

the original inspection and therefore are within current design limits.

.The Category Y evaluation in Table C-2 indicates that 74% of the observed weld discrepancies, wherein weld capacity was reduced by approximately 10%

after accounting for the weld discrepancy, are acceptable. In all cases, the

' design margin remained within design limits.

3 The Category Z evaluation in Table C-2 indicates that 19% of the observed weld discrepancies are acceptable.' The reduction in weld capacity varied after 2 accounting for the weld discrepancy. However, in all cases, the design margin remained within the specified design limits. .

C-3 e

l in summary, the engineering evaluation of subjective discrepancies has shown l 1that the weld discrapancies are not significant to the overall design of the plant. .

The results of the engineering evaluation are presented below for each contractor:

1. Johnson Controls The _ welding performed by Johnson Controis is of good quality. The work inspected and the discrepancies evaluated are representative of the highiy stressed components welded by Johnson Centrols.

2. Hunter -

The welding performed by Hunter is of good quality. There were a small number of discrepancies and the evaluation indicated that no discrepancy had design significance.

3. Hatfield Electric The engineering evaluation has shown that the welding performed by Hatfield Electric i:; of good quality. The work inspected and the discrepancies evaluated are - representative of the highly stressed

, components welded by Hatfield Electric. Ho.vever, in order to expand the data base and respond to specific questions asked by the NRC staff (refer

- to question Q7 in Appendix F), the additional inspections and evaluations discussed in Exhibit C-2 are proceeding.

4.: Powers-Azco-Pope The welding performed by Powers-Azco-Pope is of good quality. The work inspected and the discrepancies evaluated are representative of the highly stressed components welded 'uy Powers-Azco-Pope.

C-6

5. Pittsburgh Testing The engineering evaluation has shown that the we.* ding inspected by Pittsburgh Testing is of good quality. In order to expand the data base and respond to specific questions asked by the NRC staff (refer to question Q7 in Appendix F), the additional inspections and detailed evaluations described in Exhibit C-2 are proceecing.

6. Peabody Testing The welding inspected by Peabody is of good quality. The reinspected we:ds are representative of the welds inspected by Peabody.

E. CONCLUSIQN None of the observed discrepancies had design significance and, in all cases, the design margin remains within specified design limits. The work by the contractors is of good quality.

t.

l

EXHIBIT C-1 Page 1 of 4 EXHIBIT C-1

SUMMARY

OF SUBJECTIVE DISCREPANCY EVALUATION BY CONTRACTOR Table CE-1 Summary of Subjective Discrenar.cy Evaluation Johnmn Controls Type of Categy X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design by Attribute Quantity Parameters by Judgment by Calculation Significance Visual w91d

1. Tube Track / 65 15 -12 38 0 Instrument Supports No*e fy Table CE-1:

- 1. Categories X, Y, and Z are defined in Section C of Appendix C.

l

. - - . . . _ . , , _ - - . , , . , - . . ~ _ - . - , - _ _ . - - - -

EXHIBIT C-1 Page 2 of 4 Table CE-2 Summary of Subjective Discrepancy Evaluation Hunter Type of Category X Category Y Category Z No. with Discrepancy _ Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance Visual weld i Pipe supports 60 19 18 23 0 and pipe whlp restraints. ,

2. Large bore 1 2 1 0 0 butt weld (ND)

3. Socket angle 30 3 4 23 0 fillet welds (NB, NC, ND)

4. Support welds 14 1 0 13 0 (NF)

, 5. Pipe penetra- 2 0 0 2 0 tions and reinforcing saddles

-TOTAL 109 25 23 61 0 Note for Table CE-2:

1. Categories X, Y, and Z are defined in Section C of Appec.Jix C.

i f

,,w - - -. - - ., - - , , - -~

..p

EXHIBIT C-1 Page 3 of 4 Table CE-3 Summary of Sub ective Discrepancy Evaluation Hatfield Electric Type of Category X Category Y Category Z No. with

' Discrepancy Total - No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation gnificence Visual weld 2117 11 2064 42 0 Note for Table CE-3:

~ 1. Categories X, Y, and Z are defined in Section C of Appendix C.

Table CE-4 Summary of Subjective Discrepancy Evaluaticn Powers-Azco-Pope Type of Category X Category Y Category Z No. with Discrepancy Total Nc. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance Visual weld

1. Instrument -608 167 77 364 0 tubing supports

2. Socket welds 44 1 0 43 0 (NC)

3. Socket welds 24 .11 0 13 0 (ND)

4. Support 34 0 0 34 0 welds (NF)

5. Socket welds 204 22 0 182 0 (B31.1)

TOTAL- 914 201 77 636 0 Note for Table CE-4:

1. Categories X, Y, and Z are defined in Section C of Appendix C.

I

(.'

EXHIBIT C-1 Page 4 of 4 l

l Table CE-5 )

Summary of Subjective Discrepancy Evaluation l Pittsburgh Testmg l

Type of. Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance Visual weld 905 1 887 17 0 Note for Table CE-5:

1. Categories X, Y, and Z are defined in Section C of Anpendix C.

S Table CE-6 Sumraary of Subjective Discrepancy Evaluation Peabody Testing Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity- Parameters by Judgment by Calculation Significance Visual weld 22 0 11 11 0 No'te fer Table CE-6:

1. Categories X, Y, and Z are defined in Section C of Appendix C.

l

EXHIBIT C-2 Page 1 of 15 EXHIBIT C-2 -

ENGINEERING EVALUATION OF AWS WELD DISCREPANCIES A. GENERAL The purpose of the engineering evaluation was to determine the design impact of weld discrepancit.s noted in the Reinspection Program. This engineering evaluation demonstrated that the observed weld discrepancies at Byron Station do not impair the structural integrity of any structure. The design margins of welds with their associated discrepancies are within the specified design limits.

Design margin is a measure.of the excess capacity in the weld design. The design margin is reported as a factor which is the ratio of design basis allowable stress to the actual stress in the weld. If this factor is equal to or greater than one, the weld design is within the specified dcrigo limits.

Ine engineering evaluation was approached from two directions:

1. Identify the lowest quality welds from the population of observed weld discrepancies and then evaluate the daign impact of these weld discrepancies on the component capacity.

2. Identify highly stressed components from the population of observed weld

discrepancies and assess the design impact of discrepant welds associated with these components.

B. CATEGORIZATION OF AWS WELD DISCREPANCIES i The AWS DI.1 Structural Welding Code (Reference Cl) formed the basis for visual weld inspection. The visual weld inspection items are categorized as

follows:

-Category A Weld discrepancies that are structurally acceptable because they do not reduce the load-carrying capacity of the weld and i

EXHIBIT C-2 Page 2 of 15 therefore have no design significance. These are limited to arc strike, convexity, and spatter.

Category B - Weld discrepancies that result in a reduction of the size.

length, or capacity of the weld. These include craters,

. incomplete fusion,. overlap, porosity, undercut, and underrun.

Category B weld discrepancies are further subdivided into Categories B1 and B2 to quantify the significance of these weld discrepancies as follows:

Category B1 Weld discrepancies that have capacity reductions of less than 10%.

Category B2 Weld discrepancies that have capacity reductions of equal to or greater than 10%.

Category C Weld discrepancies that are assumed unsuitable for load transfer and .esult in total weld rejection. Cracks are the only c.tse for this category.

According to the American Welding Society, a weld discrepancy is defined as "An interruption of the typical structure of a weldment, such as a lack of homogeneity in the mechanical, metallurgical or physical characteristics of the material or weldment. A discontinuity is not necessarily a defect" (Reference

. C2).

The terms Category X, Category Y, and Category Z huvc been used to categorize the evaluation methods used for the AWS weld discrepancies in Chapter VI and in this appendix. Categories A, B, and C have been used to categorize the significance of the weld discrepancy. The evaluation methods and weld discrepancy significance are related. Category X is equivalent to Category A, Category Y is equivalent to Category B1, and Category Z is equivalent to Categories B2 and C.

l i

EXHIBIT C-2 Page 3 of 15 C. - EVALUATION PROCEDURE FOR AWS WELD DISCREPANCIES

1. Category A Weld Discrepancies Welds with Category A discrepancies are structurally acceptable. They do not reduce the load-carrying capacity of the weld and therefore have no design significance.

Arc strikes and spatter are cosmetic discrepancies and only have impact on weld strength when they appear in an extreme case. In this evaluation, there were no discrepancies reported associated with arc strikes or spatter in the extreme case.

Convexity is cniy censidered a potential problem in fatigue design of welds. The structures and supports in a nuclear power plant are subjected to static and dynamic loadings, but not fatigue type loading. Fatigue is defined as the damage that may result in fracture af ter a sufficient number of fluctuations of stress. The occurrence of wind, earthquake, or pipe transients is too infrequent to warrant consideration in fatigue design. The loading conditions for fatigue loading require a minirnum of 20,000 cycles during the life of the structure. The actual number of loading cycles at Byron Station for wind, carthquake, or pipe transients is less than 2,000 for the plant life. Therefore, fatigue is not an applicable loading condition in the design of structures and supports in nuclear power plants. This discussion of non-fatigue loading condition also applies to Category B and Category C weld discrepancies.

2. Category B Weld Discrepancies The procedure that was used to evaluate welds with Category B discrepancies considered those portions of the weld with discrepancies to be totally ineffective and then calculated the section properties of the remaining weld. This is a conservative approach because it has been shown li

EXHIBIT C-2 Page 4 of 15 that even the portion of the weld with this type of weld discrepancy does

. exhibit significant load-carrying capability.

The AWS code requirements for welder qualification and weld procedure q'u alification ensure tht weld discrepancies are of a localized nature and do not exist throughout tne weld length. Any localized weld discrepancies vill not propagate because the weld metal and the base metal have

adequate . fracture toughness. The discrepar.cies observed in the 7 ,

Reinspection Program were localized, as would have been expected.

m. .

The fracture toughness of material is a measure of its capacity to absorb energy without failure or damage. Generally, a material is considered y ' " tough" when, under stated conditions of stress and temperature, the material can withstand leading to its design limit in the presence of flaws, including: notches. Toughness means that, under specified conditions, the

' - material has' the capability to arrest the growth of a flaw. Fracture toughness of a' material is better at higher temperatures.

One measure of the fracture.. toughness of a material is its nil ductility temperature (NDT).' Materials used at temperatures above its NDT provis

assurance that any flaws will be arrested under service conditions. The.

criteria contained in NRC Report NUREG 0577, Revision 0.(Reference -

C3), was used to determine that the NDT temperature for weld and base

~

metals are less than the lowest l service temperature. This provides

. assurance that Category B discrepancies are nonpropagating and any flaws s -due to these discrepancies will be arrested.

~

Stress increases locally in the _ area of discontinuities on weld and base

- metal. . These stress intensification effects are negligible if the materials have adequate: fracture toughness properties. The presence of

^

D discontinuities such as Category B discrepancies are acceptable due to the fracture toughness properties of the weld and base metals.

~

h

EXHIPIT C-2 Page 5 of 15 It is, therefore, justified and conservative to disregard the portions of the weld with Category B discrepancies and compute the section properties of the remaining weld to evaluate its load-carrying capacity.

3. , Category C Weld Discrepancies The procedure that was used to evaluate welds with Category C discrepancies rejected the entire weld length. The remaining assembly of welds forming the weldment was evaluated to determine the component capacity and to show that the capacity is still within design allowables.

Additionally, an evaluation was made to ensure that the crack does not propagate into the' base metal.

D. AWS WELD EVALUATION RESULTS BY CONTRACTOR

1. Johnson Controls (JCI) Evaluation Results - AWS Weld Discrepancies The inspection work performed by 3CI included instrument supports and tube track supports. A total of 1,459 welds were inspected and 65 weld discrepancies were identified.

An engineering evaluat'cn of all 65 discrepancies was comp!ced. In all cases, the design of the component with the presence of the discrepancy

- was found to be acceptable. The results are categorized in Table CE-7.

s

EXHIBIT C-2 Page 6 of 15 Table CE-7 Results of AWS Weld Discrepancy Evaluation 7ohnson Controls Weld Discrepancy Category A B1 R2 C Weld Weld No Strength Strength Weld No. of Weld Structural Reduced Reduced Rejected Weld Type Discreparcies Impact by < 10% by 110% (Cracks)

Tube track /

instrument supports 65 15 12 38 0 Typically, supports installed by 3CI have a large design margin. The supports were designed for peak seismic accelerations rather than for location specific accelerations. Supporc types are setected from generic design tables, which envelope the various design considerations, and the supports use standard member sizes. Whenever standardization in design is used, design margin is generally greater. The component supports associated with these 65 weld discreparries have a design margin greater than 1.2, based on the origkal design. This des!3n margin is representative of the highly stressed elements in the plant. The maximum weld strength reduction based on the lowest quality weld was applied to all of the supports associated with the 65 discrepancies. In all cases, after performing a more exacting analysis, the design margin remained greater than one.

2. Hunter Evaluation Results - AWS Weld Discrepancies The inspection work performed by Hunter included pipe supports and pipe whip restraints. A tetal of 3,725 welds (see Table C-1) were inspected and 50 AWS weld discrepancies were identified. The 49 Hunter ASME weld discrcpancies are discussed in Exhibit C-3 and Table CE-13.

An engineering evaluation of all 60 AWS discrepancies was completed.

.all cases, the design of the component was acceptable with the observed discrepancy present. The results are categorized in Table CE-8.

EXHIBIT C-2 Pcge 7 of 15 Table CE Results of AWS Weld Discrepancy Evaluation Hunter Weld Discrepancy Catcgory A B1 B2 C

~

Weld Weld No Strength Strength Weld No. of Weld Structural Reduced Reduced Rejected Weld Type Discrepancies Impact by < 10% by 210% (Cracks)

Pipe supports and pipe whip restraints 60 19 18 23 0 The results of the engineering evaluation of Hunter AWS welds indicate that each of the components are adequate to carry the derign loads with the observed discrepancies present.

Based on the small number of discrepancies and the evahation which determined that no discrepancy had design significance, the AWS welding performed by Hunter has been determined to be of good quality.

3. Hatfield Electric Evaluation Results - AWS Weld Discrepancies The inspection work performed by Hatfield included ' conduit supports, junction box supports, cable tray supports, cable tray hold-down welds and ,

auxiliary steel for electrical supports. A total of 26,660 welds were reinspected and 2,117 weld discrepancies were identified.

A detailed review of the reinspection records for all 2,117 discrepancies was made. This review indicated that there were only two cracked welds.

In order to achieve 95% reliability with 95% confidence, a statistical sampling plan was chosen in accordance with Military Standard 105D. The resulting sample size for the engineering evaluation was 100 welds. The sample was conservatively biased by including the 50 welds that the third-party inspector identified as having the most weld quality discrepancies.

The two welds with cracks were part of that group. The remaining 50 welds were randomly selected.

EXHIBIT C-2 Page 8 of 15 The results of the engineering evaluation for the sample of 100 Hatfield welds are shewn in Table CE-9.

Table CE-9 Results of AWS Weld Discrepancy Evaluation Hatfield Electric Weld Discrepancy Category A B1 B2 C Weld Weld No Strength Strength Weld No. of Weld Structural Reduced Reduced Rejected Weld Type Discrepancies impact ,by < 10 % by 2 10% (Cracks)

a. Conduit /

junction box supports and associated auxiliary steel 35 2 17 15 1**

a.- Cable tray st.pports and associated auxiliary steel 36 1 17 13 0

c. Cable tray hold-down 29 2 19 7 l' TOTAL 100 5 53 40 2***

• One of the two ho'd-down welds attaching the cable tray to its support was

-cracked. It was futend that, af ter subtracting the entire length of the cracked weld, the other weld was sufficient to transfer the design loading.

• • Temporary tack weid used to aid construction was cracked. The tack weld is not required by design. There is no crack in the design weld.

• • • The potential of crack propagation into the base metal was evaluated. For the two reported cases, based on the fracture toughness of the materials, it was determined that the cracks will not propagate into the base metal under the maximum design loading and minimum plant operating temperatures.

Design margins exist in conduit and junction box supports and associated auxiliary steel because the initial design conservatively assumed maximum

EXHIBIT C-2 Page 9 of 15 cable weight in each conduit. In addition, the supports and auxiliary. steel

--are conservatively designed for peak seismic acceleration. When a more

- exact calculation - was performed using actual- cable loads and actual seismic acceleration, the design margin exceeded a 1.5 factor. This design margin is representative of the highly stressed conduit end junction box supporttt and' associated auxiliary steel in the plant. - The weld strength reduction for all but the two lowest quality welds was applied to all of the components with weld discrepancies and the weld stresses remained within design basis allowables. The two lowest quality welds were evaluated and those supports have a design margin greater than one.

- A design margin exists in welded connections for cable tray supports and cable tray support ~ auxiliary steel because the initial design was conservatively based on a maximum uniform ca.ble load. In addition, the components _ are generally designed using simplified, yet conservative, techniques. By using actual cable tray loadings and more exacting methods of -analysis,' it was shown that the actual stresses sre lower than the stresses . from the original design.' For the 36 cases where a detailed engineerir.g evaluation of the weld discrepancy was performed, the welds -

are adequate to carry the loads. The average value of the weld strength reduction for cable tray supports and auxiliary steel is approximately 13% ' This reduction is not signifh: ant to the overall behavior of the 4 support system.

3 q A design margin exists for cable tray hold-down welds because the initial-i~ . design conservatively assumed maximum cable weightimaximum cable

( tray span and peak seismic acceleration. When a more detailed calculation

.is' performed for , any- particular cable tray hold-down weld using ' the seismic values -for that particular location, actual cable tray -loads and

actual cable : tray spans, there is -additional design margin. There is

additional design m'argin for- the -29 cases where a. detailed engineering evaluation was pe; formed, and. the welds are adequate to carry the design

- loads.

' p :(

T T

% l 4_

EXHIBIT C-2 Page 10 of 15 The results of tne engineering evaluation of the Hatfield Electric welds indicate that each of the components are adequate to carry the design loads with the observed weld discrepancies present. The results of the engineering evaluation, combined with the large number of reinspected welds without discrepancies, indicate that the welding performed by Hatfield Electric results in components capable of carrying the design loads.

The evaluation has shown that the Hatfield work is of good quality However, to expand the data base and in response to a specific questica asked by the NRC staff regarding assurance "that the welds with the lowest factor of safety comply with the applicable codes and the design basis"(refer to question Q7 in Appendix F), the following two programs are being implemented:

a. The 50 welds with the lowest design r. argins from the Hatfield Electric discrepancies have been identified. A detailed engineering evaluation on each of these welds will be performed to verify that they comply with the appPcable codes and the desigr: requirements,

b. A representative sample of the highly stressed welds will be identified for Hatfield Electric inspectors not alreaay included in the Reinspection Program. A detailed engineering evaluation will be performed on each weld where a discrepancy is found to <erify that the welds comply with the design requirements.

4. Powers-Azco-Pope Evaluation Results - AWS Weld Discrepancies The inspection work performed by Powers-Azco-Pope included instrument supports and instrument piping supports. A total of 6,607 welds were reinspected and 608 AWS veeld discrepancies were., identified. The 306 Powers-Azco-Pope ANSI B31.1 and ASME weld discrepancies are discussed in Exhibit C-3 and Table CE-14.

EXHIBIT C-2 Page 11 cf 15 An engineering evaluation of all 608 AWS weld discrepancies was com-pleted. In all cases, the design of the component was acceptable with the observed discrepancy present. The results are categorized in the Table CE-10.

Table CE-10 Results of AWS Weld Discrepancy Evaluation Powers-Azco-Pope Weld Discrepancy Category A B1 B2 C

- Weld - Weld No Strength Strength Weld

. No. of Weld Structural Reduced Reduced Rejected Weld Type Discrepancies impact by < 10% by .)_10% (Cracks)

Instrument tubing supports 608 167 77 364 0 The supports installed by Powers-Azco-Pope typically have a large design margin. The supports are designed for peak seism'c accelerations. These supports are selected from generic design tables which envelope the various' design considerations and use standard member sizes. Twenty-eight.of the _ supports associated with the 608 discrepancies had a design margin of 1.1 or less. This is representative cf the highly stressed supports installed by Powers-Azco-Pope. The maximum weld strength reduction based. on the lowest quality weld was applied to all of the supports associated with the 608 discrepancies. In all cases, after performing a

~

more exact analysis, the design margin remained greater than one and had no design significsnce.

The results of the engineering evaluation of Powers-Azco-Pope AWS weld discrepancies indicate that the - Reinspection Program has captured a

. representative sample of highly stressed elements with lowest quality

welds and that there is no design significance.

t f

m e . = - . w- y - - -...v - . - . - - , -.- , . - , - - . -y-- y- g

L EXHIBIT C-2 Page 12 of 15

5. Pittsburgh Testing Laboratory Evaluation Results - AWS Weld Discrepancies The,inspe

• ion work performed by Pittsburgh Testing covered structural steel work performed by Blount Brothers, American Bridge and Mid-City Architectural. A total of 6,137 welds were reinspected and 905 AWS weld discrepancies were identified.

A detailed review of the reinspection records for all 905 discrepancies was made. This review indicated that there were no cracked welds, In order to achieve 95% reliability with 95% confidence, a statistical sampling plan wu, chosen in accordance with Mihtary Standard 105D. The resulting sample size for the engineering evaluation was 64 welas. The sample was conservatively biased by including the 50 welds that the third-party inspector identified as having the most weld quality discrepancies. The remaining 14 welds were randomly selected. The remaining 841 discrepancies were reviewed to assure that the numbers and types of discrepancies within the sample were representative of the entire group.

The results of the engineering evaluation for the sample of 64 Pittsburgh Testing weids are shown in Table CE-II.

Table CE-Il Results of AWS Weld Discrepancy Evaluation Pittsburgh Testing Weld Discrepancy Category A B1 B2 C I Weld Weld No.of No Strength Strength Weld Weld Structural Reduced Reduced Rejected Weld Type Discrepancies Impact by < 10% by1 10% (Cracks)

- 3tructural steel work 64 10 37 17 0

EXHIBIT C-2 Page 13 of 15 The welds associated with the 905 observed AWS weld discrepancies-involved 312 connections. Each was reviewed and those with the highest level of stress were identified. A representative number of connections,6 of 44, with a design margin less than' 1.3 were among those chosen for engineering evaluation, and they were found to have no design significance. Further detailed evaluation was not required because items with a design margin greater than 1.3 are capable of carrying the design loads even if the percentage reduction from the most discrepant of the 30

-lowest quality welds is applied to these items.

For the 64 cases where a detailed engineering evaluation of the weld discrepancies was performed, the welds are adequate to carry the design loads. The average value of the weld strength reduction for Pittsburgh Testing welds is less tnan ten percent. This reduction is not significant to the overall behavior of the structures.

The' engineering evaluation shows that Pittsburgh Testing welds are adequata to carry the design loads with the presence of weld discrepancies of the type observed.

However, to . expand the data base and in order to answer a specific question asked ' by the NRC staff (see Q7, Appendix F), additional inspections and evaluations are being performed.

Pittsburgh _ Testing showed an undesirable fsilure rate for inspection of

- welds with overlap and undercut. The presence of overlap makes visual weld quality inspection more difficult. In order to confirm that welds with overlap are not masking .ither discontinuities, the third-party inspectors will select a suitable number of welds which represent the most severe cases of overlap. The overlapped portion of these welds will be removed by grinding and the weld will be reinspected. - The results of this reinspection will be reviewed and evaluated. For undercut, the reduction in capacity was approximately 5%, which is insignificant.

EXHIBIT C-2 Page 14 of 15 1

I From the population of Pittsburgh Testing visual weld discrepancies, the most highly stressed welds will be identified. A suitable number of these connections will be reinspected and detailed weld maps will be prepared.

Using these weld maps, each of these connections will be evaluated to

. insure that the welds are capable of transferring the required loads.

6. Peabody Testing Evaluation Results - AWS Weld Discrepancies All'of the accessible Peabody Testing work was reinspected. This included 163 welds' from which 40 weld discrepancies were identified. Twenty-two of these weld discrepancies were in safety-related structures. The results of the engineering evaluation of these 22 Peabody Testing weld discrep-ar.cies are shown in Table CE-12.

Table CE-12 Results of AWS Weld Discrepancy Evaluation Peabody Weld Discrepancy Category A B1 B2 C Weld Weld No Strength Strength Weld No. of Weld Structural Reduced Reduced Rejected Weld Type Discrepancies impact by < 1096 by2 10% (Cracks)

All work 22 0 11 11 0 L

All of the weld discrepancies were evaluated in detail and determined to be acceptable. The work performed by Peabody Testing has been deter-mined to be of good quality and no further inspections are warranted. This conclusion is based on the s:7.all scope of work performed by Peabody Testing, the small number of discrepancies, and the evaluation as expanded to 100% of the reinsp2ctable work which determined that no discrepancy had design significance. In addition, Peabody Testing's work was primarily overinspection of other contractors, some of ivhose work was reviewed in the Reinspection Program (e.g., Blount Brothers), and the first line

~

r inspections performed by Peabody Testing were monitored by

EXHIBIT C-2 Page 15 of 15 Commonwealth Edison. This further substantiates the conclusion that Peabody Testing's work is of good quality and that no further inspections are warranted (refer to Exhibit E-9).

o E. CONCLUSION - AWS WELDING DISCREPANCIES The engineering evaluation of AWS weld discrepancies identified in the Reinspection Program demonstrates that the discrepancies at the Byron Station do not impair the structural integrity of any structure or system. The design margins of all welds with their associated discrepancies remained within specified design limits.

F. REFERENCES C1. AWS Di.1, " Structural Welding Code-Steel," American Welding Society, Miami, Florida.

C2. AWS A3.0-80, " Welding Terms and Definition," American Welding Society, Miami, Florida.

C3. NUREG-0577, Rev., " Potential for Low Fracture Toughness and Lau,ellar Tearing on PWR Steam Generator and Reactor Coolant Pump Supports," October 1979, Nuclear Regulatory Commission, Washington, D.C.

EXHIBIT C-3 i Page 1 of 8 I EXHIBIT C-3 - I ENGINEERING EVALUATION OF ASME AND ANSI B31.1 WELD DISCREPANCIES A. GENERAL An engineering evaluation was performed to determine the de:ign significance of ASME and ANSI B31.1 Code welding discrepancies identified by the Reinspection Program. The evaluation confirmed that ASME Code (NB, NC, ND, NF and MC) and ANSI B31.1 discrepant welds are acceptable based upon meeting code design criteria after taking into account the weld discrepancies.

There are three contractors whose scope of work i cluded ASME or ANSI B31.1 Code welding. It was determined that two contractors (Hunter and Powers-Azco-Pope) had discrepancies involving ASME and B31.1 code welding. A third contractor (NISCo) had no ASME or 831.1 code weld discrepancies.

B. CATECORIZATION OF,ASME AND ANSI B31.1 (CODE) WELD DISCREPANCIES The visual weld reinspection discrepancies were assigned into two basic categories:

Category P Weld discrepancies resulting in a reduction in size, length or quality which "1eet the ASME and ANSI B31.1 Code design criteria.

Category Q Weld discrepancies that do not meet ASME and ANSI B31.1 Code design criteria. (There are no discrepancies in this caterpry.)

Note: Tne discrepancies have also been subdivided into Categories X, Y, and y Z, as described in Exhibit C-1.

EXHIBIT C-3 Page 2 of 8 C. , CODE EVALUATION METHODOLOGY i

All weld discrepancies were evaluated against applicable code design criteria.

Weld assessment calculations were performed assuming na weld material existed where a relevant discrepancy was located. The reported discrepancies were deterinined to have no effect on the code fatigue evaluation. In all cases, it was determined that the weld joints with discrepancies met or exceederi code design criteria.

All discrepancies that were determined to exceed code examination acceptance criteria were repaired, even though they were determined by evaluation not to have design significance.

1. Hunter Evaluation Results - Code Weld Discrepancy Evaluation A total of 49 Hunter Code welding discrepancies were identified by the Reinspection Program. The evaluation results are shown in Table CE-13.

Table CE-13 Code Weld Discrepancy Evaluation Hunter Results P Q Acceptable Welds do not No. of Weld Weld Strength meet Code Weld Type Discrepancies Reduction Design Criteria

a. Large bore butt weld (ND) 3 3 0

b. Socket and fillet welds (NB, NC, ND) 30 30 0

c. Support welds (NF) 14 14 0

d. Pipe penetration and 2 2 0 reinforcing saddles TOTAL 49 49 0 The discrepancies in Table CE-13 are all relatively minor in nature and can be

-characterized as follows:

EXHIBIT C-3 Page 3 of 8

a. ASME Large Bore Beit Welds - ND One discrepancy involved a 1/4-inch diameter region of porosity on a 42-inch diameter circumferential pipe weld. The code considers surface indications up to 1/16-inch in diameter as not relevant. The subject indications were all 1/64 inch to 3/64-inch in diameter which fall within ASME Class ND liquid penetrant examination acceptance criteria. This discrepancy was therefore determined to be acceptable since the visual reinspection acceptance criteria exceeded code requirements.

The second discrepancy involved a convex bead on a 30-inch diameter r.ircumferential pipe weld. After grinding to remove the convexity, the weld was then examined by visual and liquid penetrant examination and determined to be acceptable.

The final discrepancy involved a section of weld at a pipe-to-weld neck flange joint which was reported to be under the surface of the flange. The

- weld was reported to be 1/32 inch under the surface of the flange over approximately 25% of the weld length. This occurred only on the flange side of the weld (the pipe side of the weld was acceptable). Flange surfaces commonly are out of exact round due to rnanufacturing tolerances. This rer.ults in the ;ype of discrepancy identified; this is not considered a valid discrepancy because it does not conflict with design requirements or intent. No sharp discontinuities were noted at the flange joint. The observed weld discrepancy was, therefore, determined to be acceptable.

Ib. . ASME Socket and Fillet Welds NB, NC, ND A total of three ASME Class 1 (NB) observed discrepancies were reported.

All three involved slightly undersized seal weld fillet welds for radiographic plugs. The piping has been qualified with the actual seal weld size and determined to be acceptable.

A total of eight ASME Class 2 (NC) discrepancies were reported. Five involved socket weld fillet sizes, one involved porosity in the socket weld,

I J

l EXHIBIT C 3 4 Page 4 of 8 one involved a slag inclusion at the toe of the fillet weld, and the remaining l discrepancy involved a slightly undersized radiographic plug seal weld. For the five fillet weld discrepancies, it.was determined that the fillet weld size indicated on the design documents was larger than required by design. All fillet sizes were found in compliance with ASME Code design criteria when compared to actual design requirements. The surface porosity was

-evaluated as acceptable based on the approved magnetic particle test for

. the weld. The visual reinspection acceptance criteria again exceeded code requirements. The slag inclusion at the toe of the weld had a depth of 0.012 inches and was buffed out; no weld repair was required. The code required

-that minimum weld size was not violated. For the slightly undersized

-radiographic plug seal weld, the piping has been qualified with the actual .

seal weld size and determined to be acceptable.

A total of 19 ASME Class 3 (ND) discrepancies were reported. These included 14 fillet welds identified as undersized,3 welds with undercut, and 2 welds with surface porosity. The 14 fillet welds identified as undersized meet design require:;.ents and therefore are acceptable per the ASME Code design criteria.. The 3 welds with undercut were evaluated and found not to be significant. The remaining 2 welds identified as having surface porosity

~

were evaluated and ' found to be acceptable. (Also, ASME porosity inspection requirements are nonexistent for Class 3 socket welds 2 inches or less in diameter.)

c. ASME Support Welds - NF A total of 14 discrepancies were reported for ASME Class NF welds. Eight of :these involved fillet weld length, three involved fillet weld size reduction, and one involved undercut. These observed discrepancies were evaluated on the- basis of _the actual design capacity of the associated supports with the observed weld discrepancies taken into consideration. All

. welds were determined to be acceptable.

l EXHIBIT C-3 Page 3 of 8 One discrepancy involved paint or, a weld which resembled a crevice snat was not actually a discrepancy (upon removal of the paint), and hence was acceptable.

The final discrepancy involved a fillet w2ld jcining a strut bracket to a steel plate. A portion (1-1/2 inches) of the 5-inch weld was reported to have lack of fusion and undercut along the steel plate. The actual length of specified weld for the bracket exceeded desid n requirements. The weld was determined to meet ASME Code design critaria when compared to the actual design requh ements. Since the region exhibiting the lack of fusion consisted of excess weld, the support weld was found to meet the ASME Code design criteria and to be acceptable. In addition, the weld with :he discrepancy was evaluated with respect to the maximum design capacity of the support component and was determined to be acceptable with a design ,

margin of 2.

f

d. ASME Pipe Penetration and Reinforcing Saddles - MC, ND One discrepancy was reported involving an ASME Class MC penetration closure plate weld. This discrepancy involved the weld joining the penetration closure plate to the pipe wall. A full penetration weld with a 5/8-inch by 3/8-inch reinicrcing fillet was specified. The 3/8-inch leg was joined to the pipe surface. The discrepancy involved a 1/8-inch reduction in reinforcing leg length cf the larger leg. The observed reduction in weld leg was evaluated (compared with actual design requirements) and determined to be acceptable.

The second discrepancy involved an ASME Class 3,16-inch diameter pipe, reinforcing pad, attachment weld. The specified attachment weld was a 3/8-inch fillet attaching a 3/4-inch thick reiforcing pad to the pipe wall.

Several sections of the weld were reported to be undersized by 1/16- to 1/8-inch. The entire weld was conservatively reduced in size by 1/8-inch resulting in a.'l/4-inch filist weld for evaluation (comp +ed with actual design requirements). It was determined that the weld was acceptable with l

l-

i-EXHIBIT C-3 Page 6 of 8 a design margin in excess of 30. In addition, is should be noted that fillet-welds on pipe surfaces are difficult to measure due to the curvature of the pipe surface; welding gages will typically indicate slightly undersized weld legs due to this pipe curvature.

e. Summary In summary, all the ASME weld discrepancies were evaluated in detail and datermined to be acceptable when compared with the design requirements and with ASME Code design criteria. Based on the small number of discrepancies and the evaluadon which cete mined that no discrepancy had design significance, the ASME work performed by Hunter has been determined to be of good quality and no further inspections are warranted. .

2. Powers-Azco-Pope Evaluation Results - Code Weld Discrepancy Evalue. tion All the accessible, ecreatable Powers-Azco-Pope work was reinspected, kr those inspectors who did not meet the program acceptance criteria. A total of 306 weld discrepancies were identified involving ASME and ANSI B31.1 piping support welds as noted below in Table CE-14.

Table CE-14 Code Weld Discrepancy Evaluation Results Powers-Azco-Pope ,

P Q Acceptable Welds No. of Weld Weld Strength do not meet Weld Type Discrepancies Reduction Code Design Criteria

a. Socket welds (NC) 44 44 0

b. Socket welds (ND) 24 24 0

c. -Support welds (NF) 34 34 0

d. Socket wd.ds (B31.1) 204 204 0 TOTAL 306 306 0 '

l

D EXHIBIT C-3 Page 7 of 8 The discrepancies in Table CE-14 are ah minor and can be characterized as follows:

a. . ASME Socket Welds - NC, ND A total of 68 ASME socket weld discrepancies were reported. These were mainly leg and throat welding discrepancies. No incomplete fusion, overiap or undercut discrepancies were reported.

The 68 ASME socket weld discrepancies were evaluated for compliance with ASME Code design criteria. All discrepancies involved a slight reduction in fillet weld size. In all cases, an evaluation was made to determine the minimum fillet weld size required by design. All fillet weld sizes were found to meet ASME Code design cri;eria.

b 'b. ASME Support Welds - NF A total of 34 support weld discrepancies were identified for ASME Class NF welds for small bore piping supports. All 34 discrepancies involved undersized tubing U type strap hold down fillet welds. A 1/4-inch fillet weM was specified; however, the installed welds were undersized by as much as 1/16-inch. The strap welds were evaluated for the maximum ,

design load and were determined to be acceptable.

c. ' Non-ASME A . total of 204 ANSI B31.1 wdd discrepancies were reportea. These

-included leg and throat discrepancies, minor reductions in weld legs (1/32-inch to . 1/16-inch), items. acceptable to current design / installation

! tolerances, surface ripples, overlap, slag and scale, and one undercut.

The minor reductions in weld leg with a resulting 1/32-inch to 1/16-inch reduction in weld size were conservatively evaluated on the basis of 1/16-inch less weld over the entire weld circumference. All welds met design requirements and ANSI B31.1 Code design criteria.

EXHIBIT C-3 Page 8 of 8

d. Summaty In summary, all - the ASME and ANSI B31.1 weld discrepancies were evaluated in detail and determined to be acceptable wiien compared with the design requirements and with code design criteria. Based on the small number of discrepancies and the etatuation which determined that no discrepancy had design significance, the ASME and ANSI B31.1 work performed by Powers-Azco-Pope has been determined to be of good quality and no further inspections are warranted.

t 1

b 1

4 g +- -.m.- y . - , ,. rv- c--,, . - ~ - y-- - ,,,-,- -. r -w , e, ,,

O D-1 APPENDIX D ENGINEERING EVALUATION OF OBJECTIVE DISCREPANCIES A. INTRODUCTION This appendix has been reformatted from the appendix submitted with the

' January 12,1984, Interim Report. A sample of the objective discrepancies was evaluated for the Interim Report and each of the evaluated dist.repancies was tabulated in the appendia. A!! objective discrepancies have been evaluated for -

this report. The tables which form part of this appendix tabulate discrepancies by type and by method of engineering evaluation.

B. QUANTITY OF OBJECTIVE INSPECTION DISCREPANCIES The Reinspection Program identified 3,247 discrepancies associated with 156,926 objective inspections. Table D-1 summarizes the number of objective inspections ana objective discrepancies for each contractor.

Table D-1 Summary of Objective Discrepancies by Contractor No.of No.of -

Contractor Inspections Discrepancies Blount Brothers 2,390 28 Johnson Controls .7,812 47 Hunter 69,624 684 NISCo 2,792 12 Hatfield Electric 60,245 2,115 Powers-Azco-Pope 8,047 295 Pittsburgh Testing 6.016 66 Peabody Testing 0* N/A TOTAL 156,926 3,247

• Reinspection of Peabody Testing involved only subjective inspections.

D-2 C. CATEGORIZATION OF OBJECTIVE DISCREPANCIES An engineering evaluaticn has been performed for each observed objective discrepancy. The evaluation methods used are divided into three catepries.

The categories and typical types of evaluation methods used in each category are shown below: -

. Category X - Evaluation by comparison witn current design parameters and tolerances.

Perform a comparison of actual component locations to the corresponding design location with applicable installation tolerar.ces to show that the actual locaticas are within tolerance.

Perform a comparison of the actual installation to the designed installation for discrepancies with - minor documentation errors to show that error was limited to the documentation and did not affect the actual installation.

Perform a comparison of actual component dimensions to the corresponding design dimensions with applicable tolerances applied to show that the actual dimensions are within tolerance.

Category Y - Evaluation based on engineecing judgment hy comparison of the discrepancy with design margins.

Perform a comparison of discrepancy to current design analysis or calculations to determine that the discrepancy was not significant.

D-3 Review the component design function to determine that the function of the component was not affected by the discrepancy.

Category Z - Evaluation by engineering calculations.

Revise the existing design documents to incorporate the design change reflected in the discrepancy.

Prepare a specific calculation to address the impact of the discrepancy on the desian.

D. RESULTS OF OBJECTIVE DISCREPANCY EVALUATION The results of the objective discrepancy evaluation for each contractor is summarized in Table D-2.

Table D-2 Summary of Objective' Discrepancy Evaluation Results Category Y Category Z No. No. No.

No.'of Category X Acceptable Acceptable with l Discrepancy No. Within by by Design Contractor Evaluations P_arameters Judgment Calculation Significance Blount Brothers 28 10 8 10 0 Johnson Controls 47 15 19 13 0 Hunter 684 614 52 18 0 NISCo 12 0 12 0 0 Hatfield Electric 1,675 1,243 74 358 0 Powers--Azco-Pope 295 232 5 58 0 Pittsburgh Testing 66 1 9 56 0 Peabody Testing

• N/A N/A N/A N/A N/A TOTAL 2,807** 2,115 179 513 0

• Reinspection of Peabody Testing involved only subjective inspections.

• • In some cases, more than one discrepancy was associated with a componen:. This results in the number of discrepancy evaluations (2,807) shown in Table D-2 being different than the number of discrepancies (3,247) shown in Table D-1.

I l.

I D-4 Table D-2 shows that 2,115 (75 %) of the discrepancies identified in the

' Reinspection Program are not considered valid discrepancies and represent work that is within current design parameters. The discrepancies in Category X result primarily from three factors:

o Minor documentation errors that do not affect the quality of the installed work (e.g., transposed numbers that are clarified by corroborating information on the affected document or other independent documents).

4 e . Tolerances that have been expanded since the time of the original inspection and therefore are within current design limits.

e Conduit as-built reinspection tolerances that were intentionally chosen to oe more restrictive than actual installation tolerances in order to assure that all actual discrepancies were captured by the Reinspection Program.

The discrepancias in Categories Y (179 or 6%) and Z (513 or 18%) represent either:

e An installation that differs from the original design but also meets the design requirement 2 (e.g., the actual location of a conduit support is different from the support location shown on the installation drawing, but the actual location also meats the design requirements).

e Design features that are not required for normal or emergency operation

- (e.g., the diesel . generator cylinder temperature indication at the local control pane: which does not serve either an operating or alarm function but is connected because it is part of the package provided by the vendor).

[

4

& D-5 In summary, the engineering evaluation of objective discrepancies has shown

'that the majority 'of the observed' discrepancies are not vaild discrepancies and that none of the observed discrepancies have design significance.

i; Objective discrepancies associated with weld length have been reviewed and all resulting _ weld strength reductions are within specified design limits.

'With respect to ASME and ANSI B31.1 Code requirements, all discrepancies were

, ' reviewed to their applicable code design criteria and determined to be acceptable.

- A more detailed breakdown of discrepancy evaluation by type of discrepancy is shown for each contractor in Exhibit D-1, which contains Tables DE-1 (Blount Brothers), DE-2 (Johnson Controls),' DE-3 (Hunter), DE-4 (NISCo), DE-3 (Hatfield), DS-6 (Powers-Azco-Pope) and DE-7 (Pittsburgh Testing).

E. CONCLUSION

.' All th'robjective discrepancies were evaluated and determined to be acceptable when compared-with the design requirements and code design criteria. It was

.. determined thatL over 80% of all objective discrepancies were within current

. design and installation parameters or were acceptable by engineering judgment ll (Categories X and Y). The number of.. discrepancies requiring detailed

~

h _ engineering evaluation _(Category Z) is very small when' compared with the total numbar of objective inspections performed. In sommary, the quantity and type of objective discrepancies reported confirmed the good quality of construction l:

at Byron.

^

f '.

s t

L

EXHIBIT D-1 Page 1of 12 EXHIBIT D-1 -

SUMMARY

OF OBJECTIVE DISCREPANCY EVALUATION BY CONTRACTOR Table DF.-l Summary of Oblective Discrepancy Evaluation Blount Brothers Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance Welds 1.- Stiffener installation 2 0 2 0 0

2. Beam end clearance 13 10 3 0 0

3. Connection plate size and location 8 0 0 8 0 Masonry

1. Block walls 2 0 2 0 0 Concrete expansion anchors-

1. Embedment length 1 0 1 0 0

2. Detail Modification 2 0 0 2 0 TOTAL- 28 10 8 10 0 Notes for Table DE-1:

1. The majority of the evaluations in Category X are beam end clearance discrepancies. Tha actual clearances are within revised tolerances which are currently shown on the design drawings.

- 2. The majority of the evaluations in Category Y deal with beam erection tolerances and with block walls. The beam erection tolerances for end clearance were shown to allow for erection clearances and have no effect on the behavior of the completed structure. The beam stiffeners were slightly out of plumb, but this attribute does not affect the capacity of the stiffeners to reinforce the beams.

-Some of the mortar joints on a block wall were 1/4 inch wider than specified.

This will not affect the behavior of the wall.

(Notes for Table DE-1 continued on next page.)

EXHIBIT D-1 1 Page 2 of 12 Notes for Table DE-1 (Cont.) -

3. The majority of evaluations in Category Z deal with connection plates and expansion anchor plates. For example, a connection plate at the end of a small beam was changed from eight inches long to seven inches long. Calculations were made which show that the seven inch plate is adequate. In one case, an expansion anchor plate was used because an embedded plate was located out of tolerance.

Calculations were made which show that the expansion anchor plate detail is adequate to transfer the loads.

Table DE-2 Summary of Objective Discrepancy Evaluation Johnson Controls Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance Visual

1. Support fit-up gap exceeded 12 0 3 4 0

2. Tube track installation 15 -15 0 0 0 discrepancies

3. Base meta!

gouges 3 0 3 0 0

4. Tube track joint 3 0 3 0 0 weld ground flush

5. Support- 11 0 2 9 0 configuration

6. Drawing

- discrepancies 3 0 3 O_ 0 TOTAL 47 15 19 13 0 Notes for Table DE-2:

1.- The Category X discrepancies involved tube track assemblies and gouges in base

- metal. The discrepancies associated with the tube track assemblies involved missing clamps, loose or missing bolts and damaged tube track. These items were the result of temporary disassembly by other contractors to facilitate other construction. The gouges in base metal were small nicks that were the result of removing fireproofing for reinspection access.

EXHIBIT D-1 Page 3 of 12

.2. Most. of the Category Y discrepancies involved fit-up discrepancies and -

cttachment relocations. The fit-up discrepancies resulted in reductions in the length of three-inch welds to two and three-quarter inches. In one case, a plate was warped, causing a fit-up gap, but the warped plate will not affect the

. capacity at the connection. In another case, a hanger was attached to an

. embedded plate rather than to a beam. The embedded plate is more than adequate to transfer the loads.

3. The Category J discrepancies required calculations to assure that the component

_ is adequate.- For two support configuration discrepancies, hangers were attached to beams, but stiffeners were not provided. In both cases, the beams are adequate without adding the stiffeners. In other cases, the weld size should have been increased to account for fit-up gaps. In these cases, calculations were

- performed to show that the existing weld is adequate to transfer the required loads.

Table DE-3

. Summary of Objective Discrepancy Evaluation Hunter Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design

. By Attribute Quantity Parameters by Judgment by Calculation Significance Documentation -

1. Controlled con- 94 94 0 0 0 struction drawings -

not signed 2.. Work process sheets -

(a) Data entered incorrectly 85 85 0 0 0 (b) Boxes not checked 73 73 0 0 0

- 3.- Weld material . 40 40 0 0 0 requisition Date not

[ initiaied/ signed Line outs Write overs Missing marking numbers

4. . Weld record not

- signed 35 35 0 0 0

' 3. . Work process sheet -

i not signed / counter-

' signed 29 29 0 0 0 l-

_~

EXHIBIT D-1 Page 4 of 12 Table DE-3, Cont.

Type of Ca_tegory X Category Y Category Z N::. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance

6. Field inspection report not signed / type 3 26 26 0 0 0 inspection

. performed

7. Concrete expan-sion anchor traveler entries 21 21 0 0 0 not signed

8. Daily surveil-lance not 18 18 0 0 0 signed

9. Discrepancy report not .

Initiated 8 8 0 0 0

10. Nondestructive examination needed to be signed 5 5 0 0 0

11. Engineering entries missing 3 3 0 0 0

12. Hardware

~ documentation 2 2 0 0 0 incorrect

13. Equipment checklist not 1 1 0 0 0 signed

14. Angle size incorrect 1 1 0 0 0 Documentation 441 441 0 0

0 Subtotal

. EXHIBIT D-1 Page 5 of 12 Table DE-3, Cont.

Type of ' Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by 3udgment by Calculation Significance Hardware

1. Support as-built 21 16 3 2 0 dimension / angle not recorded or

' incorrect

2. Support location 16 14 1  ! O or elev. missing or incorrect

3. Hardware configur- 8 5 3 0 0 ation incorrect

4. Hardware substitu- 3 0 3 0 0 tion M4x13 for W4x13

5. Concrete expansion anchor 1 0 0 1 0 1 6.: As-buit'a pipe whip restraint 21 0 21 0 0

, dimension

. incorrect

7. Pipe whip _

restraint hardware 3 0 0 3- 0 discrepancy

' 8.- ~ Pipe' bend ovality 11 0 0 11 0

- greater than 3%

-9. As-built -

isometric 138 133 5 0 0 drawing chain dimensien incorrect location

10. Mechanical joint or joiner. 12 1 11 0 0 discrepancy

EXHIBIT D-1 l Page 6 of 12 !

Table DE-3, Cont. -

Type of Category X Category Y Category Z No. with Discrepancy Total' No. Within .No. Acceptable No. Acceptable Design

- By Attribute Quantity Parameters _by Judgment by Calc"latiori Significance l

11. As-b'uilt hardware 3 0 3 0 0

. missing

12. As-built hardware -

2 1 1 0 0 <

. malfunction

13. As-built iscmetric 1 0 1 0 0 drawing configura-tion incorrect i location

' 14. Pipe whip :

. restraint incorrect weld 3 3 0 0 0 Hardware Subtotal 2,43 .g 5,2 J,,8 g

- TOTAL 684 614 52 !8 0 Notes for Table DE-3:

1. Most of the documentation discrepancies. in Category X consisted of minor data entry errors 'and omissions on work reports and process sheets. Because of other

, corroborating information on the affected documents and other independent documents, these discrepancies were determined to have no design significance.

. Hardware discrepancies in Category X consisted primarily of minor as-built F - measurement discrepancies or -incomplete 'as-built information. Many of the

. reported discrepancies for isometric chain dimensions were less than one inch off, which falls within the established accuracy tolerances for field measurement.

, , The remainder were.within specified design / installation tolerances. Since most chain l dimensions involve a fitting or elbow as a reference point a ' slight

--inaccuracy in measurements is understandable. ~ In summary, Category X items L were eva!uated by comparison with the current design /installatiors tolerances and L

. parameters ami determined to be acceptable.

2.~  : The Category Y discrepancies were compared to.the current design calculation.

. and/or analysis and determined not to .be significant. Items involving missing or damaged components that occurred after the first inspection were also reviewed.

e Each of ~these-items-was also determined not to be significant. Detection by -

- other inspection programs (preoperational tests, hydrostatic tests, Type 4 final hanger inspections, etc.) would have corrected these discrepancies.

~

U - 3. . Category Z discrepancies were resolved by preparing calculations to resolve the

-discrepancies. Calculations for'll pipe bends with ovality in excess of 8% were l-

-4 . performed as required by ASME Code. Pipe minimum wall thickness and flow

_ 1 area reduction were evaluated. Stress intensification effects were evaluated as negligible since all of the pipe bends are five pipe diameters in radius.

s EXHIBIT D-1 Page 7 of 12 Table DE-4 -

Su6amary of Objective Discrepancy Evaluation NISCo Type of Ca:egory X Category Y _ Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute - Quantity Paramcters by Judgment by Calculation Significance Mechanical .

- 1. Length of stitch welds on shim plates 12 0 12 0 0

. TOTAL 12 0 12 0 0 Note for Table DE-4:

The ditcrepancies in Category Y cover minor variations (approximately 1/4") in the length of stitch welds on shim plates. These are non-load bearing welds and were, therefore, considered to be acceptable.

Table DE-5 Summan of Objective Discrepancy'~Eve.luation Hatfield Electric Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance Conduit installation

1. Supports 3 1 2 0 0 2.- Auxiliary steel 1 0 0 l 0

3. Conduit 42 7 23 12 0

4. Junction box 13 3 10 0 0

5. Other. 7 0 7 0 0 Cable Termination

1. - . Workmanship 7 0 7 0 0

2. Wiring 2 0 2 0 0

. Cable Pan Hanger

1. Configuration change 3 2 4 2 0

2. Detail substitution 15 1 13 1 0 l

EXHIBIT D-1 Page 8 of 12 Table DE-5, Cont. -

Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance

3. Dimensions out 18 1 5 12 0 of tolerance

4. Other 1 1 0 0 0 Conduit As-Builts

1. Supports 972 813 0 159 0

2. Auxiliary steel 14 1 0 13 0

3. Conduit 298 178 0 120 0

4. Junction box 247 209 0 38 0

5. Other 27- 26 1 0 0 TOTAL 1675 1243 74 358 0 Notes for Table DE-5:

1. The majority of the evaluations in Category X are 1s-built location dimensions of conduit, conduit supports and junction boxes that are within the 6-inch installation _ tolerance. These were identified in the Reinspection Program because the reinspector was using a 3-inch tolerance to ensure that all potential discrepancies were identified.

The evaluations in Category Y cover such items as emunding discontinuities,

~

'2.

missing insulated throats in conduit fittings, cable pan hs - detail substitutions and in one case a missing junction box barrier. Alternate g ,und paths existed for all grounding discontinuities. The insulated throats while providing an extra measure of cable protection are not necessary. The substituted details were similar in member size and capacity and varied only in their attachment -

configuration. The missing barrier was not separating cables of different divisions but rather power and control cables of the same division. The cables are in fact separated without the barrier.

3. The evaluations in Category Z consist primarily of as-built location dimensions of conduit, conduit supports and junction boxes and other minor dimensional differences that exceeded installation tolerances. Generally, the dimensional difference was limited to a few inches. The installatica drawings and supporting calculations have been revised to show the actual dimensions as standard practice and the revision confirms that the actual dimensions are acceptable.

4. Discrepancy HE-129 covers a termination error for cable IVA075. This discropancy would result in damper OVA 052YA not closing automatically on the start of charcoal booster fan OE which starts automatically on a safety injection

EXHIBIT D-1 Page 9 of 12 signal. Operation of the damper is required in the event that radioactive 1

' effluents are present in the auxiliary building exhaust air. Radiation monitors in L the exhaust plenum alarm in the control room and. damper OVA 052YA can be closed ~ manually through a ccontrol switch in _the main control room. Byron operating procedure BOA-RAD-1 requires that the operator check the damper position in response to the alarm. Therefore, the termination error is judged to be not significant. However, the FSAR description of the auxiliary building

exhaust systems states that the damper operates either auto

7.atically on a safety

-injection signal or manually through a control switch in the main control room. If undetected, the design would have differed from the FSAR description, but it would not affect the safe operation of the plant. The termination error did not go undetected. ' The inspection of this design feature by the electrical contractor is only the first of three checks made to ensure the design functions prcperly. The ,

discrepancy was actually dt.tected and corrected independent of the Reinspection Program during construction testing, and the interlock would have been tested as part of the auxiliary buildmg ventilation system (VA) pre-operational test.- It is also worth noting that only two wiring errors were identified out of the 7,784

terminations reviewed in the reinspection program. The other involved the diesel generator cylinder temperature indication at the local control panel which does  :

- not serve either an operating or alarm function.

-3. The -Reinspection . Program provided a small sample for Hatfield Electric's equipment setting and modifications,~ A325 bolting and conduit support boltir.g work. . Additional inspections are being undertaken. for - these objective

, attributes. It should be noted that QC inspections for all objective attributes require similar skills and training. The Hatfield inspectors who were reinspected did e_xceptionally well on other objective attributes. Therefore the effectiveness

,. of their QC inspections is not in question. The additional work is being performed

'_ ' to complete _ the data base. The inspection plan being implemented is as follows:

ei Equipment Setting and Equipment Modifications To confirm that the safety-related electrical equipment that was installed, mounted, or modified in the field by Hatfield Electric is installed properly,

. the . total popul_ation of approximately 250 of such items has been identified,'

and a~ program for reinspection of a sample' size of 50 has been selected for

. reinspection in accordance with Mil. Std.105D single sampling plan. In case .

the success criterion is not met for tht. selected sample, the total population F' will be reinspected.

V . .

e ' A325 Bolting

The Reinspection Program in the area of A325 bolting resulted in a total p

, sample population "of 8. Although no discrepancies were noted, further reinspections are being performed to . increase the confidence level. The total population of such items has oeen determined to be approximately 170,

'and a sample of 50 has been selected for reinspection in accordance with Mil.

r Std.~ 105D single sampling plan. In case the success criterion is not met for i the selected sample, the total population will be reinspected..

~

L  !

L  !

p _

L y

[

_= .- _. . . _ _ _ _ - _ - . _ . _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ -

EXHIBIT D-1 Page 10 of 12 o- Conduit Support Bolting The Reinspection Program did not include checking the torque level on the conduit support bolting. This was considered acceptable inasmuch as the Program was designed to validate the qualifications of the inspectors rather than the quality of construction. However, in view of the specific apparent concerns expressed by the ASLB, a reinspection of this attribute has also been included. The total population of this attribute has been determined to be approximately 25,000. The reinspection sampling will be in accordance with Mil. Std.105D multiple sampling plan. The initial sample size is 125.

. Additional samples of 125 will be inspected if required.

Table DE-6 Summary of Objective Discrepari y Evaluation Powers-Azco-Pope Type of Category X Catagory Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design By Attribute Quantity Parameters by Judgment by Calculation Significance Pipe Material ~

Verification

1. - Incorrect 19 19 0 0 0 quantity on piping bill of material

2. Incorrect size on piping bill of 3 3 0 0 0 material

3. Incorrect heat 28 28 0 0 0 number on bill of material Hanger Material Verification

1. Incorrect quantity 34 34 0 0 0 on hanger drawing bill of material ' i

2. Inccrrect size 45 37 2 6 0 of hanger drawing i bill of material

3. ncorrect heat 84 84 0 0 0  ;

number on hanger

- drawing Final Hanger

1. Incomplete weld 55 22 2 31 0 length

L EXHIBIT D-1 Page 11 of 12 Table DE-6, Cont.

Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design

-By Attribute Quantity Parameters by Judgment by Calculation Significance

2. _ Hanger config-uration 18 2 1 15 0

- 3. Incomplete weld 5 1 0 4 0 length and hanger configuration Flex He<e 2 2 0 0 0 Pipe Bend 2 0 0 2 0 TOTAL 295 .232 5 58 0 Notes for Table DE-6:

1. Category X documentation discrepancies consisted of minor data entry errors and omissions on work - reports- and process sheets (transposed numbers, incorrect quantities, etc.). Because of other corroborating information on the affected documents and other independent documents, these discrepancies were determined to have no design sigmficance. Most of the hanger discrepancies involved items such as minimum weld lengths which are within tolerance when compared to current design tolerance.

2. The Category Y discrepancies generally involved material substitutions such as substituting a 3/8-inch thick plate for a 1/2-inch plate.

I

.3 Calculations were performed to evaluate the Category Z discrepancies. Two Class H pipe bends exhibiting ovality in excess of 8% were evaluated as

! required by ANSI B31.1 code. Pipe minimum wall thickness and flow area reduction were also evaluated. Stress intensification effects were evaluated as negligible since all of the pipe bends are 5 pipe diameters in radius.

The hanger discrepancies involved weld configuration chc1ges and material substitutions. Calculations were performed to show that thesa substitutions and configurations resulted in supports that are adequate to transfer the required loads.

I'~

EXMIBIT D-1 Page 12 of 12 Table DE-7 Summary of Objective Discrepancy Evaluation Pittsburgh Testina Laboratories j Type of Category X Category Y Category Z No. with Discrepancy Total No. Within No. Acceptable No. Acceptable Design

,By Attribute Quantity Parameters by 3udgmant by Calculation Significance Concrete Expan- l sion Anchors

1. Embedment length 17 0 0 17 0

2. Spacing and edge distance 16 0 2 14 0

3. Angularity, insufficient tread engagement 6 1 1 4 0 4 Enlarged holes 24 0 6 18 0

5. Insufficient torque- 2 0 0 2 0

6. Substitution of weld to embedment for concrete expansion anchor 1 0 0 1 0 TOTAL 66 1 9 56 0 Notes for Table DE 7:

1. The Category X discrepancy involved a bend in the anchor bolt. The bend was above the nut (i.e., in the part of the bolt that carries no load).

2. .The majority of the Category Y discrepancies involved expansion anchor assemblies where the loads were small compared to the design capacity of the assemblies. The discrepancies involved such items as nuts engaged slightly over 80% of the thread engagement, edge distances redeced by less than 20%, and plate washers for enlarged holes which did not completely cover the enlarged holes all were found acceptable.

3. The Category Z discrepancies were resolved by making calculations to show that the conditions found during the reinspection are adequate. For example, where an expansion anchor embedded length has been reduced from 5 inches to 4-5/8 inches, the capacity of the anchor i:: reduced proportionately and compared to the actual design loads.

t E-1 APPENDIX E MANAGEMENT APPROACH TO ENSURE QUALITY OF CONSTRtfCM

~A. GENERAL The Commonwealth Edison efforts during the course of the Byron Project have been actively, energetically, and intensely performed to ferret out problems, obtain corrective action, prevent recurrence of problems, and assure that the proper engineering information was used in generating the design documents, that the proper design documents were used in the construction of the plant, that construction was performed to approved procedures and is in accordance with the design documents, and that the inspection and testing of the facilities were performed, reviewed, and accepted by properly qualified personnel. These verification efforts have been carried out with multiple tiers of inspections, surveillances, and audits by contractors, Commonwealth Edison's Quality Assurance Department (QAD), and other outside organizations, such as the Nuclear Regulatory Commission (NRC) and the American Society of Mechanical Engineers (ASME), plus consultants (such as Energy Inc.), who have performed management audits of Commonwealth Edison's QA Department every other year to evaluate the implementation and effectiveness of the QA Program in Commonwealth Edison (Table E-1).

l

E-2 Table E-1 Quality Assurance Coverage of Commonwealth Edison Construction Projects

, - CONSTRUCTION SITES DESIGN ACTIVITIES OTHER

1. Contractors 1. Architect Engineers 1. Corperate QA Inspectiens (QC) Audits (QA) Vendor Audits Audits (QA) Other Other Commonwealth Independent Review Edison organizations System Design Surveillances Overall Design Audits

2. Commonwealth Edison .2. Commonwealth Edison 2. QA Coordinator Ouality Assurance Quality Assurance Engineering and Testing Site QA Site Design Activities Surveillances inspections Site QA Audits Independent Testing Surveillances

- Agency (documented continuously)

Audits Surveillances Corporate (documented Audits continuously- Management Audit -

800/yr) Outside Consultant Audits Corporate Outside Design Activities Audits - by Corporate QA Corporate QA of Audits of AE Organizations all site (Supplemented Team with organizations design Engineers having Management audit - special design expertise) by outside Consultants Commonwealth Edison is ultimately responsible for the assurance of quality in all phases of the design, procurement, construction, and preservice testing of

' the Byron plant. The Commonwealth Edison QA Program, which has been approved by the NRC and fcund acceptable by the ASME, addresses the requirements of the 18 Criteria of Appendix B to 10CFR50 and Section III of

m E-3 the ASME Boiler and Pressure : Vessel Code as well as the applicable American -

National Standards institute (ANSI) QA standards, in addition to NRC and other federal and state regulatory requirements.

Inl order to ensure that Commonwealth Edison QA requirements are properly communicated.to the contractors, standard QA articles are included in the

. contract specifications. These QA articles delineate the requirements to be

. imposed on the contractor during execution of the work. The articles are

, written to' include the requirements of Code of Federal Regulations 10CFR50 Appendix B and the requirements of the Commonwealth Edison QA Program.

This up-front approach establishes that the contractot s fully understand the QA b ,

requirements that must be-complied with by wr.ractors when working on

• Commonwealth Edison projects. The contractor is required to translate the f requirements of the quality articles into commitments in the contra:: tor's QA

manual and implementing precedures.

The QA programs of the various vendors and contractors involved with ASME

code- and safety-related items and activities are also required to adhere to me requirements of the applicable 10CFR50 Appendix B criteria. Further, the

$ - Commonwealth Edison QA program requires that its QA personnel review and ,

cccept= all such vendor and contractor QA programs (including those of the architect-engineer and the NSSS vendor) prior to award. .

Y  ! The contractor's QA manual is reviewed to ensure that all the necessary quality commitments are clearly stated in the contractor's QA Program.

~ After award of a' contract, the contractor is required to develop and implement work procedures to carry out-the commitments as described in the approved contractor QA manual and the applicable specification. All procedures used in safety-related work by a site contractor must be submitted to Commonwealth

= Edison for review and approval.- Procedures are required to be approved by Commonwealth Edison before the work activity described in the contractor

~

procedure is allowed to begin. Procedures are reviewed by Commonwealth 4

h[

_. .3.~., . . . .

t E-4 Edison's Construction, Engineering, and QA Departments to assure compliance j to the ' Commonwealth Edison QA manual, contractor QA manual, work )

specification, FSAR commitments, and applicable codes and standards. The i review consists of both a technical and quality assurance review. During the duration of the work, any revisions made to implementing procedures must be submitted to Commonwealth Edison for review and approvalin a like manner.

Commonwealth Edison's Manager of QA directr, the quality assurance activities

- related to the design, procurement, construction, and operation of the Commonwealth Edison's nuclear plants. Commonwealth Edison's QA organization is completely independent of the Engineering, Construction, and Operating organizati;ns, is responsible for identifying quality problems, and reports directly to senior management. Since the Commonwealth Edison QA organization was established ic: 1173, the Manager of QA has successively reported to an Ev<utive Vice President, a former President, and currently to Comraonwealth Edison's Vice Chairman, all of whom have given, and are giving, their full support to the QA Program.

Commonwealth Edison's QA organizations consist of graduate engineers from essentially all nuclear-related engineering disciplines and non-graduan:s with years of hands-on experience such as maintenance, construction, engineering, and operating Navy nuclear and other technical activities. Having qualified personnel .is a basic requirement for a QA organization. To ensure that qualified personnel have adequate knowledge to perform their work and maintain their proficiency on such quality-related items and activities cs codes, standards, materials testing, nondestructive examination, welding, auditing, inspection, coating, and fire protection, formal training and refresher training 1

- are routinely given. Generally, the training of QA personnel is an extension of l their formal education or work experience. In addition, the Commonwealth

- Edison QA personnel are required to be qualified and, in turn, certified for specific functions involved with inspection, surveillance,'and audit activities, l including those for which training is provided. They are also required to I adequately demonstrate their qualifications to perform specific functions

E-3  :

Lbefore they are allowed to perform any QA activities for which they are qualified.- Furtnermore, Commonwealth Edison QA personnel are required to be recertified on a 2-' to 3-year cycle.

Qualified- QA: personnel are assigned to .the nuclear construction sites and

. operating stations and Commonwealth Edison's nuclear testing and engineering organizations. - A QA Coordinator, who reports to the cognizant Director of QA (Engineering-Construction), has been assigned full time, since 1975, to the ,

Commonwealth Edison engineering organization which has overall design responsibility for'the nuclear' stations. The QA Coordinator reviews procurement documents, specifications, 2nd procedures and performs periodic audits and surveillances of. project ' activities to verify compliance to the Commonwealth Edison QA program and applicable procedure requirements.

Each Commonwealth . Edison corporate department having nuclear related responsibilities, such as engineering, purchasing, and testing, is annually audited to ithe requirements of the - QA program, their respective department

~

procedures, and other - requirements by Corporate 'QA ' using qualified and certified Le'ad Auditors and audit personnel who are not directly involved with the organization rseing audited. The nuclear construction sites are also audited by Corporate QA~on about a 4- to 6-month cycle. 'In addition, audits of other construction site organizations and cor tractors are performed . by g _

Commonwealth Edison,QA in accordan~c e with an approved schedule ranging

. from monthly. to -annually depending on need or ~ established requirements.

ildentified deficiencies are - required to be committed to be corrected or-

' dispositioned in a timely manner. Audit deficiencies involving Commonwealth Edison departments that are not corrected or dispositioned in a timely manner are referred to top corporate management for their handling. Where serious problems or inaction to solve problems are identified, stop work action is initiated.

4-<

l

rt I

, ,,.-,-,,w, , ,,----.--,.,,..-m

E-6

~

- The Manager of QA has the responsibility and authority to stop unsatisfactory work or to stop further processing of unsatisfactory material during design and construction of the plant. Conditions found by quality personnel that are adverse to quality and require prompt action and corrective action but that cannot be resolved at the respective ;ocations are required to be promptly

. reported to the Manager of QA or his designated alternate for action.

Furthermore, resumption of work under a stop work order requires concurrence of the Mareger of QA. There have been nine' cases of stop work involving significant concerns at Byron plus two at a vendor plant due to QA program implementation and welding problems ranging from a very short time to many weeks. The extended stop work cases, for the most part, involved administrative, training, inspection and/or procedure problems which had to be corrected before work activities could proceed.- Currently, at the Byron site, a stop work order - is in effect against Reliable Sheet Metal, the HVAC contractor. The corrective actions by ReHable Sheet Metal consisted of a complete reinspection of all installed work, correction of all noted deficiencies,

and acceptance of 'all reinspected items prior to release for new work to proceed.

.The Commonwealth Edison . construction QA organization functions independently of the Commonwealth Edison Construction site organizations

- under the direction of the site QA Superintendent. He reports offsite to a QA Director (Engineering-Construction) who reports to the Manager of QA. The Byron site construction QA group has expanded over the years to keep pace -

with the construction activities and is augmented with an Independent Testing

. Agency contractor.

l Also, .the Commonwealth Edison Operations QA group was established early in the project -to cover plant operation activities at the Byron Station. It has overall responsibility to ensure that the activities and functions performed by the station staff are carried out, as required, during the formulation, preoperational testing, and startup stages of the project as well as during plant operation after issuance of the plant operating license. The Operations QA

E-7 group consists of seven or eight individuals who perform inspections, surveillances, and audits of plant operation activities such as mMntenance, procurement, preoperational testing, piant start-up activities, and associated adininistrative and other functions related to plant operations. Special emphases were placed on performing surveillances and audits of each system preoperational test to ensure requirements were fulfihed.

B. ON-SITE QUALITY ASSURANCE During the construction phase of the v;ork, many levels of checking, review, and evaluation are periormed of contractor work activities to ensuie that the construction work is correctly completed and is of high quality. The methodology used by Commonwealth Edison QA to achieve an appropriate confidence level regarding adequacy and accuracy of the construction includes inspections, surveillances and audits (Table-E-1). This approach has resulted in examining,-in great detail, a wide range of selecteo items and areas, including those reported to be of concern at other project sites such as troubles with vendors or contractors, construction, equipment and operatmg problems.

. failures relating to implementation of QA Programs, design and design control problems, welding, nondestructive examination and inspection deficiencies. A brief description - of each - of the checks which comprised the iaulti-layer coverage imposed on the construction activities at Byron is as follows:

1. . Audits The Commonwealth Edison constrt.ction QA group at Byron monitors quality related construction activities through audits and surveillances.

Commonwealth Edison has implemented a comprehensive pragram of audits covering the contractors performing safety related work at Byron.

Audits are conducted to approved audit schedules and to approved checklists. The audits include verification that the contractors' QA programs and work procedures are properly implemented. Also, sizeable portions of the fieid installations are reviewed. At a minimum,

= - -

E-8

' : Commonwealth Edison audits each contractor quarterly, and in many cases audits a contractor with suspected problems on a monthly basis.

'u - The scope of the audits conducted through the course of the project cover not only -the eighteen criteria.of 10CFR50, Ap;endix B such as design control, document ' control, inspections,~ tests, personnel qualification,

- corrective action and audits,-but other areas such as ANSI Standards, ASME Code, FSAR, applicable NRC and other governmental regulations,

. the requirements or the Commonwealth Edison QA program and the -

respective contractors' QA programs, procedures, and plant installation

. adherence to specifications,-drawings, and procedures requirements.

~

To date, the Site QA group has performed about 500 audits of contractor activities at an average audit interval of three months. .

- The audit process provides for identification of problems, obtaining

_ -appropriate corrective action including. action to prevent recurrence,

- ' follow-up and' close-out by the QA groups. This audit program has been H -effectively executed during - the course of the project and has been i

p  :

instrumental in the identification of problems at Byron.

p P 2. Surveillances I In addition to audits, the Commonwealth Edison Site QA group performs

• regular surveillances of construction activitien To date, approximt,tely 5,550 surveillancas have been perfumed. Surveillances are less formal

[~

. than: audits in that they are generally not conducted in accordance with 1

approved checklists but instead are conducted to. verify ' adherence to :

- procedures, codes,. standards, etc. These documented surveillances are

- performed at a greater frequency and independent of audits although they j cover .the same types of activities as audits.- The surveillances cover 2

, broad range of contractor activities.

M l

s. \

l I T i-I

_ J ' '

, ., l 4 '*-

wr=r a+-p ,w g.y,wwg.. 4,.,y _.g_.,, e ,y,,4 y,g ,p . , . , . , , ,, ,, , _ _ _ ,

h 3 E-9 Surveillances are fundamental in assurir.g contractor compliance to his -

work procedures. Examples of areas observed during surveillances are as

. follows:

e. Storage of materials and equipment prior to installation.

e Housekeeping in the area of work to preclude potentially detrimental conditions.

e Welder qualifications review to assure that welders are qualified to do the welding performed by them.

e QC inspector qualification review to assure, as with welders, that QC inspectors are properly qualified to perform inspections.

e Weld rod control review to assure that only approved weld materials 1 are used for the activity being performed.

e _ Drawing control to assure that only the last approved drawings are used for construction.-

e Record storage to verify proper controls are in place to assure records will not be lost, destroyed, etc.

e -- Preservation of equipment verification by surveillance to assure insta!!ed equipment is not damaged.

e -Nonconforming conditions review to verify that the cor. tractor identifies conditions adverse to quality utilizing the appropriate

- deficiency reporting document.

l 3 -.

k 9 -,-g g w , -,c----,.e t=-e -

I' E-10 e- Field verification that contractors are installing and inspecting items -

in accordance with the design documents. These surveillances are

' performed both in-process and after completion of contractor activities.

Deficiencies identified during surveillances must be corrected in a timely manner or they are elevated to audit deficiencies.

- 3. Contractor QA Audits Each site contractor also performs inspections and audits of their work activities in accordance with their own QA programs, which are approved by Commonwealth Edison. The contractor is required to have an independent QA auditing function. This means that the contractor must have, -within his organization, qualified raiditors who audit the site activities on a regular basis in accordance with our approved schedule.

- Audits are performed from the early stages of construction and de continued till project completion. Contractors are required to audit all aspects of their QA program including implementation of procedures. To date, the site contractors have performed approximately 1,200 audits of their construction and quality control inspection activities. The performance of such activities are monitored by Commonwealth Edison QA to assure adequate implementation by each contractor.

4. Contractor QC Inspections All contractors performing rafety-related work at Byron are required to

[ have certified QC inspectors. These inspectors are qualified to ANSI N45.2-6 (1978). At present, there are a total of about 250 QC inspector:

employed by contractors at the Byron Site. Site contractors at Byron are j l

responsible for acceptance inspection of work performed. This includes j receiving inspection of material and equipment purchased by others but i l

installed by the contractor. Also, the contractor performs all first line l inspection for acceptance with the exception of those inspections for l which responsibility has been delegated to me independent inspection I

1

_ _ . . , _ - _ . _-- i

l l

l E-11 agency (NDE and concrete expansion anchor inspections are examples of -

inspections delegated to the independent inspection agency). . In addition to first. line inspection, the contractor is responsible for final walkdown inspection which includes documenting the as-built condition of the installations for which the contractor has responsibility. Detailed iriformation on the coverage of contractor QC inspection activities is provided in Exhibits E-1 through E-9.

5. Inspections by Independent Testing Agency

'An independent testing agency reporting to the Commonwealth Edison Site QA Department, performs independent inspections, destructive testing and nondestructive testing of many of the key work activities of the site contractors. Site QA also uses the independent testing agency to perform special inspections as deemed prudent or necessary to check constructior:

work performed and inspected by the contractors. The Byron independent testing centractor.(Pittsburgh Testing Laboratory) currently has about 65 inspectors and technicians performing this work. This independent

- inspection / testing activity is the " official acceptance" required for selected key work. For example, the independent testing agency performs the testing required to accept concrete, reinforcing bars, cadwelds, soil composition, concrete expansion anchor instal!ctions, structural bolting, etc. In addition, ti.ey perform the radiogreAic testing required to accept welding and most of the magnetic-particle, liquid penetrant and ultrasonic acceptance testing.

, ~ , , , ~ . . , , , , ,,-

e--,- - , . . - - y .--

% ,, . _ _ 3

l 1

l E-12 l

Table E-2 l Byron Station Independent testing Agency Testing 9

and Inspes. tion l

Test and Inspection Area soils Concrete Aggregate Grout Cadwelds 55,000 Inspection Reports (approx.)

Concrete Expansion Anchors Bolting Coatings Shear Studs Non-Destructive Examinations Magnetic Particle Liquid penetrant 44,000 Inspection Reports (approx.)

Radiography Ultrasonie

6. Independent Overinspections In Mdition to the audits and surveillances conducted by Commonwealth

. Edison and site contractor QA groups, Commonwealth Edison Site QA has directed the independent testing agency to perform "overinspections" of certain portions of the contractor's work. These overinspections are irl addition to the QC inspections required to be done by the site contractor.

The purpose .of these overinspections is to give Commonwealth Edison another level of confidence that the field work and the inspection activities havs. been done properly. These independent overinspections have been in effect since about 1980 and generally cover about 10% of a work activity and have been concentrated in the areas of welding, electrical installations and HVAC installations. In addition, this Agency

s E-13

- has performed over-check surveillances of many contractor activities-in the structural, mechanical and electrical disciplines for Commonwer.lth Edison Site QA.

- 7. Reinspections t

Another level of dssurance of quality is achieved through reinspections. A few contractors -have found it necessary to reinspect their own work.

Reasons vary from identification of questionable or improper workmanship to- insufficient documentation of the original QC . inspections. Such reinspections result . in having both the questionable . work cnd the acceptable work rechec?ced for a second time. Detailed information on the moverage of the construction reinspection work is provided in Exhibits E-1 through E-9.

8. Project Construction Surveillances The' Commonwealth Edison Project Construction Department also has

, responsibility for performing surveillances of all construction activities.

F These surveillances cover such activities as plant installations work inspections, materials and equipment handling, safety, fire protection, construction ~ tests, . welding, - housekeeping, storage of equipment and materials, concrete pours, batch plant operation, etc.

9. ASME Surveys Also, the~ American Society of Mecnanical Engineers (ASME),-starting in

-197te, conducted surveys of Commonwealth Edison, to assure that the Commc,nwealth Edison Program is correctly and properly implemented.

ASME conducted surveys of the Commonwealth Edison corporate offices in 1974,1977,1980, and 1983 which resulted in ASME awarding Engineering

(. Organization and "N" Certificates plus NA and NPT Certificates to Commonwealth Edison that grant it authority -to desi,p,' fabricate, and 8

- e r-e - ,. . v3e.. , , , , ,,e - - - - , , - - - - ,-, . , . - , ~ -, , - - , _e,,-,,n

- , . , y----m,-~~ y-m, ,- . - - - - _ , ,-,--,-,-v

p E-14 s

install items that must meet ASME Code requirements. In all cases the ASME Survey Teams reported that the procedures and management controls were sufficiently effective to implement the Commonwealth Edison QA program. Also, ASME survey teams have been on-site to aud't Hunter and Powers-Azco-Pope.

L

10. Unit Concept In pection

[-

In September 1982, another form of inspection was initiated at the Byron Site by Commonwealth Edison QA. This new inspection is called " unit concept inspection" (UCI) and is in addition to all other forms of inspection already in effect. The UCI was designed to give another level of confidence that the essentially completed plant elements are constructed properly. These UCIs are performed using a team of six inspectors from

, the ' Independent Testing Agency, who are qualified in the various disciplines per ANSI N45.2.6 (1978).

All items installed within specific spatial boundaries or in conjunction with specific equipment are inspected for compliance to vendor and engineering design documents. The UCI group inspects a different area or " unit" on a

' weekly basis. This inspection encompasses all contractors who performed work activities witbin a given area. As of February 1,1984,68 UCis have L

been completed. The UCis are in addition to the normally assigned inspections and the specific overinspections performed by the testing agency as directed by Site QA.

The inspection performed by this group is an independent reinspection of-work and inspections performed by the contractors during plant construction and a verification. of the acceptability of manufactured j

equipment-installed in the plant. The purpose of the UCI is to perform a  !

complete check and reinspection of all items of an element of the plant l each week (such as all items in a volumetric space within four buildmg '

' columns and the floor and ceiling of the plant, or a specific item), against i the fint.1 design and vendor. drawings to provide Commonwealth Edison QA I 1

l l1

- - . e , . - -

. . . . ~ . . - - . . . . .- .

E-13 v 4 an additional level of assurance of the construction quality of thefacilities and . equipment : in the pla'nt at the essentially completeu stage of installation.-

The extensive nature .of' the inspection and the percent of total plant

. components inspected can be measured by the percentage of the-total Byron. plant . volume inspected as part of UCI. This has been done by comparing the' volume of the plant inspected to the volume (ft3 ) of the

. plant that -has inspectable items contained in it. In - the containment

_ building, approximately.13% (147,000 cubic feet out of 1,155,000 cubic feet) of' the volume 'was inspected. Much of the critical equipment is ,

contained inside the missile barrier wall. For the portion of the containment between the missile barrier and the primary shield wall, 4

where part of ' the critical equipment is ' contained, approximately 26%

(51,000 cubic-feet out of 196,000 cubic feet) was inspected. Examples of the items inspected within'the missile barrier wall as part of the UCI are

~t he reactor coolant pumps, steam generators, pressurizer, primary loop ,

whip. restraints, and jet deflectors. - Other equipment inspected inside b containment includes the in-core monitoring system, reactor containment fan coolers, an' d the charcoal filters.

+ -

.The UCIs of the auxiliary building were conducted in a fashion similar to

those conducted in the containment building. The inspections focused on major components important to operation and safety.- Examples of major components inspected in the auxiliary building are the diesel generator, the

residual heat removal pumps,' essential service water pumps, containment spray pumps, safety injection pumps, centrifugal charging pumps, remote

. shutdown control' cabinets, motor - driven pumps, auxiliary feedwater l pumps, 4160V switchgear, cable spreading rocms, and control room HVAC p _

systems. Approximately 16 % (391,000 cubic feet out of 2,450,000 cubic p - . feet) of the auxiliary building was inspected on a volumetric basis, i<

~

l j --  !

L a

c=

t E-16 i.

Some of the UCIs are not included in the volume assessment described above because volumes were difficult to define outside of the containment or auxiliary buildings. The more significant of the items delineated on this list are the main control board, mechanical, electrical, and HVAC

. penetrations, and the Halon fire protection systems.

t in addition to the large total . volume inspected,' the inspections were-sufficiently distributed throughout the plant to encompass a wide spectrum of ' electrical, . mechanical, HVAC, and structural items. The UCI categories included 30 mechanical, 18 electrical, 11 HVAC, and 9

. structural categories.- These are listed in Exhibit E-10.

7_ ,

= The 68 Byron UCIs performed as of February 1,1984, have inspected and evaluated the following approximate quantities of installed items:

l e 101,300 square feet of structural concrete and masonry block wall e 2,500 structural items e 3,200 square feet of special coatings e 16,700 mechanical ite:ns e -- 5,300 linear feet of piping and insulation e 16,500 electrical items

e. 1,500 sections of cable pan and conduit e 900 HVAC items e 1,100 -linear feet of HVAC duct installation

.Because of the extensive nature of the UCI, both in terms of plant volume -

and categories -.of items inspected, and . the total number of items inspected, the results o? the program are a good measure of the quality of

. construction at the Byron pl ant. l l

An analysis of the results of the 68 UCis indicate that no generic trends exist relevant to the quality of the installations or the QC inspections.

Additionally, all deficiencies identified in the UCIs are tracked until corrected or evaluated to assure conformance to the design documents.

l

a ,-,,

E-17 Based upon the positive results of the UCIs, Commonwealth Edison QA has a high level of confidence that the contractors are performing installation and inspection activities in an acceptable manner, thus assuring that the

. plant is constructed in accordance with design requirements.

11. Other Audits In addition to the audits performed by the Site and Station QA groups at Byron, extensive and comprehensive audits are performed by General Office QA using qualified QA auditors not directly involved with the Byron Project QA activities. These large off-site audits,' which are performed two or three times each year, cover all contractors on site. If there are concerns about a special subject (such as small bore pipe, whip restraints, or QC document verification), then the general office audit is designed to be very specific in order to thoroughly review and analyze the specific subject relative. to documentation, design and installation requirements. A major advantage of the general office audit is that it brings qualified personnel from other locations such as LaSalle or Zion to indepenc'ently examine work results and work activities being performed by Byron for acceptability. . .

Also, a 25-man team of off-site senior Commonwealth Edison personnel

. and QA auditors with management, design, and construction expertise performed an evaluation to the performance objectives and criteria for construction sites that were established by the Institute of Nuclear Power Operations (INPO).

In addition, a management audit is performed every other year of the implementation .of the QA program and other requirements, by an

- independent outside the Commonwealth Edison consultant organization, covering all nuclear-associated activities in Commonwealth Edison, all of which provides additional assurance relative to construction quality.

, , . , , , . . . ., - - m .-

I h

E-18 As a result of the Management Audit and Assessment performed in 1981, it was -the assessment of the Energy Incorporated Audit team that Commonwealth Edison's QA Department and the Commonwealth Edison

- organizations and contractors demonstrated a high level of capability in implementing the Commonwealth Edison QA Program. It was current with respect to the applicable regulatory requirements and codes; and it _was effective as evidenced by its implementation, such as the timely responses and corrective actions . taken- by the audited organizations and the recognition and understanding of the program requirements by personnel outside the QA Department.

~

The most recent independent Management Audit completed at Byron in June 1983 stated, "the Commonwealth Edison QA audit program is a very comprehensive program that is continually being improved."

C.- ~ CONCLUSION Commonwealth Edison QA is quite confident that the ASME code-related safety-related aspects of the Byron Plant are acceptably designed and built.

t

,-..e .. . -

- ~

't Y

EXHIBIT E-1 Page 1 of 3 EXHIBIT E-1 -

BLOUNT BROTHERS CORPORATION 6

Blount Brothers Corporation (Biount Brothers) has been on site since December 1975

. to execute a contract in accordance with Sargent & Lundy Specification F/L-2722.

F The scope .of work performed by Blount Brothers includes most major civil work including concrete masonry, . installation of post tensioning _ tendons, miscellaneous structural ' steel, and fireproofing. Detailed information,on the coverage of the

' construction inspection and reinspection activities performed by Blount is described as

-followst ' '

' A. QC INSPECTIONS .

The areas'.of certification of QC inspectors _has changed, over the project, to conform with the various types of work activities required for each stage of construction. As an example, _ inspectors are no longer certified in the area of post tensioning because that phase of construction is now complete. ,

r t Currently,'Blount Brothers has QC inspectors who are certified in one or more of'eight different' inspection disciplines. These disciplines include areas such as

, concrete,' masonry, concrete expansion' anchors, structural welding, structural

,  ; bolting, cadweld,' calibration, fireproofing,' and receiving. The~ inspectors are y, , Lcertified :in each discipline as either a Levei ! or Level !!~ inspector. The 3,. . inspection force has' fluctuated commensurate with the work load as have the lnumberiof inspectors certified .in any given discipline. Types of inspections iwhich have been or are currently'being performed include:-

5( e i Welding inspections including configuration, weld quality,'.weld size and placement, welder identification, and proper ' materials.

- e Structural; steel: bolting inspection including bolt type,' use of . washers,-

1 (quantity, and ~ verification of bolt torque by sampling in accordance with

.." . ASTM standards.'

v 'E

[.

i' a e i

t ,

Y 1. -

F -

EXHIBIT E-1 Page 2 of 3 e Cadweld inspections to serify embedment length and void area.

e Concrete inspections including excavation, rebar, form work, location of embedments, joints, conveyance equipment, placement, finishing, curing and hot and cold weather protection.

e Fireproofing including preplacement and postplacement activities.

e Grouting including preparation of surfaces, mixing, placement, and curing.

e Masonry laspections including mortar mixing, use of reinforcement, wall alignment, joint thickness, and appearance plus cold and hot weather precautions.

e Post-tensioning inspections including storage, staging, sheathing, trumplates, tenden installation, buttonheading, stressing, greasing, and bearing plate preparation, o Concrete expansion anchor inspection including size, location, plumbness, and shimming.

e Receipt inspection of materials used or insta. led by Blount Brothers.

B. REINSPECTIONS PERFORMED During the construction period, several reinspection programs have been

-initiated by - Blount Brothers.. The following is a summary of those reinspections:

o Concrete expansion anchors installed prior to 1979 were completely reinspected to verify proper installation. This reinspection was performed ,

I

.as a result of NRC IE Bulletin 79-02.

l 1

, - - - - r , ._. - _

. _ - . - - - . - .-. . . - - = _ _ . . . - . --

EXHIBIT E-1 Page 3 of 3

-e Containment structural steel slip connections were reinspected in 1981 to verify r;ial lengthening and washer condition and again in 1982 torque  :

verification was done due to changes in design. All belted slip connections  !

In both containment buildings were t einspected. .

e High strength structural bolts in the auxiliary building, river screen house and fuel handling -building were reinspected on a sample basis to demonstrate the acceptability of this botting.

I j e Post-tensioning.teridons were reinspected to assure that no unacceptable

, degradation of tendons.by rusting had occurred during storage. Another reinspection was . performed of tendon button heads due to prcblems

identified at other . sites. Also, reinspection was performed of post-

tensioning anchor heads for material properties and thread engagement.

e A reinspection of a sample of structural welding was performed to verify acceptance of welds to revised undercut criteria.

e . Blount Brothers . has been ~ included in the Reinspection Program

implemented as 6 resuit of NRC Inspection 32-05/82-04. A total of 8 of 23 Blount Brothers inspectors had their first 3 months work reinspected under-

_ the Reinspection Program.' A . total of more than 2,200 items were reinspected. it was found that the work of allinspectors during the first 3 months post-certification was acceptable, with a cumulative acceptance of 3

98%+.

's

s. .

t t  :.,-- <

l; +

4 '

u ,

I

~

EXHIBIT E-2 Page 1 of 3 EXHIBIT E-2 JOHNSON CONTROLS INCORPORATED Johnson Controls has been onsite cince February 1979 and has been performing safety related activities since January 1930. The scope of Johnson Control's work covers installation'of HVAC controls and instrumentation, including tubing and tube-track installation, hangers, instrument, and instrument panel installation. Johnson Controls work activities are delineated in Sargent & Lundy Specification F-2783.

Detailed information on the coverage of the construction inspection and reinspection activities performed by Johnson Controls is described as follows: ,

A. QC INSPECTIONS in the early stages of construction, Johnson Controls certified their QC inspectors in the basic area of visual inspection which included inspections of a wide range of work activities. As construction progressed, the certification method was revised to provide certification in more specific areas.

Currently, Johnson Controls quality control inspection certifications include:

receiving, calibration, visual welding and solder-joints, hangers / tube-track / tubing / torquing and panels / penetration / leak tests / instruments. An inspection can be certified in one or more of these disciplines as either a Level I or Level II inspector. Types of inspections which have been or are currently -

being performed include:

e Welding inspections including configuration, weld quality, weld size and placement, welder identification, and proper materials, e - Hanger inspections including configuration, location proper materials, s

welding, and damage.

EXHIBIT E-2 Page 2 of 3 e Solder jent inspections including appearance, size, and freedom from excess solder.

e Tube track inspections including damage, size, span distance, welding, torque, material, tube clamps, and location.

e Tubing inspections including damage, bends, size, and freedom from contact with tube track, o Panel installation inspections including location, welds and damage.

e Instrument installation inspections including location, welds, damage, torquing, and identification. ,

.e Penetration inspections including location, damage, materials and identifft.ation.

e Leak testing including verification that the tubing system is air tight and ,

has no blockages.

e Receiving inspections including physical inspection -of ' the material and documentation reviews to assure that the items and materials conform to procurement documents.

E. REINSPECTIONS PERFORMED During Johnson Controls' time onsite, two reinspection programs were instituted. The following is a brief sum' mary:

e- - All installed tube track hangers were reinspected in 1983 to verify the inclusion of stiffeners.

r , --

, . . - , . . . - ..-,e # - - .

y EXHIBIT E-2 Page 3 of 3 o Two QC inspectors' work was reinspected due to concerns raised by an audit deficiency in one case and a request from the Commonwealth Edison Project Construction Department in the other case.

e lohnson Controls has been included , in the Reinspection Program implemented as a result of NRC Inspection 32-05/82-04. A total of 5 of 7 Johnson Controls inspectors had their first 3 months work reinspected under the Reinspection Program. A total of more than 9000 items were reinspected. It was found that the work of allinspectors during the first 3 months post certification was acceptable with a cumulative acceptance of 98%. '

W l-f l

l h

EXHIBIT E-3 Page 1 of 3 EXHIBIT E_-3 HUNTER CORPORATION

- Hunter Corporation (Hunter) has been onsite sir.ce December 26,1976. The scope of work performed by Hunter includes mechanical erection activities associated with equipment sett!ag, piping, component supports, and pipe whip restraints. Hunter's work activities are delineated in Sargent & Lundy Specification F/L-2739.

Detailed information on the coverage of the construction inspection and reinspection activities performed by Hunter is described as follows:

5. QC INSPECTIONS Hunter's current quality control inspection certification includes the disciplines

.of hangers, piping (including equipment erection and as-built), and receiving.

An inspector can be certified in one or more of these disciplines as either a Level I or Level II inspector. Types of inspections which have been performed during the early stages of construction or are currently being performed under the present areas of certification include:

e Welding inspections including joint configuration / fit-up, weld quality, weld size and placement, welder identification, and material traceability.

e Pipe bending inspections including bend radius, pipe ovality, and surface discontinuities.

e Mechanical joint inspections including seating surface condition, torquir.d of bolts,~ torquing sequence, and orifice plate orientation, if applicable.

e- Component support inspections including configuration and assembly per design documents, welding, threaded connections acceptable and locked, pins secured with locking devices, movement unobstructed, variable settings correct, and use of proper materials.

I 1

l 1

l

qq m

EXHIBIT E-3 Page 2 of 3

- e. Pressure test' inspections including verification of test boundaries, test garge calibration, test pressure and duration, and weld inspection for l

' leakage.

l

.e Concrete expansion anchor inspections including anchor length and

' diameter, edge' distance and spacing, rebar contact, embedment length, number of washers, anchor plumbness, thread engagement, plate size, and j acceptable bearing surface between plate and concrete. l e Equipment' erection inspections including location and orientation, i foundation bolts and grouting, leveling and alignment, and lube oil and cooling water piping connections.

I e Piping system walkdown inspections including verification that installed

' Items. are -in place, intact, and . undamaged, and as-built information

verification.

e - Pipe' ' whip restraint inspections including location, orientation, configuration, proper materials, and welding.

e Receiving inspections including physical inspection of the material and .

documentation reviews to assure that the items and material conforms to i- the procurement documents. .

B.' REINSPECTIONS PERFORMED

-e - All hangers installed by Hunter prior to March of 1980 were reinspected in accordance with Hunter's . revised inspection. program to assure conformance to the design docunients.

e: Concrete expansion anchors installed prior to 1979 were reinspected to verify proper installation. This reinspection was performed as a result of

, - NRC IE Bulletin 79-02.

~ s

EXHIBIT E-3 Page 3 of 3 e Hunter ' Corporation has been included in the Reinspection Program implemented as a result of NRC Inspection 82-05/82-04. A total of 22 of 84 Hunter inspectors had at least their first 3 months work reinspected

- under the Reinspection Program. A total of more than 73,000 items were reinspected.' It was found that the cumulative acceptance rate was 980.

e. ' Additionally,' Hunter's lead inspectors perform periodic reinspections of

. inspector's work to verify acceptable ' performance. This is an informal management tool used to assure adequate inspections were being performed.

y l

l _.

i p

i n-J . - _

i EXHIBIT E4 Page 1 of 2 EXHIBIT E4 NUCLEAR INST &LLATION SERVICES COMPANY Nuclear Innallation Services Company (NISCo) first came on site at Byron in June of 1978. They have left and returned several times to meet the need for their services.

They hwe performed work under S&L Specification F/L-2834. The scope of work under this specification is to install specific portions of the NSSS system including centrol rod drive mechanisms (CRDMs), setting of the reactor vessel, reactor cavity preparation,-installatioc. of new and spent fuel racks, fuel transfer system and fuel handling equipment and tools, setting of reactor coolant pump internals and preparation of the reactor head.

Detailed information on the coverage of the construction inspection and reinspection

' activities performed by NISCo is described as follows:

A. QC INSPECTOR]

Nuclear Installation Services Company currently has one inspector onsite who is certified in six different areas including visual welding, mechanical, liquid penetrant, ultrasonic, documentation, and receiving inspections. The inspector

- is certified as a Level 11 inspector in all treas. Types of inspections which have been or are currently being performed include:

e Weld fit-up for various welding activities.

e Final weld inspection of safety-related welds performed by N!SCo.

O e Liquid penetrant inspections when required by the ASME Code or the work specification.

e Bolt torque for bolting operations included in the NISCo scope of work.

z, EXHIBIT E-4 Page 2 of 2 i e Visual inspection of mechanical installations.

e ' Receiving inspections of materials to be installed in the plant.

.B. REINSPECTION NISCo has been included in the Reinspection Program implemented as a result of NRC Inspection 82-05/82-04. A total of four of eight NISCo inspectors had their first 3 months reinspected under the Reinspection Program. A total of

~

more than 3000 items were reinspected. It was found that all inspectors' work during the first 3 months post certification was acceptable with a cumulative acceptance of 99b.

• w-

..s'

!!E V _ ,

h

~

A ,

OT _ EXHIBIT E-5 r Page 1 of 4 x'

EXHIBIT E-3 HATFIELD ELECTRIC COMPANY

~ ~

Hatfield Electric-c Company- (Hatfield Electric) has been at the Byron Station I Construction . site L since June !!976 to implement contractual requirements in h ~

accordanceiwith 'Sargent & Lundy.J5pecification F/L-2790. The ' scope of work performed by Hatfield Electric encompasses the majority of electrical work onsite, which consists of installation of both embedded and exposed conduit and underground

duct.; Also, Hatfield Electric is responsible for the installation of all cable pans and risers including associated hangers,' ladders and covers. Upon installation of electrical

.g -. penetration : sleeves. (by others), ' Hatfield places 'and maintains the . penetration

~

. internals prior to terminating activities. Paramount to Hatfield Electric's work is the <

- Jinstallation, and subsequent termination, of cable in the aforementioned conduits, c _ ; cable' pans,:and raceways. Electrical equipment installation and modification and

-' miscellaneous fire detection, fire protection and security system installations are part

- of the above listed scope of work.

e -

' Detailed information 'on; the coverage of the construction inspection and reinspection l ' activities performed by Hatfield is described as' follows:

A.  ! QC INSPECTIONS .

The areas of certification of 'Hatfield QC inspectors have evolved through the

1. ~ . construction period and in some cases have been grouped in different ways. As Lan~ example, early;in- the construction. period. separate areas of certification

~

. were established for-cable _ pan hanger, cable pan, conduit hanger and conduit L --installation,(whereas currently, the above ' four areas are grouped into two:

conduit / hanger installation and cable pan / hanger installation. Another example l'

L 'is thel inclusion. of bolting . inspections with cable pan / hanger- inspections, h :whereas in the past bolting was a separate area of certification.

f 3

pa qV

(: /

^

}

$ -g

+_ - . -

EXHIBir E-3 1

'E l Page 2 of 4 Hatfielo Electric presently has six differeat inspection disciplines including:

visual weld, conduit and hanger, equ;pment turnover / cable installation /termin- ,

ation, cable pan and hanger (including bolting), field and engineering design changes, and receiving. The inspectors are certified in each discipline as either a . Level I or Level II inspector. The inspection force has fluctuated cornmensurate with the work load as have the number of inspectors certified in any given discipline. -Types of inspections which have been or are currently being performed include:

e Embedded and exposed conduit inspections including size, type, condition, configuration, and location.

e Underground duct run inspections including size, type, condition, connections configurations, and location.

e Cable pan hanger inspections including configuration and assembly per design drawings, required bracing and location, condition of assem~ o ly, bolt torquing require:ntnts, and acceptability of welding.

e Cable pan inspections including size and type per design drawings, drain location and support adequacy, radii, elevation, and segregation per design -

drawing. Also, routing markers and segregation code marker location are reviewed as is metal condition and panel weld down.

e Cable installation inspections includa cleanliness, pan or conduit damage, weld down, edges for sharpness, routing and segregation markers, cable protection (half rounds, etc.), minimum bend radli, cable coils, cable height in pan and end seals. Also, reviewed are cable protection from welding,

~

set up of pulling equipment, temperature requirements, and cable pulling tension.

EXHIBIT E-3 Page 3 of 4 e Cable termination inspections document the size and manufacturer of the lug, number and calibration due date of crimp tool, nuraber of crimps, size and type of bolts, torque values and wrench identification, minimum bend radii and termination, and segregation per design drawings.

e Equipment installation and modification inspections include cleanliness and condition of mounting surface, joint anchoring details, alignment, leveling,

_ plumbing, and squaring of the equipment, bolt torquing, and/or welding.

e Electrical penetration inspections include condition of conductors and penetration, preparation and assembly of "O" type ring, pressure tests per instructions, bolting lubrication and torquing, rigging and installation per design drawings, and manufacturers' instructions and welding.

e Weld inspections include visual examination-for weld quality and size, welder qualification, completeness of travellers, and installation per design drawings.

e Receiving inspections include physical inspection of the material and documentation and reviews to assure compliance of the items and material to the procurement documents.

e Conduit Installation walkdown inspections included verification that installed items were inplace, intact, and undamaged, and as-built information verification.

I B. REINSPECTIONS PERFORMED During the construction period several reinspection programs have been initiated by Hatfield Electric. The following is a summary of those reinspections.

~

EXHIBIT E-5 Page 4 of 4 e Cable pan hangers installed in plant as of July 1977, and 10% of cable pan -

welds installed in plant as of July 1977.

e The containment isolation panels were reinspected to verify weld quality and size.

e- Concrete expansion anchors installed prior to 1979 were reinspected to verify proper installation. This reinspection was performed as a result of NRC IE Bulletin 79-02. -

e Cable crossover, bridges, and risers, conduit supports, cable tray stiffeners

-and cable routing were reinspected in January 1981 to verify compliance with design documents. Also, a sample reinspection of cables for proper routing and installation damage was conducted.

e All field welds' which did not have a complete and an acceptable weld traveller card were reinspected to determine the weld's acceptability and completeness.

e Hatfield Electric has been included in the reinspection program implemented as a result of NRC Inspection 82-05/82-04. A total of 23 of 87 Hatfield inspectors had at least their first 3 months work reinspected under the Reinspection Program. A total of more that 32,000 items were reinspected. It was found that the cumulative acceptance rate was 95+.

- -, ,._,..,m . _ _ - .- . . , , _ _ , _ . ,

EXHIBIT E-6 Page 1 of 2 EXHIBIT E-6 POWERS-AZCO-POPE Powers-Azco-Pope (PAP) has been on site since late 1978 to perform work under S&L Specification F-2906. They began safety-related work in early 1979 on small bore piping. The scope.of the contract includes the installation of small bore instrument piping and miscellaneous small bore (2" and under) systems.

Detailed information en the coverage of the construction inspection and reinspection activities performed by PAP is described as follows:

'A. - QC INSPECTIONS In the early stages of the project, inspection of installation attivities was grouped under the general certification area of welding. More recently, this area was revised to include the specific arcas of piping, hanger and documentation. Throughout the project, receiving has been an area of certification. The inspections which are performed or have been performed in the past are as follows:

e Receiving inspections include physical inspection of the material and documentation and review: to assure compliance of the items and material to.the procurement dccuments.

e Cold pipe bending to. verify proper material was used and to verify pipe ovality and bend radius acceptability.

e Hanger installation inspection to verify use of proper materials, configuration, location, welding and bolting, where applicable.

e Concrete expansion anchor inspections including size, location, plumbness, shimming, cut rebar, and use of washers.

s

N EXHIB;Y E-6 l Page 2 of 2 l e . Flex hose installation inspections to verify satisfactory configuraton, alignment, interferences and anchors. l e Mechanical joint inspection to verify acceptable seating surfaces, gaskets, studs and nuts, line-up and torque values.

e ' Instrument attachment to verify proper identification of fasteners and torque values as well as instrutnant identification.

e Piping inspections to verify use of proper materials, fitup, and final visual inspection.

e Hydrostatic and pneumatic test inspections, to verify the integrity of the test by assuring . vents ,re properly installed, test equipment is acceptable, test boundaries are correct, test pressure was held for the appropriate time interval, test caps were used, and temperature was acceptable.

B. REINSPECTIONS PERFORMED Powers-Azco-Pope was included in the QC inspector Reinspection Program as a result of NRC Inspection 82-05/82-04 A' total of 19 of 21 inspectors' first 6 months of accessible work was reinspected. The other two inspectors not reviewed had no work accessible during the first 6 months of inspection. A total of over 14,000 items were reinspected. The inspections performed for 18 inspectors was 100% reinspected. All discrepancies have been or are being corrected.

EXHIBIT E-7 Page 1 of 3 EXHIBIT E-7 -

RELIABLE SHEET METAL COMPANY Reliable Shees Metal Company has been at the Byron Station Construction site since September 1977 to implement contractual requirements in accordance with Sargent &

Lundy Specification F/L-2782. The scope of work conducted by Reliable Sheet Metal enconipasses the majority of heating, ventilating and air conditioning activities on site which consists of installation of duct, penetrations, fans, filters, smoke detectors, dampers and plenum housings incloding associated hangers, access doors, grills, registers, and diffusers. Upon installation of the air handling system or subsystem, Reliable Sheet Metal tests and balances tL2 installation.

Detailed infermation on the coverage of the construction inspection and reinspecticn activities performed by Reliable Sheet Metal is described as follows:

A. QC INSPECTIONS Reliable Sheet Metal presently has three different inspection disciplines including visual weld, detail configuration, and receiving, and the inspectors are certified in each discipline as either a Level l~ or Level 11 inspector. The inspection force has fluctuated commensurate with the work load as have the number of inspectors certified in any given disciplines. The types of inspections which have been or are currently being performed are as follows:

e Type 1 inspections cover the initial installation of hangers, penetrations, panels, racks, filters, fans, housings, dampers, and detectors. This inspection includes size, type, condition, configuration and assembly per design drawings, wcld quality and size and location.

e Type 2 inspections are a walk down review of completed systems or subsystems.

EXHIBIT E-7 Page 2 of 3 This inspection includes cleanliness, completion of system or subsystem in -

i accordance with design documents and damage thereto.

l

-e Type 3 inspections are performed on systems or subsystems which have not been turned over to the owner within 90 days after completion of the type two inspection. This inspection includes the same attributes as a type two inspection. .

e Type 4 inspections cre final reviews of all documentation and personnel qualifications associated with types one, two and three inspections. This inspection includes completeness and acceptability of all tests and inspections, certification of welders and inspectors and material documentation acceptability.

o Receiving inspections indude physical inspection of the material and documentation reviews to assure the items and materials are in compliance with the procurement documents.

B. REINSPESTION PERFORMED During the construction period, two reinspection programs have been initiated by Reliable Sheet Metal. The following is a summary of those reinspections.

e Concrete expansion anchors installed prior to 1979 were reinspections to verify proper installation. This reinspection was performed as a result of NRC IE Bulletin 79-02. (Again it should be noted that all HVAC concrete

! expansion anchors are installed by Blount Brothers and the above listed reinspection was of Blount Brothers installations.)

f. .

. . . _ _ - . . . . = . . . _ _ . _.

EXHIBIT E-7 Page 3 of 3 o All HVAC work installed prior to September 1982 is being reinspected for -

compliance to design drawings and specification requirements. ibis reinspection was a result of a "stop work" action imposed by Commonwealth Edison QA which is still curretitly in place.

e The QC Inspectors were not included in the Reinspection Program as their work is being a 100% reinspected.

5 i

r I

EXHIBIT E-8 Page 1 of 2 EXHIBIT E-8 -

PITTSBURGH TESTING LABORATORY Pittsburgh Testing Laboratory (PTL) has been on site since September 1977. The

_ scope of work performed by PTL includes nondestructive testing of welds, concrete testing, aggregate testings, concrete expansion anchor tecting, soils testing, calibration, bolting inspections, etc. PTL's work activities are delineated in Sargent &

Lundy Specification F/L 2850.

Detailed information on the coverage of the contractor construction activities performed by PTL is described as follows:

A. CONTRACTOR QA AUDITS AND SURVEILLANCES Pittsburgh Testing Laboratory is required by their QA Program to audit safety-related areas of activity. This has been accomplished through a continuous, comprehensive audit and surveillance program utilizing trained auditors. Since arriving on site in September 1977 and through the end of 1983, PTL has conducted 297 audits and surveillances of their own activities. Where deficiencies have been identified, PTL has taken corrective action to resolve the concerns.

B. COMMONWEALTH EDISON SITE QA AUDITS AND SURVEILLANCES Since the beginning of PTL's activities, Site QA has performed 31 audits and 435 surveillances of PTL to assure that their activities are in compliance with applicable codes, specifications, and procedures. The 31 audits resulted in 14 findings and 20 observations. While deficiencies were identified in both the audits and surveillances, corrective actions were taken to resolve the concerns.

e

b EXHESIT E-8 Page 2 of 2 C. COMMONWEALTH EDISON GENERAL OFFICE AUDITS As part of the Commonwealth Edison Program, site activities are audited at least semi-annually by a QA Department General Office audit team. Since Pittsburgh Testing Laboratory commenced activities at the Byron site,15 General Office audits have been conducted. Though these audits cover all site contractors, PTL has been included in the scope of the audits. These audits did not result in any major revisions to the PTL QA prcgram or procedures.

D. INDEPENDENT EVALUATIONS '

Pittsburgh Testing Lab.aratory's activities have been subjected to independent evaluation by the authorized Nuclear Inspectors and the ASME. The Ahls periodically observe 'the performance of nondestructive testing to assure compliance with the ASME Code. Additionally, the AN!s routinely review PTL inspections to assure that the correct and acceptable NDE has been performed on work within the jurisdiction of the ASME Code.

Pittsburgh Testing Laboratory's activities are also scrutinized by ASME Survey teams since PTL provides the NDE services to site ASME contractors and Commonwealth Edison. These surveys have identified no significant concerns with PTL.

r

EXHIBIT E-9 Page 1 of 2 l

EXHIBIT E-9 -

PEABODY TESTING SERVICES Peabody Testing Services (PTS) waf, on site from February 1976 to September 1977.

The scope of work performed by PTS included nondestructive testing of weias, soils testing, aggregate testing, concrete testing, bolting inspection, cadweld, calibration, etc. Peabody Testing Services' work activities were delineated in Sargent a Lundy Spelfication F/L 2850. Peabody Testing Se. vices was replaced with Pittsburgh

~ Testing L.aboratory due to their inability tc adequately staff the site as required in the scope of the aforementioned contract. However, the staffing shortcomings did not adversely affect the quality of PTS's inspection and testing activities.

Detalied information on the coverage of the contractor construction activities performed b'y Peabody is describad as follows:

A. CONTRACTOR QA AUDITS AND SURVEILLANCES Peabody Testing Services was required by their QA Program to audit its inspection and testing areas of activity. This was accomplished through an audit program utilizing trair.ed auditors. During PTS's twenty months on site, Peabody conducted three audits of their activities. Additionally, field activities were monitored by PTS surveillances.

B. COMMONWEALTH EDISON SITE QA AUDITS During the activities of PTS at Byron, Commonwealth Edison Site QA

. performed three audits and numerous surveillances of PTS to assure that their activities . were in compliance with applicable codes, specification, and procedures. The above referenced audits identified deficiencies generally associated with administrative implementation for which corrective actions were taken to resolve the concerns. Also, qualified Commonwealth Edison Site QA. personnel performed field surveillances of PTS's activities such as field

- concrete; testing, concrete cylinder tests, calibration, cadwelds, sieve analysis

EXHIBIT E-9 P2ge 2 of 2 of soils, structural steel installation including stud welding and halting, aggregate tests, welding, etc. Additionally, safety-related concrete placement was witnessed by Commonwealth Edison Site QA qualified personnel.

C, COMMONWEALTH EDISON GENERAL OFFICE AUDITS As part of the Commonwealth Edison QA Program, site activities are audited at least semi-annually by a QA Department General Office audit team. During Peabody Testing Services' period of activities at the Byron site, two General l Office audits have been conducted. Though these audits cover all site l contractors, PTS was included in the scope of the audits. These audits identified the same administrative concerns previously identified, which led to PTS's replacement.

~ , .

F-1 APPENDIX F  ;

1 ANSWERS TO NRC STMF OUESTIONS 1 Reference 1 -NRC letter dated February 3, 1984, from R.L. Spessard to C. Reed of Commonwealth Edison on the 1-12-84 Interim Report to the QC Reinspection Program Reference 2 - NRC letter dated November 18, 1983, from R.L. Spessard to C. Reed of Commonwealth Edison on the 10 28-83 Preliminary Report for the QC Reinspection Program The following are in response to the questions and comn.ents raised in Reference 1 (see questions Ql-Q12) and Reference 2 (see questions Ql3-Q15):

Ql. "The data should be presented in clear, concise, and accurate manner and in terms that a layperson can understand."

Response

l The report has been reformatted, clarifying notes have been added to l

data tables, and the presentation has been simplified to assist all readers in understanding the report.

Q2 "The data should be reviewed to determine if ttm various tables are accurate and compatible, e.g., Tables 3.1 and 4.7 relating to the number of Pittsburgh Testing personnel in the program."

F-2 Response: -

1 The data and tables have been reviewed and footnotes provided to clarify the information presented. The data in the January 12,1984, Interita Report was provided to respond to Question Ql4 (in response to Reference 2). Therefore, Table 3.1 (now Table 111-1) and Table 4.7 (now Table Q14-7) were accurate and compatible except that a clarifying footnote has been added to the latter table to point out that the inspectors tabulated thereon included inspectors performing both subjective and objective inspections in the inspection area. Other clarifications (see Table 111-4) have been provided where appropriate.

Also, refer to the response to question Q14 below.

Q3. " Provide results of the Hatfield computerized data base established to reconcile weld travelers to hangers. (Page 19 of 1-12-84 Interim Report)."

Response

The computerized data base is complete and is available on site.

After completion of the review, it was found that direct correlation of weld traveler inspection records to design drawing cable pan hanger and conduit hanger data could not be made for approximately 2% of the data. The augmented efforts to achieve complete correlation is

!. expected to be complete in March 1984.

Q4. " Elaborate on how Appendix C sample sizes were obtained for evaluating design limit compliance and how tSe sample selected bounds all of the remainder of the items in that category."

1 1

F-3 Rec,ponse:

-l An engineering evaluation has been performed for all observed discrepancies, as presented in Appendix C. In addition, Exhibit C-2 details the evaluation approach, including a description of how samples for the Hatfield and Pittsburgh Testing subjective weld discrepancies were obtained.

Q5. "On Table D.4, items 3C-27 and 3C-28 SdtL, resolution as written is unacceptable, e.g., in order to determine acceptability of a weld it would be necessary to determine extent of lack of penetration. (It is our understanding that these items were actually lack of fusion; this should be clarified)."

Response

Discrepancy Description The Johnson Controls Discrepancies, JC-27 and 3C-28, identified two 4-inch-long fillet welds with lack of fusion on the ends. As stated on Table D.4 of the January 12,1984 Interim Report, the extent of the weld discrepancy is " limited to either first and last 1/4 inch (or both) of weld". As detailed herein, the nature of these weld discrepancies was lack of fusion at the weld root rather than lack of penetration.

Weld Process Most discontinuities, except for lack of fusion and lack of penetration, manifest themselves at the surface. The following discussion (that the

! defects identified at the start and stop of the weld do not exist through the entire weld) is restricted to fusion type discrepancies.

e- 1 T

F-4 The welding procedures and the welder qualifications at Byron, which -

are in accordance with AWS DI.1, ensure that lack of fusion at the start and stop of a fillet weld is a local discrepancy and does not exist threugh the entire length of the weld.

7 The possible causes of fusion defects, in general, are improper heat (volts, amperes, and -travel speed), incorrect electrode angle, and improper arc manipulation. All of these parameters are identified in

- the welding and welder qualification procedures.

When the weld is just started, the base metalis cold compared to the heat required for obtaining proper fusion. When the welder does not wait long enough at the beginning of his v/ eld, enough heat may not be built up at the root for melting of the base metal for good fusion.

This localized condition is rectified as the heat of welding builds up.

This is typically within a length of weld equal to the weld size. Prcper electrode angle and arc manipulation are important throughout the making of a weld. At the termination of a weld, a phenomenon called

" arc blow" may be . encountered. This phenomenon is caused by magnetic fields induced in the base metal by high welding currents. It causes the arc to deflect in directions other than the intended direction. This deflection can in itself cause local fusion type defects; however, most of the time the welder counters this " arc blow" by manipulating his electrode. In general, this electrode manipulation works, and the weld is terminated - properly. Sometimes, the mcnipulation may cause the arc to play off of the base metal and thus proper termination may not be obtained.

I- .

The AWS . Structural Welding Code DI.1-79 acknowledges this start /stop phenomena in sub-paragraph 2.3.2.1 for fillet welds: "No reduction in effective length shall be made for either the start or crater (at the end) of the weld if the weld is full size throughout its i

length." AWS is acknowledging that this type of defect at the ends of D

. - . . _ . _ . - , _ _ _ . _ . . - ~ ~_,,_--,.-. - ..w_. -

F-5 the-weld is not uncommon and that it has little effect on the weld strength.

It is concluded that lack of fusion encountered at the start and stop of

- a fillet weld'is localized and does not affect the overall integrity of the weld.

- Q6.

. "With regard to item 5 above, for all welds having lack of fusion and lack of penetration, was the defect ccmpletely removed in order to assess how ' much of the weld remained upon completion of the

- repair? Did the weld quality and adecuacy determination consider the stress intensification resulting from the notches created by various defects, i.e., undercut, overlap, etc.?"

Response

A conservative approach' was taken in the engineering evaluation presented in Appendix C by considering the portion of the weld with incomplete fusion as ineffective.

~ There is no reason to remove these defects for the reasons stated in the response. to Q5. - As discussed in Exhibit C-2, Section D.5, and Appendix F, item Q15, the overlap portion of certain welds inpsected by Pittsburgh Testing will be removed by grinding and the welds reinspected. As discussed in Exhibit C-2, Section C.2, stress intensifi-cation resulting from -notches is- negligible, because the fracture toughness properties of the weld and base metals are adequate.

Q7. "In order to address the issue raised during the January 27, 1984,

-meeting regarding the- amount of loading seen on degraded welds, provide your analyses to assure that the welds with the lowest factor of s&fety comply with the applicable codes and the design basis. For example, one method would be to perform a detailed engineering t

,,,3- ,-,,w- v ,.,c e - - - - . . - , - ,-,

-a,-.- _ _ * . -

- . .=,, ,,,-, .,e.-, . , . . - - - c------w-- , - - . . . - , , , , - - - , - - , , - -

F-6 evaluation of fifty welds from the entire population of discrepant -

welds with the lowest factor of safe;y. Another method would be to select the worst weld in each category and the weld with tle lowest factor of safety in each category. Then perform a detailed engineering evaluation to determine if the worst weld would meet the design intent for the weld with the lowest factor of safety."

l

Response

1 As stated in the response to Q4, all weld discrepancies have been evaluated. Refer to Exhibit C-2, Sections D.1 through D.6, for the engineering evaluation of highly stressed welds and their compliance -

with design criteria.

Q8. " Provide a summary regarding the number and type of code (ASME and AWS) rejectable items found during the reinspection for each contractor. Further, with regard to the number of rejectable ASME Code items, please cxplain how you are going to assure that the items that have not been repaired are acceptable. This "mcludes both items that have and have not been reinspected."

Response

All discrepancies that were determined- to exceed ASME code examination acceptance criteria were repaired, even though they were determined by evaluation to have no design significance. The ASME l' discrepancies are tabulated and discussed in Exhibit C-3. The AWS L discrepancies are tabulated and discussed in Exhibit C-2.

Although some work was not reinspected as part of the Reinspection Program, a significant quantity of work was reinspected. The reinspected work is represc:.tative of all similar work that was not reinspected. The result of the engineering evaluation of discrepancies

F-7 Response: -

All discrepancies that were determined to exceed ASME code examination acceptance criteria were repaired, even though they were determined by evaluation to have no design significance. The ASME discrepancies are tabulated and discussed in Exhibit C-3. The AWS discrepancies are tabulated and discussed in Exhibit C-2.

Although some work was not reinspected as part of the Reinspection Program, a significant quantity of work was reinspected. The reinspected work is representative of all similar work that was not reinspected. The result of the engineering evaluation of discrepancies

yielded no discrepancy vith design significance. This provides assurance that work not reinspected was constructed with similar quality.

. Q9. ' "Provide your evaluaticn with regard to the acceptability of past inspections involving not recreatable attributes and inaccessible work performed by the inspectors whose accessible work was reinspected in accordance with the program. Your evaluation should consider the results of the reinspections performed to date as well as available information obtained from past QA/QC audits, surveillances and inspections involvin6 this activity."

Response

Under the Reinspection Program, a significant quantity of werk was reinspected. Also, the reinspected work is representative of work that p was not reinspected. Furthermore, the enginer. ring evaluation yielded no discrepancy with design significance. As a resuit, this provides assurance that work not reinspected was constructed with similar quality.

F-8 Inaccessible and . not recreatable work was either inspected by

' inspectors involved in the Reinspection Program or by inspectors of similar qualification. The verification of the acceptability of the inspected work provides assurance that work not reinspected was constructed and inspected with similar quality. Although inaccessible and not recreatable work was not reinspected, there are other assurances that this work was done properly. The information in Table Q9-1 summarizes the work that was not reinspected and the assurance l that this work was properly performed.

In addition to the above, construction testing is performed by Commonwealth Edison-Operational Analysis to functionally check the operation of equipment and controls. This testing program is conducted prior to and in addition to preoperational testing and system demonstration testing.

The approach utilized by Commonwealth Edison to check construction work provides a high level of confidence of the acceptsility of not recreatable attributes and inaccessible work. This approach provides many independent layers of inspection and review of field installations to assure compliance to requirements. These levels of overview include contractor audits, independent inspections by the independent testing agency, overinspections by the independent testing agency, reinspections performed by the contractor, unit concept inspections (UCIs) by the independent testing agency under the direction of Commonwealth Edison QA, surveillances by both Commonwealth Edison Project Construction and QA, audits by Commonwealth Edison General Office and Site QA and independent evaluations performed by outside agencies throughout the term of the project. Appendix E describes the various icvels of overview in detail.

As an example, cable pulling performed by Hatfield Electric is a not

+- recreatable inspection. To provide a high confidence that the work

/

F-9 was performed properly, the independent testing agency has been performing a continuing overinspection of cable pulling activities.

This is in addition to the required inspection performed by the contractor. Another independent layer is the overview surveillances of cable pulling activities performed by Commonwealth Edison QA.

Another exarnple is noncreatable concrete inspections performed by Blount Brothers. To assure that concrete was mixed and transported properly, the independent testing agency performed tests on the concrete. Cylinders were taken and laboratory tested after curing to verify the strength of the concrete. The resuits of these tests were provided to Sargent & Lundy for examination and trending purposes.

In addition to these levels of overview, Commonwealth Edison QA performed surveillances of concrete mixing, testing, and placement activitics. These in-proces, overviews provide an assurance that even not recreatable or inaccessbile inspections were performed in accordance with the established requirements.

In addition to the layers of overview discussed above, Commonwealth

- Edison QA performed trending of deficiencies throughout the project. Any identified trends required positive corrective action.

Moreover, since no trends adverse to quality were identified through this. process, additional assurance of the adequacy of the work is provided.

In conclusion, this overview coverage of the contractors on site as well as the additional eva!uations performed have given Common-wealth Edison confidence that all the work has been performed correctly.

y L

y F-10 Table 09 Summary of Work Not Covered by Reinspection Pronram Work Not Covered By Assurance That Work i Contractor' Reinspection Program Was Done Properly l

' Hatfleid Electric - Embedded Conduit overinspection performed by Commonwealth Edison Project Construction prior to concrete pour.

l Embedded Duct Runs Overinspection performed by Commonwealth Edison Project Construction prior to concrete pour.

Cable Installation Overinspection since 1981 and (Pull force on Cable) inspection of work done prior to 1981 by Pittsuurgh Testing.

Pulling tensicc calculations by S&L were performed.

Receiving Inspection Overinspection by Common-wealth Edison QA was per-formed.

Storage Periodic Surveillance by Commonwealth Edison Project Construction and by Common-wealth Edison Operational Anal-

- ysis.

'- Blount Brothers . Encased Structural Engineeririg evaluation by S&L Steel / Block Wall Columns of the quality control structural eteel review (QCSSR) performed by Commonwealth Edison per statistical sampling plan.

Concrete Placement . Concrete mix design by S&L and

- tested in laboratory. Concrete was mixed in a computerized batch plant thas limiting human errors. Surveillance and over-

~

inspection performed by Commonwealth Edison. .

~

Receiving Inspection Overinspection by Common-wealth Edison QA was per-is formed.

i; l -.

w ,

f -..,,,..~.o - . . . , -..m,, , ,

y .,.,%._.. . - - . . - . - - , _ , , . , , . ,y.,,.#,,.,,,,,y,_,. , y

F-il Table 09-1 (Cont.)

Summary of Work Not Covered by Reinspection Program Work Not Covered By Assurance That Work Contractor Reinspection Program Was Done Properly Rework on Elements These elements were revised due Due to Revisions to design revisions and, there-fore, the work performed prior to these changes has no impact on design.

Pittsburgh Testing Embedded Structural Bolts Engireering evaluation by S&L of the structural bolt reinspec-tion work performed by Commonwealth Edison as a result of a 50.55e to account for .

missing inspection reports.

Engineering evaluation also per-fo med tural steelforreview qualityQCSSR)

(control struc-review done by Commonwealth Edison.

VisualInspection of Engineering evaluation by S&L Embedded Field Welds of weld size, length, and under-cut for structural steel inspec-tion performed by Common-wealth Edison.

Concrete Placement Concrete mix design by S&L and tested in laboratory. Concrete was mixed in a computerized batch plant thus limitin5 human errors. Surveillance and over-inspection performed by Commonwealth Edison.

Electrical Cable Pulling Pulling tension calculations by S&L were performed.

Calibt ations Review of structural riigh strength bolts as a result of a 50.55e. Periodic recalibration of torque wrenches and other inspection devices.

W -

? et9 -

F-12 Ta_ble 09-1 (Cont.)

Summary 3f Work Not Covered by Reinspection Program Work Not Covered By Assurance That Work l- - Contractor Reinspection Program Was Done Properly

, Peabody Testing Embedded Structural Bolts Engineering evaluation by S&L I ot the structural bolts rein-spection work performed by Commonwealth Edison as a result of 50.55e to account for missing inspection reports.

Engineering evaluation also performed for quality control structural steel review (QCSSR) done by Commonwealth Edison.

Embedded Cadwelds Overinspection By Commonwealth Edisor. Project Construction.

i L Soll Tests Subsequent settlement measure-ments.

Concrete Placement Concrete mix design by S&L and tested in laboratory. Concrete was mixed in computerized batch plant thus limiting human errors. Surveillance and over-inspection performed by Commonwealth Edison.

Powers-Azco-Pope Socket Weld Fitup No socket weld cracks detected during reinspection program.

Hunter Mechanical Joint Torques Leaks would be determined by -

- hydrostatic pressure testing.

Johnson Controls : Receiving Overinspection by Commonwealth Edison QA.

i..

4 l-f

F-13 Q10. "Make available at the site data describing those welds which have lack of penetration, lack of fusion or cracks. The data should include the initial QC inspector and welder's name."

Response

This data exists in the primary inspection record and is available on site.

Qll. " Based on the results of the Reinspection Program, provRJe your

- evaluation of the quality of the work inspected by inspectors whose work was not reinspected as part of this program. Your evaluation should also include available information from QA/QC audits, surveillances and inspections involving this activity."

Response

As discussed in the response to Q9, the approach used by Commonwealth Edison to check construction work provides a high level c! confidence in the work performed by Byron. This approach does not focuc on one type of inspection or on one contractor, bat rather provides an overall check and balance on each of the contractors' activities. Consequently, the approach applies equally to work inspected by inspectors whose work was not reinspected as a part of the Reinspection Program as well as those inspectors whose work was reinspected. Appendix E provides further details of this independent checking.

As an example, Hatfield Electric is required to inspect the welds performed by them on conduit and cable pan hangers. To pravide a high confidence that the work was performed properly, the independent testing agency performed an overinspection of a sample of the work (generally 10%). Another independent layer of review is the UCI performed by the f independent testing agency under the direction of Commonwealth Edison QA. This overview inspects all the work in an element of the plant. As a l

F-14 i

I result, work performed much earlier in the project, as we'l as the j recently completed work, is checked to the carrent vendor and I i

l engineering drawings. In addition, Commonwealth Edison audits and 1 surveillances cover verification of installations including inspection of welds.

These multiple layers provide an assurance that work by inspectors not included in the Reinspection Program were performed correctly. Trend analyse of deficiencies are also performed throughout the project and positive action is required for any trends identified.

In summary, since the Reinspection Program showed inspectors to have performed acceptable work a,d the layers of inspections, surveillances cad audits also demonstrated the inspectors' work was acceptable regardless whether they were included in the Reinspection Program population or not, Commonwealth Edison concludes that the inspectors' work involving the various contractors at Byron was performed properly.

Q12. " Based on the results of the Reinspection Program and avaliable information from QA/QC audits, surveillances and inspections, what conclusions can be drawn relative to the appropriateness of the acceptance criteria of 90% for subject attributes and 95% for objective attributes."

Response

The basis of the acceptance criteria is presented in Chapter 111. In our judgment the acceptance criteria provide an appropriate and high degree of assurance that no discrepancy of design significance remains undetected in the plant.

F-15 For the objective attributes, a total of 156,926 items were reinspected and 3,247 discrepancies were noted. For the subjective attributes, a total of 44,980 items were reinspected and 4,132 discrepancies were evaluated. The discrepancies which had potential for affecting the design basis calculations were evaluateo. This engineering evaluation showed that these discrepancies had no design significance, thus validating the original program objective that first line inspectien activity provide reasonable assurance that discrepancies of design significance do not go undetected.

Q13. "The report should be draf ted in accordance with the original program.

Specifically, the tables and conclusions based on those tables should be based on the findings of the Level II examiner or the independent Level III examiner. Use of a Commonwealth Edison Level III examiner to change the results of the independent Level III findings is not in accordance with the original program."

Response

The tables and conclusions are based on the findings of the Level II or

-third party Level III examiners (see Appendix A and Tables A-1 through A-8). Results from Commonwealth Edison Level III examiners have been removed from the tables and conclusions.

Q14. "We understand you will provide tabulation of the results of inspection attributes (weld overlap, undercut, etc.) in order to determine the need, if L

any, for further inspections. The tabulation could be made availat:le to our inspectors, and need not be in the report, but as a minimum, the conclusions ym have reached regarding the tabulations should be included in the report." >

l

[

y 9 '

F-16

? ,

. Response: -

^

The results of inspection attributes (weld detail, visual weld, material

verification, etc.) are presented in Appendix A and Appendix B. The

, i attributes of work reinspected for each inspector were those attributes

/ covered by inspections done by that inspector during his first 3 months of inspection work. In the ca:e of the visual weld inspection attribute, the elements of this attribute'are discussed in detail in Appendix C. Overlap '

,- undercut, and other weld attribute elements are discussed further in

- Exhibit C-2 for AWS welds and in Exhibit C-3 for ASME and ANSI B31.1 ,

welds. All attributes were broken down into elements to determine

' whether significant trend's exist.' A review of this assessment is discussed

. in response to Q15. - It is on the basis of that assessment and the work

~

coverage discussed below, that no futher inspections were considered to

., _ ~'be necessary.

To ensure the representativeness of the inspectors to be reinspected,

^

Commonwealth Edison also reviewed all of the areas of inspector

]

. qualification covered by each contractor's inspector certification program and the number of inspectors qualified in each area. Commonwealth

~ Edison then compared the number of inspactors previously selected to the

~ total number for each work area to ensure that all were adequately

, represented for reinspection. These areas of qualification determined the inspection attributes to'be reinspected.

4 The results of this comparison are presented by contractor in Tables -

Q14-1 (Blount Brothers), Q14-2 (Johnson Controls), Q14-3 (Hunter), Q14-4 I . (NISCo), Q14-3 (Hatfield Electric), Q14-6 (Powers-Azco-Pope), Q14-7 (Pittsburgh Testing), and Q14-8 (Peabody Testing). As can be seen in the stables, all areas of work are well represented, except for those which

- could not be recreated for reinspection. . For example, the work of

-fireproofing and coating inspectors could not be reinspected because

-those inspections were performed on the surfaces to be coated or they ee h !

, - - .,+a - , - - - ,--. ..--w ,--.:._v,m-w--mwwv_., - ~ , , ,,r-,re-wy- ., v...c,p.n---,,w~,,,%3-.,. e--,.w,.-,,,,,~,n.-,y,r.4..w--.-,--,

F-17 were performed on the coating process itself. These inspections cannot be performed after the application was completed. The reinspection work was as representative o1 all work as was feasible. These tables also demonstrate that the sarnpling basis included a significant portion of the total inspector population in the various areas of work certification.

Table Q14-1 Comparison of Inspectors Qualified / Reinspected by Area of Qualification Blount Brothers No.of S

No. of Percent Area of Inspectors Qualified Inspectors in Included in Qualification in Area Area Reinspected

• Reinspection Concrete 12 2 17 %

Masonry 6 2 33 %

Concrete expansion anchors 5 2 40 %

Weld inspection / structural 11 4 36%

• • Post-tensioning 10 0 0%

• • Cadwelding 4 0 0%

• Calibration 5 0 0%

• • Fireproofing 3 0 0%

-** Receiving 6 0 0%

TOTALS 62 10

• ' As a point of clarification for. understanding the various numbers used in this report, it must be recognized that the sample selt.ction pocess produced an individual. This' individual may have been certified to perform inspections in only one area or he may have been certified in many areas. As an example, as shown in

. Table !!I-l in Chapter.111, Blount Brothers has a total population of 28 inspectors.

In Table Q14-1 above, Blount Brothers has nine areas of qualification and shows 62 inspectors qualified. This means that of the 28 individual inspectors, several were qualified to inspect in more than one area. Again, referring to Table 111-1, Blount Brothers has a total of 8 inspectors reinspected; again referring to the above Table

-Q14-1, Blount Brothers shows 10 inspectors reinspected in various areas. By referring to Table B-1 in Appendix B, it can be seer that individual inspector E and individual inspector G were each certified in two ar2as, which accounts for the difference between 8 and 10.

o* Areas of work which cannot be recreated for a reinspection.

J

- p g ,

.--+r-- c _p

F-18 Table Q14-2 Comparison of Inspectors Qualified / Reinspected by Area of Qualification Johnson Controls No. of No. of Percent Area of Inspectors Qualified Inspectors in Included in Qualification in Area Area Reinspected

• Reinspection Visual inspection - 7 5 71%

• See discuss.on l in note (*) in Table Q14-1. For Johnson Controls, five individual inspectors were r 9ected. No inspectors had multiple certification (see Appendix B, Table b- i.

Tchle Q14-3 Comparison of Inspectors Qualified / Reinspected by Area of Quahfication Hunter No. of No. of Percent Area of Inspectors Qualified Inspectors in Included in Qualification in Area Area Reinspected

• Reinspection Piping / hangers 57 16 28%

Piping 6 1 17 %

Piping as-built 21 5 24%

TOTAL 84 22

• See d!scussion in note (*) in Table Qle-1. For Hunter, 22 individual inspectors were reinspected. Sixteen inspectors had multiple certifications as piping inspectors, hanger inspectors (see Appendix B, Table B-3).

Table Q14-4 Comparison of Inspectors Quallfled/ Reinspected by Area of Qualification NLSCo No. of No. of Percent Area of Inspectors Qualified Inspectors in Included in

_ Qualification in Area Area Reinspected

• Reinspection Visual welding 6 4 67%

Mechanical 6 4 67%

'~

-TOTAL 12 8

• See discussion in note (*) in Table Q14-1. For NISCo, four individual inspectors were reinspected. Four inspectors had multiple certifications (see Appendix B, Table B 4).

~

F-19 Table 014-5 Comparison of Inspectors Qualified / Reinspected by Area of Qualification Hatfield Electric No. of . No. of Percent Area of Inspectors Qualified Inspectors in Included in Qualification in Area Area Reinspected

• Reinspection

Visual welding 20 8 40%

Conduit installation 21 6 2%

Cable terminations 21 5 2'vM Equipment installation 14 0 0%

Equipment modification 12 2 17%

Cable pan installation 21 1 5%

Cable pan hanger 22 2 9%

Conduit as-builts 28 8 29M A325 boiting 11 1 9%

TOTAL 170 33

• See discussion in note (*) in Table Q14-1. For Hatfield Electric, 23 individual inspectors were reinspected. Six inspectors had multiple certifications (see Appendix B, Table B-5).

Table Q14-6 Comparison of Inspectors Qualified / Reinspected by Area of Qualification Powers-Azc@ ope No. of No. of Percent Area of Inspectors Qualified Inspectors in Included in Qualification . _

in Area Area Reinspected

• Reinspection Welding 20 19 95%

c* Receiving _2, ,0 0%

TOTALS 22 19

• See discussion in note (*) in Table Q14-1. Fnr Powers-Azco-Pope,19 individual l inspectors were reinspected. No inspector had multiple certifications (see Appendix B, Table B-5).

'c* Areas of work which cannot be recreated for a reinspection.

i

1 i

F-20 l Table Q14-7 Comparison of Inspectors Qualified / Reinspected <

by Area of Qualification  ;

Pittsburgh Testing '

No.of No.of Percent

. Area of Inspectors Qualified Inspectors in included in Qualification in Area Area Reinspected

• Reinspection Concrete expansion anchors / structural 43 9 21 %

Visual welding 23 14 61 %

• Concrete: field-44/

lab-44/ plant-5 .93 0 0%

• • Soils: field-15/ lab-14 29 0 0%

• *Cadweld 10 0 0%

• Post-tensioning 3 - 0 0%

• Fireproofing 4 0 0%

• Coatings 2 0 0%

• Calibration 17 0 0%

• Electrical 12 0 0%

TOTAL 236 23

• See discussion in note (*) in Table Q14-1. For Pittsburgh Testing, 23 individual inspectors were reinspected. No inspector had multiple certification (see Appendix B, Table B-7).

• • Areas of work which cannot be recreated for reinspection. Electricalinspections were of cable tray cleanliness and cable pulling tension.

l

-Table Q14-8 Comparison of Inspectors Qualified / Reinspected by Area of Qualification Peabody Testing No.of No.* of Percent Area of Inspectors Qualified Inspectors in Included in Qualification in Area Area Reinspected Reinspection l

Visual welding /

structural steel 6 6 100 %

• Concrete 23 0 0%

• Soils 20 0 0%

• Cadweld . 8 0 0%

• Coatings 1 0 0%

• Calibration 1 0 0%

TOTAL 64 6

• See disct.ision in note (*) in Table Q14-1. For Peabody Testing six individual insper tors were reinspected. No inspector had multiple certification (see Appendix E6, Table B-8).

• • Areas of work which cannot be recreated for a reinspection.

_ - _ . - - , . -~ _ _ _ _ _ _

I F-21

. Q15. "We understand that you will review different inspection activities and determine if certain _ areas'such as final hanger inspections warrant further review based on reject rates."

Response

The results for all attributes were evaluated on a contractor by contractor basis to determine whether any trends existed in the observed discrepancies (i.e., " reject rates") that might warrant further review.

This evaluation involved a sorting of the observed ciscrepancies into discrete elements and a comparative assessment of these elements. If any element demonstrated a significant contribution to the discrepancy total, its significance was reviewed and any inspection practice ramifications were considered. The discussion that follow; describes the three trends that were identified and the oisposition of each that was implemented.

.a. For the subjective attribute of visual weld inspection the results for each contractor were analyzed using approximately five elements.

Three contractors (Powers-Azco-Pope, Pittsburgh Testing, and Peabody Testing) had an apparent problem in recreating visual weld inspection. Specifically, the lack of fillet weld gauge usage and drawing length tolerances resulted in non-reproducible inspections. In the case of Powers-Azco-Pope, this resulted in certain inspectors failing. Currently, both dimensional tolerances and weld gauges are being used to make these inspections more objective than was the case at the time the original inspections were done. Both Pittsburgh

-Testing and Peabody Testing showed an undesirable failure rate for the attributes of undercut and overlap. Constant training during the visual weld inspector's tenure has much improved the consistency of their juogments made in the areas of undercut and overlap.

Discrepancies of'this type were insignificant (see Chapter VI and Appendix C).

l

F-22 1

l 1

b. In evaluating observed discrepancies associated with Hatfield visual j weld inspections, it was noted that a disproportionately large fraction j of the discrepancies were related to the inspection of sheet steel  !

welds. This is not necessarily an indication of a specific problem with 2 particular. Inspector but rather a manifestation of the historical  ;

question throughout the industry, related to visual inspection of sheet steel welds. This trend is not unexpected. The standard applied in

, the- past and which was used in the reinspection program was AWS DI.l., a structural steel code.

From a design impact standpoint, these discrepancies are insignificant. Not only were these discrepancies spu;ifically evaluated and deter.nined to be of no consequence but also sheet steel welds generally have very low load requirements. The strength afforded by even. a code rejectable weld is almost always much greater than that needed to fulfill the design requirements. This conclusion has been validated by actual tensile tests previously performed on a similar sample of rejected welds on another project.

The tests showed that welds which would be rejectable under AWS DI.1 criteria had margin in excess of what is required by design. In fact, in'atmost all cases, the failure under load resulted in failure of the sheet metal rather than the weld itself.

c. For the objective attributes,.each contractor was analyzed by area of certification, each of which averaged approximately 5 to 10 attributes. - A total of 22 objective areas of certification were analyzed for the site contractors. The final hanger inspection performed by Powers-Azco-Pope which included a check- for weld completion, identified a large number of observed minor discrepancies on hanger length and fillet leg size. This observation was due to lack of specified length tolerance and the absence of fillet

. weld gauges during the initial inspection. Currently, both dimensional

iw '

F-23 i

l tolerances and weld gauges are employed by the contractor. Because

. observed discrepancies of this type were also shown to be insignificant (as discussed in Chapter VI and Appendix D), we have concluded that no additional inspections are warranted. No other objective attribute trends were identified.

In summary, all observed discrepancies have been assessed for unacceptable trends.- The types of discrepancies found art minor discrepancies, and had no design significance. Therefore, no further inspections are necessary.

J C

4

, ..,n.-,

- ..~ . . , - . - .

.. , , . , , , , . - . , , _ . . ~ , . e ,