ML20138H455

Testimony of Ef Trainor on Conservation Council of North Carolina Contention WB-3 Re Capabilities & Performance of Applicant Const QA Program.Related Correspondence

ML20138H455
Person / Time
Site:	Harris
Issue date:	10/25/1985
From:	Trainor E CAROLINA POWER & LIGHT CO., CYGNA ENERGY SERVICES
To:
Shared Package
ML20138H418	List: ML20138H413 ML20138H423 ML20138H435 ML20138H455 ML20138H464
References
OL, NUDOCS 8510290046
Download: ML20138H455 (22)

Text

,

October 25, 1985 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of )

)

CAROLINA POWER & LIGHT COMPANY )

and NORTH CAROLINA EASTERN ) Docket No. 50-400 OL MUNICIPAL POWER AGENCY )

)

(Shearon Harris Nuclear Power )

Plant) )

APPLICANTS' TESTIMONY OF EUGENE F. TRAINOR ON THS; CAPABILITIES AND PERFORMANCE OF APPLICANTS' CONSTRUCTION QUALITY ASSURANCE PROGRAM (CCNC CONTENTION WB-3) 8510290046 851025-t PDR ADOCK 05000400 T PDR.

0.1 Please state your name.

A.1 Eugene F. Trainor.

0.2 Mr. Trainor, by whom are you employed and what is your position?

A.2 I am employed by Cygna Energy Services, Inc., an engineering and management consulting firm located at 286 Congress Street, Boston, Massachusetts, as a Senior Vice President.

O.3 Please describe your educational background and membership in professional organizations.

A.3 I was graduated in 1953 from the United States Coast Guard Academy at ?!ew London, Connecticut, where I received a Bachelor of Science degree in General Engineering. I subsequently received a Master of Science degree in Management from Rensselaer Polytechnic Institute, Troy, New York, and a certificate in Executive Management from Northeastern University, Boston, Massachusetts. I received training in U.S. Navy nuclear power plant testing and operations at the Mare Island Naval Shipyard, Vallejo, California.

I am a member of the American Society of Mechanical Engineers and have been a member of its Main Committee on Nuclear Quality Assurance since the Committee's inception in 1975. I have been the Chairman of its Subcommittee on Personnel Qualifications l

since 1978. I might note that Mr. Harold Banks, CP&L's Corporate Quality Assurance Manager, has worked with me on both the Main Committee and Subcommittee. In addition, I am a Senior Member of the American Society for Quality Control and the Chairman of that Society's Construction Technical Committee. I am a registered Quality Engineer in the State of California and a registered Mechanical Engineer in the State of Massachusetts.

0.4 Please describe your professional experience.

A.4 I have more than 25 years of experience in the nuclear industry, most of which has been in the area of quality assurance and quality control. My nuclear experience began in 1959 at the Quincy shipbuilding Division of the Bethlehem Steel Company as a member of the DlG/D2G Project Production Engineering Group. In that assignment, I was responsible for the development of inspec-tion, material control, preventive maintenance and system comple-tion procedures used in the construction of the nation's first nuclear-powered destroyer, the USS BAINBRIDGE DLG(N)25 and its land-based prototype. I was also responsible for the preparation of reactor cote installation and handling procedures, the assurance of component installation suitability, liaison with the Atomic Energy Commission / Knolls Atomic Power Laboratory and ulti-mately the management of the power unit installation.

Additional assignments while at Bethlehem Steel Corporation included participation in the reactor plant test program for the i

I l

U.S. Navy nuclear submarines SS(N) SCAMP, SS(N) PLUNGER, and the SS(N) PERMIT while at the Mare Island Naval Shipyard, and as a consultant to the Portsmouth Naval Shipyard, Kittery, Maine, where I developed a process control manual for the handling and installation of SSW submarine reactor plant components. In 1962 I was made Assistant Chief Test Engineer, SSW project, and in late 1963 I was promoted to Chief Quality Control (Nuclear).

Shortly thereafter, the Quincy facility was sold to the General Dynamics Corporation.

In 1964, I became the Division Nuclear Quality Control Manager initially reporting to the Manager of Nuclear Projects and ulti-mately reporting to the Quincy Division General Manager for all aspects of Quality Assurance and Control in the design, construc-tion and overhaul of naval nuclear ships at the Quincy Division.

I was responsible for the conversion of the General Dynamics Electric Boat Division Nuclear Quality Control program from a single product program (i.e., submarines) to a multi-faceted pro-duct mix (i.e., surface ships and submarines) as well as develop-ment of procedures necessary to implement the program. In March 1968, I was assigned the additional functional responsibility of Process Engineering Manager of the Quincy Division. In that capacity I was responsible for manufacturing and industrial engi-neering, as well as applied research and development. During the period between 1964 and 1971, the Quincy Division completed and delivered four nuclear powered submarines, two nuclear submarine tenders, and a floating nuclear decontamination facility.

L

In March 1971 I left the shipbuilding industry and joined the Stone & Webster Engineering Corporation to participate in the then rapidly growing commercial nuclear industry. My initial assignment was as a Nuclear Engineer in the Maine Yankee Project in the area of core assembly, fuel handling, and quality assurance. Six months later, in October 1971, I was assigned as Chief Engineer of the Engineering Assurance Division with the responsibility for the development and successful implementation of in-depth quality assurance programs in the Engineering ,

Department for all nuclear and selected fossil power plants designed by Stone & Webster. In January 1973, I was promoted to Manager of the Quality Assurance Department with responsibility for the administration and management of that department as well as the establishment and implementation of the corporate quality assurance program. During this period the Stone & Webster topi-cal quality assurance prcgram was developed. This program was the first such program developed by an Architect / Engineer to be approved by the Atomic Energy Commission. During my eight-year association with Stone & Webster, I was involved in approximately 18 nuclear power projects, including Maine Central Power Company's Maine Yankee project; Virginia Electric and Power l

Company's Surry and North Anna projects; Long Island Lighting l

Company's Shoreham and Jamestown projects; the Power Authority of l

the State of New York's James F. FitzPatrick, Indian Point 3 and Green County projects; as well as the Sun Desert project for San

_4_

Diego Gas & Electric; the Fulton project for Philadelphia Electric Company; the River Bend 1 & 2 project for Gulf States Utilities; the Nine Mile Point #2 project for Niagara Mohawk; the Millstone 3 and Montague projects for Northeast Utilities; and the Clinch River Liquid Metal Fast Breeder Reactor project for the Department of Energy.

s In 1979 I left Stone & Webster to pursue a career as a consultant with an initial assignment at Boston Edison Company to assess the

. Pilgrim Nuclear Power Plant Unit #1 Operations Quality Assurance Program and the quality assurance program for the design and construction of Unit #2. As a result of the accident at Three Mile Island, I did additional studies relative to the impact of that accident upon the Boston Edison Nuclear Management and Quality Assurance Systems.

In May of 1980 I joined Cygna Energy Services, then known as Earthquake Engineering Services, as Vice President and Principal with the initial assignment to establish management and quality assurance consulting services for the firm. Cygna Energy Services, which at present is a broad-based, multi-service con-sulting engineering firm, was founded in 1973 and specialized initially in structural and seismic engineering. Its entry into the nuclear industry was the seismic evaluation of the Humboldt Bay nuclear power plant for Pacific Gas & Electric in the early 1970's. In 1980, a corporate decision was made to expand its services to cover all f acets of the energy industry and to become, in effect, a " mini" Architect / Engineer. ,

In the past five years, what was essentially an in-house quality assurance service has expanded to be a significant part of Cygna's consulting business. Cygna has evaluated, developed, and/or assisted in the implementation of quality assurance programs for fabricators, manufacturers, architect / engineers, constructors, utilities and testing facilities. In addition to my managerial role as Senior Vice President, I have participated

-in an active consulting role with such assignments as the review for Northern States Power Company of its Spare Parts Program at its Prairie Island Nuclear Power Station as well as an evaluation of the implementation of the Quality Assurance Program at that station. Other assignments have included the evaluation of the proposed operations Quality Assurance Program to be used by Houston Lighting & Power Company at its South Texas Project and participation as a member of the Senior Review Team for the inde-pendent design reviews of the Grand Gulf Unit 1 plant for Mississippi Power & Light Company, Fermi 2 for Detroit Edison Company, and Comanche Peak for Texas Utilities. I have par-ticipated in the review of the quality assurance programs of Bechtel Power Corporation, Sargent & Lundy, Ebasco, Gibbs and Hill, Brown & Root, Boston Edison Company, Long Island Lighting Company, and Portland General Electric Company.

O.5 What is the purpose of your testimony?

A.5 The purpose of my testimony is to present the results of my review of the Construction Quality Assurance Program utilized by Carolina Power & Light Company (CP&L) at its Harris Plant project, specifically as it relates to the contention that alleged widespread drug abuse has impaired the quality of c( astruction at the Harris Plant. In this regard, I will discuss the ability of a well-conceived and properly implemented construction quality assurance program to deal with widespread drug abuse. I will present my evaluation of the data from various CP&L reinspection programs and my conclusions regarding overall inspector profi-ciency.

0.6 What did you review in preparing this testimony?

A.6 I reviewed the CP&L Quality Assurance Program as applied to the construction and testing of the structures, systems, and com-ponents of the Shearon Harris Nuclear Power Plant. Special emphasis was directed to those parts of the quality assurance program associated with inspector selection, trainin;;, eva-luation, error detection, and error data. I reviewed in detail the data obtained to date by CP&L as a result of the implemen-tation of its OA attribute surveillance program and the results

! of reinspections of the work performed by quality inspectors L

implicated in possible drug activity. I also reviewed evaluations performed by the Institute of Nuclear Power Operations (INPO) and

the NRC's Construction Assessment Team (CAT) and Systematic Assessment of Licensee Performance (SALP) programs of work per-formance, test performance, training, qualification, and construction activity.

0.7 What is a well-conceived, properly implemented, and moni-tored QA program capable of ensuring?

A.7 As defined in 10 C.F.R. Part 50, Appendix B, a well-conceived, implemented , and monitored quality assurance program is capable of ensuring " adequate confidence that a structure, system, or component will perform satisfactorily in service."

0.8 What are the essential elements of such a OA program?

A.8 The Commission's regulations at 10 C.F.R. Part 50, Appendix B, set forth 18 criteria which must be addressed in the esta-blishment of such a OA program. I would point out that these criteria are not unique to the nuclear industry in that they can be found to some degree in any effective quality management system. Essential elements of such a OA program are organization and staffing; the identification and application of OA management tools such as inspection, audit, and defect analysis; the periodic evaluation of implementation and program effectiveness; and lastly the documentation of results.

l 0.9 How does a well-conceived, properly implemented, and moni-tored OA program ensure quality?

l l

l

A.9 Ouality is ensured by the implementation of a series of planned activities designed to minimize to the greatest extent possible, a repetition of past errors and by doing so provide

" adequate confidence" that the end product will perform its required functions. The well-conceived program is designed in conjunction with the process work flow to provide a series of

" hold points" or " gates" at which the work product is evaluated for correctness and completeness. If acceptable, it passes to the next work station. The number and location of the gates or hold points is based on the number of steps in the process, the type of process, and consequences of past errors. The concept of pro-cess flow with performance gates or hold points is equally appli-cable to the design process and the procurement process, as well as the construction and test processes.

Redundant gates ensure that errors missed at one gate are detected at subsequent gates. Given the significance of failure of a safety-related system to public health and safety, the well-conceived quality assurance program in the nuclear industry is replete with redundant " gates," as well as quality system reveri-fications to determine the adequacy of performance and the completeness of implementation. Such reverifications generally take the form of surveillances, audits, and sampling reinspec-tions.

O.10 can a OA program ensure that construction work is 100 percent free of discrepancies?

l 1 l

l 1

l A.10 A OA program cannot ensure or provide a 100 percent con-l fidence that construction work is 100 percent free of discrepan-Cies.

To expect a quality assurance program to provide perfection is indicative of a failure to understand that the assurance process must rely on the human element which, irrespective of the checking process, with its built-in redundancies, allows for error. The framers of 10 C.F.R. Part 50, Appendix B, demon-strated their knowledge of a OA process and the human element when they established in the definition of Quality Assurance the requirement for " adequate confidence" of performance. Inherent in the word " adequate" is the understanding that it is less than perfection.

0.11 If something less than 100 percent proficiency is achievable, what then are accepted industry norms for inspector proficiency?

A.ll To the best of my knowledge, nuclear industry norms for inspector proficiency have never been established in an industry i

standard. However, reinspection programs established by either the NRC in the past few years or various licensees in conjunction l with the NRC have established a 95 percent acceptance rate as an l

acceptable standard for repeatability for objective inspection I

and a 90 percent acceptance rate for subjective inspection.

Objective inspections are defined as those inspections amenable

to quantitative measurement. Subjective inspections, however, are those inspections that rely on subjective interpretation as for example visual weld inspection. Based upon reinspection program criteria, it could be established that acceptable performance for an inspector in the nuclear industry lies in the 90 to 95 percent repeatability range. (The concept of objective and subjective inspection utilizing the 95 and 90 percent acceptance level was discussed and accepted by the Atomic Safety and Licensing Board and Atomic Safety and Licensing Appeal Board in the operating license decisions involving Commonwealth Edison Company's Byron Nuclear Power Station, Units 1 and 2.) In comparison, it is interesting to note that it is generally accepted practice in industrial quality control that " inspector errors result in a performance of about 80 percent accuracy in finding defects."

(J.M. Juran, Quality Control Handbook, Third Edition at Page 12-51, 1974) 0.12 How does anything less than 100 percent error-free construction provide reasonable assurance of public health and safety?

A.12 Human error will not allow for 100 percent error-free construction. Nuclear power plants, airplanes, bridges and l

skyscrapers are designed and constracted with this lack of per-

! fection taken into account while still assuring public health and safety. In a nuclear power plant, those systems essential to l

O assure public health and safety are designed with a high safety factor (or conservative design margin) ar.d with system redundancy to compensate for any deficiencies in construction. The degree of design conservatism and system redundancy in nuclear power plants are exceeded in modern industrial applications only in the NASA space program.

0A inspection programs are directed to safety-related systems and components. Inspections are designed with their own built-in redundancies -- sequential layers of screening or " gates" to minimize the probability of residual error. For example, a single weld on a safety-related pipe that is subject to ASME Code requirements will be subjected to inspection by the craft super-visor; Ir MQC and NDE inspectors both during and after welding; by the Authorized Nuclear Inspector at hold / witness points; during walkdown inspections, hydrostatic tests and start-up tests; and finally during independent, third party pre-service inspections as part of the ASME Code,Section XI, inservice inspection base- i line determination. Additionally, at the Harris Plant the work of inspectors is subject to audit by inspection supervision and reinspection pursuant to the Harris Plant OA attribute sur-veillance program.

l Industry experience with reinspection programs has, in general, shown that deficiencies found during reinspections were not safety significant -- were not of a nature that, were they "to

have remained undetected, could have adversely affected the safety of operations of the nuclear power plant." This general industry experience has been confirmed by the results of reinspections conducted at the Harris Plant, which have demonstrated that the Harris Plant Construction OA program is 99.5 percent effective and that no deficiencies with safety significance remain undetected.

0.13 Can a OA program ensure safe construction if there is wide-spread drug usage among construction workers?

A.13 Yes. OA programs are designed to prevent and detect errors.

Errors are caused by people for any number of reasons. Drug usage can be an error'cause, as can boredom, hangovers, family problems, cold medicines, etc. In fact, if drug usage is a causal factor in error generation, then widespread drug usage should be easily detectable due to the alleged potential for widespread error generation. Nevertheless, drug use by construction workers would not negate the ability of the OA program to detect and correct errors. A Construction OA program, properly implemented, does in fact detect and correct errors and will ensure safe construction, even in the face of widespread drug usage.

0.14 Based on your review of CP&L's OA program, what opinion l have you formed regarding its procedures, implementation, and its program for surveillance and audit?

i.

A.14 Our review of the CP&L OA program indicates that it is a well-conceived, properly implemented and monitored QA program.

Particularly noteworthy are its methods for the selection and training of inspection personnel, while its surveillance and audit programs relative to inspector performance are the most comprehensive that I and my colleagues at Cygna have observed to date.

0 15 What evidence can you point to in support of your opinion?

l A.15 To support our opinion, we reviewed applicable sections of the following program documents: the Harris Plant Preliminary

Safety Analysis Report, CP&L Corporate Quality Assurance Program, CP&L Ouality Assurance Manual, Shearon Harris Nuclear Power Plant OA Procedures (COAs), Shearon Harris Nuclear Power Plant OC Procedures (COCs), Technical Procedures and Corporate QA Department Administrative Procedures, and the Harris Nuclear Plant Quality Check Program. We found these to be responsive to the requirements of 10 C.F.R. Part 50, Appendix B.

In addition, we reviewed the flow and documentation assuciated with inspection programs including pipe hanger inspection program, electrical inspection program, instrumentation erection I

inspection program, mechanical and pipe welding inspection program, HVAC installation inspection program, civil inspection  ;

i programs, and safety-related and seismic Category I equipment installation inspection program. We found the inspection program to be well-conceived and complete.

l

l We also examined documentation associated with the reinspection of 79 seismic pipe hangers, which included 376 welds and 5,264 l attributes. Of this population, we field checked 8 hangers, which included 30 welds and 420 attributes, all of which were correctly reflected by the documentation reviewed.

We reviewed the reinspection data obtained in the implementation of the OA attribute surveillance program and in evaluating the work of inspectors who were implicated in possible drug activity.

As we discuss in some detail below, the data establishes the overall mean quality inspector proficiency for the Harris Plant at 99.5 percent.

j Our review also incluced NRC's CAT and SALP inspections and INPO's evaluation of the Harris construction project and found that they supported our opinion regarding the adequacy of the CP&L OA program implementation.

0.16 Applicants have adopted Military Standard 105-D (MIL-STD-105D) as the basis of the sampling system for reinspec-tions and QA attribute surveillance pursuant to Harris Plant l

Procedure COA-7. Is this an acceptable system?

A.16 Yes. MIL-STD-105D is a sampling system that has acceptability and widespread usage in both the defense and i

nuclear industries. It presents a series of sampling plans designed for use where the units of products are produced in a 1

l l

series of lots or batches over a period of time. It is utilized for acceptance sampling by attributes. In CP&L's Harris Plant Procedure COA-7, the units cf product in the pop 01ation to be reinspected are completed inspections of safety-related work packages and MIL-STD-105D is utilized to establish the sample size for reinspection. Inspector proficiency is the ratio of acceptable inspection attributes to total inspection attributes.

Applicants are properly applying MIL-STD-105D in selecting the sample size for attribute surveillance and for reinspections of the work of inspectors whose proficiency has been questioned.

0.17 What evaluations did you perform of the results of reinspections that have been performed as part of the OA attribute surveillance program?

A.17 We plotted the inspection proficiency per surveillance activity and determined the mean inspection proficiency to be 99.5 percent. We observed that the mean exceeds the specified proficiency for objective attributes of 95.0 percent by 4.5 per-cent. We also observed that in all cases individual inspection proficiency per surveillance activity exceeds the specified pro-ficiency for objective attributes from 0.3 percent to 5 percent.

l REINSPECTION NINCEEN SURVElLLANCE ACT1v1T'.ES

'00 ir i A_ / Fw _, .

.h A, , c 99 N '

Ss -

97 -

y 98 -

y 95  :  :  :  :

E 94 -

F Ss -

4 W 92 -

W 91 -

n . 90 - ,

89 -

es -

57 -

88 . . . . . . . . . . . . . . . . . .

1 2 3 A 5 6 7 8 9 10 11 12 13 14 15 i t 17 18 19 ACTMTY PJUMBER D PROF CATA + P R O T M E!*.N e PROP SPEO O.18 Are the results of the reinspection program statistically cignificant?

A.18 Yes. The results are statistically significant based on an analysis of variance (ANOVA) that determined the variation of the mean surveillance activity proficiency (99.5 percent) about the cpecified proficiency (95 percent). This technique uses a ratio of the variation of the mean and the variation of the experimen-tal error (F-ratio) to test for statistical significance. The results of the test show that there is at least 99 percent con-fidence that there is statistical significance. Furthermore, based on the average sample size of surveillance activity attri-butes, the technique will detect a change in proficiency as small cs 1.0 percent.

0.19 Based on the sample size and surveillance results, what conclusions can be drawn regarding the inspection activities at the Harris Plant?

A.19 Based on the sample size and the reinspection results, it can be concluded with 99 percent confidence that the inspection proficiency at the Harris Plant is greater than 95 percent. We can also conclude with 99 percent confidence that the range of inspector proficiency will lie between 99.9 percent and 97.7 per-cent. This would appear to be consistent with other industry reinspection programs, as for example the reinspections at Commonwealth Edison Company's Byron Plant which showed a range of 96.3 percent to 99.6 percent for inspector proficiency based on the results of " objective" inspection.

0 20 In your opinion, is this an acceptable basis for evaluating the performance of inspectors implicated in possible drug activity?

A.20 In my opinion, the reinspection program utilized by CP&L at the Harris Plant is an acceptable method'for determining the per-formance of inspectors. In this particular instance, we are reviewing the performance of a group of inspectors alleged to be implicated in possible drug activity. A method that is totally satisfactory for performance review of the general inspector population must be satisfactory for the performance review of a segment of that population given that you are comparing the per-formance of the segment to the total population to determine i

variances in performance. In this particular instance, there appears to be no appreciable variance between the performance of the total population, the population segment, and the apparent r I

industry norm.

0.21 Have you reviewed the results of the reinspections of the work of QA personnel who were implicated in possible drug activities?

A.21 Yes, I have.

0 22 What conclusions can you draw from the results of such reinspection?

A.22 It can be concluded with 99 percent confidence that the l inspector proficiency of the alleged drug users will be greater than 95 percent and that it will lie in a performance range of 99.9 percent to 96.1 percent.

0.23 What prediction could you make, based on this data, of the performance of all inspectors and the performance of those inspectors who might be implicated in drug activity?

A.23 our review of the results of the OA attribute surveillance l data and the reinspection data of those inspectors who might be implicated in drug activity indicate that there is no apparent l

difference in performance. This being the case, it can be pre-dicted with 99 percent confidence that inspector performance for

objective inspection will be greater than 95 percent and will lie within a performance range between 99.9 percent and 96.1 percent.

Additionally, industry experience, including the experience at the Harris Plant, would indicate that inspector errors found as the result of such reinspection will have minimal significant safety significance.

REINSPECTION OF FIF iEEN INDIVIDUALS -

100 l /

h_ f e f, /%

~ -

99 JF V 95 -

97 - .

g 9s -

95  :

E SA-93 -

W 92 -

9 91 -

n.

• 90 -

89 -

58 - .

57 -

82 , , , , , , , , ,

13 1A 15 4 5 5 7 8 9 10 11 12 1 2 3 ACTMTY NUM2EM PROP'3pCO

+ PROF' MEAN e D PROP CATA I

0 24 Based on your review of CP&L's reinspection program and data, can you draw any conclusions about the likelihood that there has been widespread drug usage by Harris Plant inspectors at a time and place when such usage could influence the quality of Harris Plant construction?

A.24 As stated above, there are no apparent differences between the performance of the total inspector population and that of the alleged drug users. Furthermore, if one accepts the hypothesis that widespread drug usage will cause a significant reduction in inspector performances, then there must be an indication of significant performance reduction to conclude that there was widespread drug usage among inspection personnel. This hypothe-sis is not substantiated by the reinspection data which indicates that inspector performance is well above the accepted standard.

0 25 Mr. Trainor, please summarize your overall assessment of CP&L's QA program.

A.25 I have stated that a well-conceived, implemented, and moni-tored OA program is capable of ensuring " adequate confidence that

, a structure, system, or component will perform satisfactorily in service" and that I have found that CP&L has such a program in l

place with regard to the construction and testing of Harris Plant structures, systems, and components. I have commented favorably l on CP&L's inspector selection, training, evaluation, error detec-tion, and error data; and I have indicated that my findings have i -

l

been' substantiated by activities performed by INPO and NRC's CAT and SALP programs. CP&L's surveillance and audit programs rela-tive to inspector performance are the most extensive I have observed.

Statistically, we have identified that the results of the reinspection programs for the Harris Plant far exceed acceptable industry parameters, and that the sample size of the reinspection data is sufficient to conclude with a 99 percent confidence level that the range of inspector proficiency will lie between 99.9 percent and 96.1 percent.

Based on my review of the Harris Plant Construction OA program and the results of the QA attribute surveillance program and reinspections of inspector performance, I conclude that there is l

adequate confidence that the Shearon Harris Nuclear Power Plant i

' is well-constructed and will perform satisfactorily.

l l

l l

i 6

9 i

I l  !