IR 05000508/1986013

Insp Rept 50-508/86-13 on 861208-19,870108 & 12-13. Violations Noted:Const Spec for Concrete Installed for Reactor Bldg Did Not Incorporate Appropriate Reinforcing Steel Clearance Criteria & Concrete Grout Nonconforming

ML20212N139
Person / Time
Site:	Satsop
Issue date:	02/26/1987
From:	Richards S, Toth A NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
To:
Shared Package
ML20212N112	List: IR 05000508/1986013 ML20212N107 ML20212N123
References
50-508-86-13, NUDOCS 8703120373
Download: ML20212N139 (20)

Text

7

.

3 -; ,

1 .~ s

" -

o-

, .

, ,

,

'

"4 '

,p

<! ;. , ~t

. .-

,

,

'

U.'S.LNUCLEAR~REGULATORYCOMMISSION

, , -)

, , ; REGION V'

'

Report No.750-508/86-13 Docket No. 50-508~ ,

.

Construction Permit No. CPPR-154 Licensee: Washington Public Power Supply System (WPPSS)

P. O. Box 1223 Elma, Washington 98541 Facility Name: Washington Nuclear Project 3 Inspection Conducted: December 8-19, 1986 (On-site)

January 8, 12, 13, 1987 (In-office Review)

Inspector: , M, I J- 24-# 7 A. D. Toth, Project Inspector Date Signed Consultant: C. G. Bruch, EGG Idaho Idaho National Engineering Laboratory Approved By: M_ F-M M7

. Richardpf Clifef, Date Signed

,gineering Sdetion, Region V Sumsiary:

Inspection on Deceniber 8-19, 1986 and January 8, 12-13, 1987 (Report No. 50-508/86-13)

Areas Inspected: Routine unannounced inspection by a regionally based inspector, and a cons.ultant, of the implementation of the readiness review program / construction assurance program for concrete work completed prior to commencement of extended construction delay. Inspection procedures 30703, 47051B, 47056B, 92700, 92701, and 92702 were considered for inspection '

guidanc Re s ul_t_s In the area inspected,.two violations were identified with respect to rebar clearance requirements of ACI-318-71 in applicable concrete specifications (Reference Paragraph 13), and failure to correct identified deficiencies (Reference Paragraph 14).

The program implementation did not appear to meet licensee commitments relative to consideration of prfor problems in the selection of work, samples and attributes, and in establishing that construction work procedures had implemented design requirements and were satisfactor Implementation of statistical random sampling appeared overly simplistic and incomplete, with conclusions subject to challenge of their validit PDH ADOCK 05000D00 0 PDR

-

.

-

. .

b DETAILS 1. Persons Contacted Washington Public Power Supply System P. D. Olson, WNP-3 Program Director

• C. E. Love, WNP-3 Project Support Manager

• C. M. Butros, WNP-3 Planning & Administration Manager

• D. R. Coody, WNP-3 Project Quality Assurance Manager

• K. Drinkard, WNP-3 Quality Assurance Engineer

• A. G. Carlyle, WNP-3 Quality Assurance Engineer

• D. W. Coleman, WNP-3 Licensing Manager

• M. M. Monopoli, WNP-3 Plant Manager

• E. A. Stauffer, WNP-3 Plant QA/QC Manager L. J. Garvin, Readiness Reviews Program Manager

• G. A. Block, Engineering Assurance Manager

• R. L. Knawa, Construction Assurance Program Manager

• T. A. McCormick, WNP-3 CAP Concrete Module Team Leader F. Teague, WNP-3 CAP Concrete Module Engineer

• N. F. Blais, WNP-3 CAP Concrete Module QA Engineer S. Michaels, WNP-3 CAP Concrete Module QC Inspector L. Fields, WNP-3 CAP Concrete Module QC Inspector EBASCO

• M. Taylor, WNP-3 Project General Manager

• P. Gupta, WNP-3 Site Project Engineer

• Torturgul, WNP-3 Civil Lead Engineer (ESSE)

• C. Bennett, WNP-3 Contract Administration Manager

• E. Niemi, WNP-3 Resident Civil Engineer

• P. L. Pitman, WNP-3 Quality Program Site Manager (Acting)

Adams Associates (WNP-3 Owners Group Agent)

• J. A. Adams, Site Representative

• Designates persons in attendance at exit meeting 9:00-10:30 AM December 19, 1986, 2. Construction Assurance Program (CAP) Status The first CAP Module (C3-01 Earthwork) report was submitted to NRC September 10, 1986 and was under final NRC review at the time of this inspectio The second CAP Hodule (C3-02 Concrete) report has not yet been prepared by the licensee. The licensee CAP review plan was issued to CAP staff October 30, 1986 and the reviews were approximately 75% complete at the time of this inspectio The WNP-3 project is in extended cocstruction deferral status, with no construction in progress, except preservation and maintenance activitie p --

.

The EBASCO site erigineering staff level was nine (9) at the time of the inspectio . Background A July 12, 1985 NRC letter to WPPSS accepted the licensee proposal to conduct a readiness review program.for WNP-3 (and WNP-1). The proposed two phase program would address adequacy of completed construction during phase I, and all aspects of new or changed design and construction and readiness for operation in phase II. The phase I program consists of a Construction Assurance Program (CAP) and Engineering Assurance Program (EAP).

The licensee responded to specific NRC questions on the CAP through a January 9, 1986 letter to NRC, along with a revision to the CAP. Based upon the WPPSS replies, Revision 1 of the CAP was accepted via February 25, 1986 letter to the licensee. The NRC acceptance noted that planned module report format and content appeared to be tentative and would be the subject of further NRC inspection and review, to determine if the licensee construction reviews provide confirmation that completed plant construction complies with FSAR commitments and NRC regulations. The attachment to the NRC acceptance letter noted that the plan appeared capable of providing assessments which would ascertain whether or not all such commitments and regulations had been me The NRC acceptance of the program recognized that the validity of the final product would be a function of the effectiveness of the general program describe The described CAP accepted by NRC discussed detailed and in-depth reviews, with:

... selection of work activities based upon known or suspected licensee and/or contractor problems and industry problems;

... checklists and reviews which consider past special programs that addressed known problem areas; and

... sample selections based upon detailed review of contractor performance history and the extent of contractor activit In response to specific NRC questions, the licensee also described in the January 9, 1986 response that the CAP team will:

... establish that (contractor procedures) implement the design requirements and would result in satisfactory construction products;

... select known or suspected problems areas (including NRC open items)

that pertain to their respective modules (e.g. concrete) which will be considered by the team in developing review plans, samples, and attributes for inspectio The above descriptions of CAP consideration of past problems appeared to satisfy the NRC interest in assuring adequate licensee synthesis of individual past inspection and audit results (to draw conclusions about the

-

-

.

,

.

-3

'

, ,

q -

-

'

.

. .

&

overall extent,-significance, and root'causes of problems), and the effectiveness of' evaluation of inspection findings (to ensure that they were not treated as isolated items to be dealt with individually as opposed to being symptoms of more pervasive underlying problems requiring broader management' action). Purpose of this Inspection The purpose of this inspection .wa's to assess implementation of the CAP during work in progress on the Concrete module, at a time when CAP review personnel were available for interview and records samples were readily available for examinatio The assessment was intended to provide details of the program implementation, to allow future meaningful interpretation of the report'

expected to be submitted by the licensee to NRC after completion of the licensee reviews, analysis of findings, and definition of follow-up action The assessment was intended to identify any weaknesses in the program implementation which may require additional licensee actions for inclusion in the final report, or alternatively,; additional NRC efforts to supplement the final report (as a basis for NRC conclusions regarding acceptability.of completed concrete construction.)

The inspection was heavily oriented toward assessing implementation of the licensee commitment to in-depth and detailed reviews. This included evaluation of: The application of a review of past known and suspected problems to the selection of work, samples, and attributes to be considered by the review tea . The extent of licensee examination of contractor work procedures, especially in process controls which cannot be assessed by examination of completed construction. (Prior NRC inspection findings indicated a weakness in control of contractors and implementation of FSAR commitments in procedures.) The distribution of sample selection by work / structures, time period, and contractor involve The selection of the populations from which samples were taken for assessing particular attribute . Inspection Approach A regional. inspector consultant participated in the inspection. The consultant assisted in preparation of a preliminary list of signif tcant elements of concrete construction, and in evaluation of overall and specific aspects of the WPPSS CAP concrete module on site implementatio lie was on site December 10 through 19, 198 ~

- .

.

'

.

'

The inspector and consultant examined the CAP' Concrete review plan which was issued to the WPPSS reviewers for implementation, to assess the nature of the instructions, the nature of problem review conclusions which had been incorporated and the manner in which procedure reviews were to be addresse The inspector compiled a list of previously identified NRC inspection findings which NRC had not yet determined the final resolution, and reviewed the history of past inspection findings whose resolution had previously been accepted by NRC. He interviewed CAP program personnel and project QA personnel regarding the nature of the current problem review -

efforts and status of previously identified NRC inspection issues. The inspector also reviewed checklists and backup documentation associated with the CAP review of past audit and surveillance reports and findings relating to concrete work. He examined data relative to the currently incomplete CAP review of past nonconformance reports relating to concret The consultant developed a preliminary list of significant elements relating to concrete quality. This list was used, along with a similar list prepared by an NRC Quality Assurance Branch consultant, as a basis for examination of the scope of the licensee's review checklist The inspector and consultant examined nearly all of the CAP review checklists which had been implemented, or were substantially in progress, and the associated plant records in many cases. They particularly considered the sample selected compared to the amount and dates of work by each contractor, the significance of the review attributes, the nature of the population from which samples were selected, and the size of the sample ,

The inspector also reviewed the status of development of plans for conducting reviews for which checklists have not been define . Overview of Program Commitment Implementation The approved licensee Construction Assurance Program was to: " Provide additional assurance that there are no undetected generic quality problems associated with the fabrication and installation work of Supply System Contractors during their period of performance at WNP-3." For concrete work, the generic limitation materialized in the selection of random samples from a population of work by all contractors involved, without specific focus on individual contractors (other than by proportionate amount of work done) and with little consideration of their individual performance historie The NRC report of CAP program review (Inspection Report 86-01) concluded that the CAP was capable of providing an assessment which will ascertain that all FSAR commitments and regulatory requirements have been me Actual implementation for the concrete module is reflected in notes by a member of the licensee's Overview Committee (T. Bishop) which stated in item I.A: " Program designed to provide added assurance (versus total verification)." A substantial FSAR review was conducted, however, the CAP matrix of FSAR Commitments versus Specifications and CAP Checklists showed that not all FSAR commitments were addressed by CAP review checklists. For

..

. , - , .- - --- .

• '

. e' ~

y ' ,

,

• * . i

'

.

,

.- 5 s

+ -

,

P

,

.;. . FSAR:conditments:identifiedwithaCAPchecklist,thechecklistoften

.

addressed only a part of the commitment; the checklists were notably weak in assessing commitments involving work processes which could not readily be confirmed by examination of hardware-or existing direct verification records. Little effort was directed to assure that individual-contractor generated procedures had incorporated such PSAR/FSAR requirements during the various periods that1 work was performed. The CAP efforts were, rather, i

to consider only selected significant attributes of concrete construction which might be confirmed by recorded numerical data, or signatures on i

'

records, or inspection of in place concrete surfaces and exposed

'

- reinforcing steel. This partial review appeared consistent with the licensee commitment to provide assurances additional to that arising from existance of . prior quality assurance _ program The program implementation considered some past problems at WNP-3 and some significant past problems in the industry, and these were noted in the

-

review plan. However, the sample selections, individual checklists of

,

attributes to be examined, absence of supporting documentation, and

interviews indicated that the overall CAP problem review effort was

minimal. During the inspection, the inspector noted that undefined i additional data analysis of past audit and surveillance reports was planned, and a review of contractor past nonconformance reports was planned

(with some trending work already accomplished). However, the undefined and l' incomplete status of these-reviews existed at a time when implementation of the review checklists was already 75% complete, with little evidence that interim results of the problem reviews had been incorporated. The CAP management representatives also clearly stated during the inspection and exit meeting that the purpose of the CAP was not to examine resolutions of

'

>

'

previous problems for adequacy, nor to effect resolution of previously identified open items. The extent and manner of consideration of the

, problem histories was not consistent with NRC perception of the intended reviews.

l Specific inspection findings which illustrate the above observations are

'

presented in subsequent sectious of this inspection report.

) Overview of Statistical Premises i

A basic premise has been implemented in the CAP concrete reviews, which l involves selection of a sample of 60 items of various types for review of

some selected attributes. The absence of discrepancies in the reviewed

] 1 sample then was considered as statistical basis for concluding that 5% or

less of the population (from which the sample was selected) contains such

!

discrepancies.- This promised to provide a substantial contribution of

,

assurance that prior quality assurance programs had functioned as intende !

The licensee's stated CAP purpose of " additional assurance that there are no undetected generic quality problems" led to a random sampling program

, which ignored the potential differences between contractors, and considered I

work activities of all contractors as one homogeneous population. The

!_ premise of homogeneity implies that only deficiencies which may be common f

'

to all contractors will be assessed with the probabilistic confidence described in the. CAP review plan. Similarly, the random sampling approach

L

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

,

,. . - -

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

'

. -

-

" '

? n S -

.

1. m' [ '

, .

c .

[

. did.not consider the~ variance in. contractor performance with time,.and

-

again presumed ~a~ homogeneity in the work activity populatio '

The val'dity i of the homogeneity premises appears questionable, in view of

, : variability of individual contractors' and time ' dependence of the following-

_

. ;, '

-

,t '

'*

_

. Individual quality assurance programs,

_

,

,

' Individual management systems effectiveness,

'

.

• -

Individual procedures 1 approaches; and

Individual problem' histories' with tim The licenne indicated'a' great deallof confidence that the differences between contractors was minimal in viewIof EBASCO having reviewed each -

'

contractor's, procedures lfor compliance with' commitments, and in view of the existence of audit' and surveillance overview programs in effect; for each

~ ~ ~

contractor. However,' the inspector noted that a cursory review of WNP-3 problem histories shows'that early int'ernal audits and NRC inspection

. findings both had challenged the effectiveness of licensee. control of-contractors, effectiveness of procedure reviews, and effectiveness of  !

contractor internal programs to control quality.' One. contractor was

, _ determined.to have inadequate training of QC-inspectors and an inability to identify and correct internal work control problems. These issues appear to' undermine the. basi ~s for the licensee confidence, particularly in view of=

absence of concerted: CAP effo~rt'to ascertain the effectiveness of related prior corrective action programs. A sampling approach which addresses each contractor's work as a separate population, and which considered time periods of good and known or suspected poor performance, would seem mor appropriate.

{-

In two cases the licensee premise did appear to be valid with respect to selection 1of_ contractors, but not necessarily with regard to contractor

,

performance history in time: 4-t

!- Assessment'of the concrete batch plant operations, wh'ich provided concrete to all site contractors, and r

' Assessment-of the concrete testing laboratory, which provided testing

- services to all contractor "

On the other. hand, concrete and reinforcing steel placement, curing and other related activities were performed by at least four different contractors with different performance historie To illustrate the vulnerability of improper assumption of homogeneous population, one.may consider the evaluation of a contractor who had a known

>

, poor performance history during a specific time period.. The work activities of this. contractor over-the period in question may be considered the population to be sampled. The fundamental Binomial Probability

, ' Distribution sampling approach.(referenced.by the licensee) would dictate a

.

[' sample size of 60. activities; absence of defects in.this sample would theoretically ascertain less than 5% defects in the population of 7 *.

i activities, with a 95% confidence level. But when the CAP considered similar work activities of several contractors as a single population, the CAP sample of 60 items from tLis population resulted in only a few selected i

'

, >

[

. w y , .

-

f . , . . _ . . . . , _ , , , . . . , _ . . .,, ,. _.;__.,,,m..,.,__.m_

-.

.

~

.

,

1. 7 for the aforementioned contractor for the questionable time perio It is difficult.to conclude that this few items would provide the same l conclusions as the aforementioned 60 samples ~ specific to that contractor !

and the troubled perio It"would seem that a higher defect rate, or a I lower confidence level may result', (presuming that not all contractor j activities were homo'geneous). I In view of"the above, it appeared th'at some .of the concrete module sampling

.

plans did not include representative samples (as suggested by Sections and 6.~0.A and .B of the NRC. accepted-January 1986 Construction Assurance Program Plan.) The licensee actions on this matter will be considered I during the. future review of the pending CAP Concrete Module Repor I Follow-up item 86-13-01.

, 8. Overview of CAP C'oncrete Module' Review Plan The review plan for CAP Module C3-02 (Revision 0) was approved by the CAP Manager on October 30, 1986. The plan described contractors a'nd their period of involvement in concrete work, and the amount of concrete work performed. It briefly described results of a review of past WNP-3 and

industry problems and mentioned several NRC past inspection finding (e.g. item 79/10-03, missing dowels, was noted and a check for missing dowels was included in a checklist for walking down exposed concrete blockouts and other unfinished work having exposed reinforcing steel). The l plan included 29 short checklists for conducting reviews, and mentioned !

several other reviews which were to be performed (checklists or actions j

, plans were not included). A general discussion of the random sampling '

approach was included, as was identification of sample selections. 'A matrix of FSAR commitments was included, identifying implementing specification paragraph numbers and identifying CAP checklists which addressed the general subject matter. A copy of each of the 29 checklists and instructions was included in the plan. Qualifications of-the review team were also described. The review plan appeared to be an accurate description of work to be performed by the team, and a record of

preparatory activities conducted prior to commencement of the review However, the review plan was imprecise in cross-referencing some prior CAP modules, which indicated a need for the review team to specifically consider the content of referenced prior modules relative to the current module under review. Examples identified by the inspector included the following:

L The C3-02 plan mentioned that Module C3-01 had included review of Batch

Plant Scales and equipment calibrations. However, the C3-01 review was I associated with soil-cement Earthwork, and the period of work considered did not encompass the full period that concrete activities were performed at WNP-3. (The team leader stated that he would respond l to the inspectors' concern via preparation of a new checklist number

,_ 30, which would consider batch plant calibrations for concrete work including water meters and admixture dispensers.) Follow-up item

'

86-13-02.

I i The C3-02 plan mentioned that Module C3-01 had included a review of the EBASCO Site Surveillance Program for contractors. However, a review of i

._ .

2 ~' ,* ' a w 2

. ,

.

.

,

'

i [- -

s

.

.' ; t : .

m d

.

,

-

- a+

.

.

,

(s-

8 h-- ,

,-y  ;

+

.,

~ .

, _ ,

s

.

- ,

. .

~

' ,fi . , .

-

.. , : . .

• =

. 'the reports.showed that few'of-the reports related to concret .

~

' '

. ,

,

,

-construction. See paragraph 12.~d for additional detail regarding this',

'

ite ~

,

-

.

. . . ,

"

' The review plan dkd not- cur enty address the impact on concrete ; work, ~of . ,

activities which had been deferred for consideration during future' modules.-

ExamplesLidentified by the inspector included the following:

'

.

. . . .

Conciete Anchor Bolts - Evaluation of anchor bolt installation was ' . ,

~ deferred.for consideration in several future modules. However, hole ^"

~

j

. drilling for anchor bolts' has potential .for damage 1to concrete .

'

reinforcing steel. ' Acceptability of completed concrete work' cannot be

'

fullyfascertained without assurance that anchor > bolt drilling was performed under well. controlled procedures.'. Follow-up item 86-13-0 ~ *

Embed Plates - Evaluation of embed plate installation was deferred for consideration in the future structural steel module. However, '

~

difficulty of concrete consolidation and field changes for reinforcing steel interferences:could affect the integrity of the concrete in-the vicinity of the embeds. Acceptability of completed concrete work in-such areas cannot te ascertained without assurance that complex embed

~

+ '

plate installations were performed correctly.' Follow-up item 86-13-0 '

'

The current review plan did not provide details of action plans for special1

_ reviews described in the review pla Contribution of these" efforts to assurance of concrete quality, and resolution of concerns expressed elsewhere in this report, could not be assessed by the inspector in-the-

'following areas (CAP treatment of these areas will be examined in future

~

NRC reviews):

'~ -* NCR Review -'An engineering review of nonconformance reports was described. -The nature of this review and planned use of findings was-undefined. -(Some licensee effort had been made to. categorize the past NCRs, but the review was described as not yet complete. See paragraph

1..e for^ additional details regarding this item.

,

Trending of Compressive Strength Tests - The nature of this proposed

,

review was not defined. Follow-up item 86-13-0 i l_

Review of Correlation Data For Truck / Pump Sampling Points - The nature.

/ of this proposed review was not defined. Follow-up item 86-13-0 ~

I *

Statistical Analysis of Rebar Tensile Test Results - The nature and - _ extent'of this proposed review was not defined. Follow-up item-t 86-13-0 Statistical Analysis'of Cadweld Tensile Test Results - The natureaof

~

this proposed review was not defined. ; Follow-up item 86-13-08.

,

< * Review' Construction Manager Actions Relative to NRC/QA Activities - The

{' nature of this proposed review was'not defined. Follow-up item

86-13-09.

l L

r

! .

.

, 9

.

. .

o

Review- of Applicability of ;IN-83-40 (relating to environmental qualification of epoxy grout) - The nature of this proposed review was not defined. Follow-up item 86-13-1 . Review Plan Omissions The inspector judged that the following matters were of interest to the NRC assessment of concrete-adequacy, but'were not included in the CAP Concrete Module review plan:

^ Core Drilling - Consideration of the potential impact of core drilling activities on the adequacy of installed reinforcing steel. Follow-up item 86-13-1 Bending of Reinforcing Steel - Consideration of past known problems

~

with proper bending of reinforcing stee (However, the exposed-rebar walkdowns planned by the CAP do include inspection for cracks in exposed bent bars.) Follow-up item 86-13-1 Grout Walkdown Inspections - Visual assessment of installed grouts to determine evidence of cracks, proper thickness, filling of cavities, and evidence of distres (The inspector observed cracks in non-shrink grout on safety related pump pedestals.) Follow-up item 86-13-1 . CAP-Concreta Review Team Personnel Qualifications'

The review plan description of personnel qualifications, and interviews of individual team members showed that management, civil engineering, quality assurance, and civil discipline quality control inspection capabilities had been adequately incorporated into the team. The personnel appeared

. qualified to perform their respective dutie . Overview of Significant Elements Review The licensee compiled a listing of elements which which can affect concrete and grout work quality. This included materials, installation, tests, and inspection attributes. These were then prioritized to reflect those which had greatest impact on quality and which could be readily assessed through review of existing records. The prioritized elements were then incorporated into checklists for the review of records and hardwar Many elements, important to obtaining adequate concrete, were not selected for direct incorporation into reviewschecklists, particularly where such items could be indirectly inferred through test results on completed wor For instance, various batch plant materials controls and materials tests were considered indirectly via review of concrete compressive strength test results for concrete that was placed into the structures. This approach placed a great deal of reliance on the validity of test results, and was accordingly augmented by review of test equipment calibration and control and test personnel qualification The resultant CAP checklists and special review topics were considered by the inspector relative to a similar preliminary list of important elements prepared independently by NRC consultants. Both lists were derived from , , , .

, ,

'

. . ,

' *

-

-

.

~

-

c

,

.

consideration of ACI and ASTM _ standards applicable to concrete. construction-of nuclear plants. The basic CAP listing appeared generally comprehensiv . 'The11icensee-;prioritized and abbreviated.this; basic list. The abbreviated

> list did not address some areas of NRC interest, as.noted elsewhere'in this

. report'. A more. critical evaluation of the the selected elements, and the associated nature of-reviews of each element, will'be reported after CAP completion'of reviews, special studies, and the final, repor '

12. Probles' History Reviews

The inspecto'r found that the CAP review preparation' included limited review of past' problems at WNP-2 and other nuclear facilitie In some cases the subject of the problems was included in CAP concrete. review checklists, alth~ough there appeared to be little evidence of root'cause analysis and focused probing to assure that the problems.had been resolved and nonrecurring.- There appeared to be no . synergistic evaluations of past problem records.to assure that work process weaknesses were isolated and

.not indicative of a need for more general management action which may not have been recognized and taken. The CAP management stated that the CAP purpose 'does'not include assess' ment of the adequacy of disposition of previously identified problems. The absence of a1 comprehensive evaluation of past problems and contractor performance history does not appear consistent =with the Sections 6.A'and,6.B of the NRC accepted January _1986 Construction Assurance Program Pla The'foliowingaretheresultsoftheins'pectorinquiryintovariouslimited problem' reviews which were conducted by:the CAP team; identified follow-up

'

items will:be considered by NRC during.. future review of the pending CAP Concrete Module report: NRC Previously Identified Inspection Findings The licensee /EBASCO quality assurance staff stated that action on-j- certain previously identified items appeared to have been completed,

'

but had not yet.been verified by the QA department (e.g. 81-08-03, 81-08-04, 81-09-02, and 82-06-01). The CAP management stated that the

.

CAP purpose did not include closing prior issues, and such issues were i not evaluated by CAP. These items involved quality assurance program i aspects which could impact quality of construction, such as item

.

81-08-03 relating to ineffective engineering review of contractor work procedures. Although it appeared appropriate for the CAP to defer action on hardware problems requiring future engineering / construction forces to correct, it did not appear appropriate to defer inquiry into i: the status of corrective actions on program matters which could have l affected quality broadly.

' Licensee Reported Construction Deficiency Reports (50.55.e)

!

[ >

A cursory review of the CAP office copy of the 50.55.e index showed i notations and handwritten summary sheets indicating CAP staff l cognizance of concrete related prior significant construction

! deficiency rep-ts. However, there was no evidence of CAP analysis of

,

prior reports of quality breakdown and effectiveness of corrective

,

, .

.

,

<

actions. Report number 35 involved significant concrete voids in the auxiliary building and corrective actions. Report number 25 involved cracks in welds of embed plates. The CAP efforts did not appear to have elements to address satisfactory resolution of the generic elements of these item The deficiency described in report 25 involved a vendor (Fought Company):

Use of partial penetration welds where full penetration welds were required,

Failure to perform specified post weld heat treatment,

Failure to. identify and reject assemblies with cracked weld The final documented corrective' action available in the licensee /EBASCO QA department stated:

All specific hardware items have been correcte *

Training of Vendor Quality Assurance Representatives completed

Vendor surveillance inspection plans revised to include appropriate inspection requirement These actions did not appear to address the generic implications of the inadequate vendor quality control, inadequate vendor surveillance training and checklists for other similar procurement In absence of other documentary evidence to support such generic issue, the licensee actions appear to have been incomplete. Since the CAP reviews of embed plates has been deferred to a future module, this item remains open pending completion of licensee generic actions. (Item 81-05-H). Construction Manager Records of Audits of Contractors The CAP review activities considered audit team qualifications, audit planning and preparation, documentation and follow-up of findings. The CAP checklist (CAP-D-GC-02-27) required assessing appropriateness and timeliness of corrective action, including stated actions to prevent recurrence. The CAP reviewer was required to attach copies of all audit findings to each completed checklist, fer information of the team leader during his review of the completed checklis The inspector examined the concrete related aspects of CAP completed checklists and attached audit findings for WPPSS/EBASCO audits of contract #265 (J. A. Jones Company) numbers 265-1 through 265-7; these involved an audit period of April 1980 through August 1982. The checklists noted acceptable audit performance with respect to the {

g checklist criteria. The review was completed by the reviewer on m November 5, 1986 and approved by the CAP manager December 8, 1986. The CAP manager stated that analysis of the audit review data was not yet complete, at the time of this inspection. The records reviewed by the CAP show that an aggressive audit program was in effect by WPPSS/EBASCO

.

.

'

, 12

-

during concrete work activities. However, it does not appear that CAP checklists for concrete activities have incorporated elements to specifically address certain known and suspected quality assurance / quality control problems revealed by such audits for contract

1. 265 activities between start of work in January 1980 through November 1981. Neither sample selection nor special reviews address the validity of contractor records and work quality during the period of the contractor's unwillingness to acknowledge deficiencies in quality control inspector knowledge and training and internal audit effectiveness. (The auditors had found demonstrated knowledge deficiencies in methods of concrete consolidation, installation of reinforcing steel, layout of concrete delivery equipment, and methods of checking vibrator frequency. Several of the current CAP checklists rely upon QC inspector signatures as evidence of satisfactory compliance of these items.). This issue had remained unresolved for the period October 1980.through November 1981, involving slowly escalating management attention culminating in a WPPSS/EBASCO August 1981 issuance of a show cause order to the contractor. The CAP reviewers appear to have relied upon the fact that the audit findings and management corrective action-request (CAR) were subsequently

" Closed" by WPPSS/EBASC0; independent CAP review apparently was not made to assure that appropriate evaluation of prior work had been conducted as part of the closure process. Additional CAP attention appears warranted to evaluate concrete work performed during this troubled period. Follow-up item 86-13-1 Records of Construction Manager Surveillances of Contractors In addition to periodic audits of quality assurance programs of each contractor, WPPSS/EBASCO conducted a program of quality assurance surveillance of each contractor. On a daily basis, quality assurance personnel would visit various contractor work activities in progress and observe conformance to program requirements. A daily report was prepared for each such surveillance, and the CAP reviewer determined that the logs showed 7346 such reports were generated. Through random number generation, the reviewer selected 60 random surveillance reports as a sample. The sample included 6 of the 681 reports for the contract

1. 265; only three (3) of these involved concrete work; similarly, one of 120 reports for contract #216, one of 96 reports for contract 219. The CAP review of these documents was' conducted using checklist number CAP-D-GC-01-16, and was completed June 18, 1986 in conjunction with the first CAP module (Earthwork). Review criteria consisted of verifying existence of the report identified in the log, determining if the ysurveillance report contained approval signatures, and ascertaining fthat the recorded discrepant observations contained notation / record of g follow-up and resolution. This review did not include efforts to

,

, detect and evaluate trends for concrete quality, but rather were

, oriented toward ascertaining that a project wide surveillance program

, did exist, surveillances were performed and documented, and mechanisms exercised to resolve identified specific discrepancies; thus the absence of emph' asis on concrete work, and consequent little value in assessing adequacy of concrete work. Follow-up item 86-13-15.

i

.

.

• 13

. 1 Nonconformance Report (NCR) Records The CAP reviewers utilized the existing project nonconformance report trending system to obtain a printout of concrete related NCRs, grouped j by_ type of' deficiency. At the time of this inspection, the CAP review and analysis of this data was not complete. There was no definitiv analysis to identify which of the NCRs were prepared by contractor QC inspectors prior to final QC acceptance of work, and which were prepared after final acceptance of the work. The inspector noted 200

~

NCRs categorized as missing reinforcing steel dowels, but the significance of this could not be ascertained without review of the-individual NCRs for specific discrepancy, time period (repeatedly occurring?), and noted generic corrective. actions. This observation was typical of several NCR categories e.g. see paragraph 15.f regarding

<

batch plant' review. Since the implementation of CAP checklists was about 75% complete at the time of this inspection, it appeared that a detailed consideration of the results of the NCR review had not been incorporated into the' concrete CAP review. The CAP staff did not l identify any' currently specific" plans'for completing the NCR review nor

'

method of application of findings. Follow-up item 86-13-1 . Nonconformance Report Examination l

For the subject of missing dowels, the CAP checklists include review ;

for a signature that preplacement inspections were accomplished I (including rebar placement); also one checklist includes a walkdown of exposed steel in>blockouts and incomplete placements; the inspector did not consider these to be a thorough exploration of this issue. Other data bases, such as old construction photographs, sketches on field changes and nonconformance reports, and other in process documentation would appear to offer the opportunity for a more thorough assessment of I reinforcing steel placements. The inspector examined one such NCR (N I 265-3845 and related 3864), for contract #265, which identified concrete void problems in the reactor building shield wall. The voids i occurred in an area of congested reinforcing steel and involved areas '

about 8 feet wide and 1 -1/2 feet high at both the inside and outside faces of the 3-foot thick wall. The NCRs included detailed sketches of the discrepant areas after excavation to sound concrete; the sketches showed rebar patterns in an apparent discrepant configuration. The rebar clearance criteria of ACI-318-71 and construction specification WPPSS-3240-412 had not been achieved (more than 1-inch clearance between bars). This appeared to be a contributing factor in the I resulting concrete voids from ineffective concrete consolidation during l

'

placement. The void was still exposed in the field, and inspection l showed inadequate clearances. Although the actual field condition varied somewhat from the NCR sketches, the representation of inadequate clearances was clearly vali Further review showed that the construction specification did not implement the ACI-318 Part 7.4.1 and 7.4.5 requirement for one nominal bar diameter clear distance between parallel bars (the specification called for minimum of 1-inch spacing).

l

-

i .

,

_

im The original inspection and acceptance of the reinforcing steel

" spacing / orientation" was documented on the Concrete Preplacement Checklist Record for Blockout F2 and F3, dated February 4, 1982; no

' discrepancies were noted. The NCRs which documented the concrete void repairs were dated by the EBASCO Resident Engineer office November 29, 1982/ October 25, 1982; there was no evidence of identification and evaluation.of the discrepant rebar clearances shown on the NCR sketches. The inadequate specification, the nonconforming rebar installation, and the failure of project personnel to recognize and correct the nonconforming installation appear to be violations of FSAR commitments and regulatory requirements. However, this item would be expected to be identified by the CAP reviewers'during the rebar walkdowns, yet to be conducted (detailed checklists have not yet been prepared). Follow-up item 86-13-1 The failure to incorporate appropriate quantitative acceptance criteria into the' construction specifications appears'to be a violation which would not necessarily be identified through the CAP process, which commences with the presumption that such details of the specifications are correc (Such discrepancies are not necessarily identified through the Engineering Assurance Program, which only samples specifications to arrive at probabilistic determination of acceptability of all contractor specifications.) Violation item 86-13-1 . Field Inspection of Concrete Grout The inspector visually examined the grout placements at safety related equipment in the reactor auxiliary building: Diesel Generator 1, Auxiliary Feedwater Pumps, and Containment Spray Pumps. Minor cracks were observed in several area. Particularly noteworthy were cracks in the non-shrink grout on the pedestal of the containment spray pumps. These cracks radiated outward from the base of the pump and it could not be readily determined if they extended under that base. These cracks were through the 3-inch thick grout layer, and at pump "A" a triangular piece (6" by 6") had broken off and was missing. The opposite corner showed a small piece broken off and ~ clear evidence of lack of bonding along the surface of the chamber.of the supporting pedesta <

The placement records designated " Proprietary Non-shrink Grout", and indicated that Masterflow 713 (Non-metallic, Non-shrink) had been use Acceptance of surface preparation and pre-soak were indicated by QC signatures on the placement records. Specification WPPSS-3240-470, Section 4, defined required grout materials and properties to be demonstrated by qualification test Part 4.02.I required testing to demonstrate proprietary non-shrink grout (cast in a 4" by 6" by 6" block) "not develop any cracks on the surface nor shall it have pulled free from the form at the sides". The observed through-cracks in the grout on the containment spray pump pedestals appeared inconsistent with such qualification, and lack of bonding reflected improper surface preparation. During the inspection, the project QA department could not produce any evidence of prior identification and planned action regarding this discrepanc During a subsequent January 13, 1986 telephone conference, the licensee identified to the inspector the existence of Quick Fix Project Change Proposal

-

.

.

, 15 PCP-19254'(originated in 1983). This document identified the damaged grout and noted that the grout manufacturer did not recommend such grout for the pedestal configuration for which it had been used. The PCP called for rework of_the installation; however, the PCP had been stamped " VOID", since it had been subsequently incorporated into the design drawings in accordance with procedure. With shutdown of construction, the contractor never implemented the rework, such that the deficient condition still existed December 13, 1986. The licensee stated that there was no PCP work status tracking system which would assure that the void PCP would be implemented. However, the licensee stated that eventual as-built walkdown activities would detect the discrepancy in the futur The reliance on a final as-built activity, in lieu of a nonconformance control system, appears contrary to QA program requirements to maintain identification of the acceptability of items and to control the ideu-ification and rework of nonconformin8 items. In the above case, the nonconformance is clearly not identified nor under control and relies upon future special activities for detection of completed work which is not in accordance with latest desig Failure to identify and correct the discrepant condition appears to be a violation of 10 CFR Appendix B quality assurance program requirement Violation item 86-13-1 This finding indicates the need for the-CAP to include grout walkdown elements in the program, in addition to the records reviews currently underwa See paragaraph 9.c, abov This finding also indicates that the general as-built condition of the plant may in some respects not be in accordance with the latest revisions of design drawings. Work may thus not actually be considered complete until as-built walkdowns are conducted to identify discrepancies, and implementation of construction to correct discrepancies (e.g. the PRIDE program mentioned in the CAP program plan). However, this is not a significant concern for concrete reinforcing steel and embeds, since pre placement inspections for concrete apparently included verification of conformance to both the latest drawings and interim design change documents I

(e.g. PCP's). This item will be considered during NRC consideration of future readiness review modules and at the restart of construction.

l 15. Findings Regarding Specific CAP Checklists The inspector and consultant reviewed specific CAP checklists relating to concrete and grout work, and with the following specific comments: CAP-D-C-02-01, 02, 04, 07, 08 and 09 (Compressive Strength, Delivery, Slump and Air Tests, Preplacement Inspection, Post Placement Inspection and Repairs, and Curing Inspection)

A detailed review was made of the data used by the Licensee for one of

, the 60 samples. A step-by-step review of the process used and the data l viewed was made for a second sample with the Licensee reviewer. A third sample was given a cursory review. All of the data examined by l

l i

__

.

.

~ ~

,

the inspector, and the response of the Licensee reviewer were acceptabl A few cracks in the concrete were noted during the walkdown observation for Licensee checklist 21. These did not appear to reflect an abnormal amount of cracking. Some cracks showed efflorescence which indicates that they are flowing water. It is considered that this should be added to Licensee checklist 28 (Preventive Maintenance) or reviewed separately for potential rebar corrosion. Follow-up item 86-13-2 CAP-D-C-02-10 (Cement Tests)

A spot check of the data within the Licensee sample of 60 was made and did not disclose deficiencies in specified requirements, CAP-D-C-02-11 (Alkali Reactivity and Abrasion Resistance Tests)

The data within the Licensee sample of 60 laboratory tests for both alkali reactivity and abrasion resistance was spot checked. Additional examination was made of the actual laboratory results of 35 aggregate reaction samples representing fine and course aggregate dispersed over the period of July 1978 through February 1983. The results of each test were verified on the Figure 2 plot from ASTM C289. All but one set of results plotted well into the zone of " Innocuous Aggregates" based on' reduction of alkalinity. The one exception was borderline between innocuous and potentially deleterious. It was noted that the numerous petrographic examination reports listed the need for the ASTM C289 reaction tests due to a significant chert content in both the Steilacoom and Weyerhauser aggregates. The C289 tests indicated low reactivit In addition, low alkali cement was used. The licensee's conclusion of no expected alkali-silica reaction was supported by the data examine CAP-D-C-02-14 (Rebar Tests)

A detailed review was made of the data compiled by the licensee for the sample of 60 rebar heat numbers. All but one heat was above requirements for elongation, yield and ultimate. The one exception was reviewed. The slightly less than specified yield had been dispositioned in an acceptable manne CAP-D-C-02-15 (Cadweld in-Process Testing)

A detailed review was made of the data compiled by the licensee for the sample of 60 cadweld strength tests. All results met or exceeded specification requirements. The sample dates were reviewed and verified as reflecting a broad range of tim This instruction only requires review of tensile test results of a random sample of 60 cadwelds of those which have been tested. Thus, it would not detect failure to perform qualification testing of cadweld crews, nor does it ascertain achievement of proper frequency of test splices (as shown in the logic diagram of the CAP review plan page 16).

The checklist does not ascertain that cadweld inspections were l

l

[

.-

L e *

-

' ,

,

,

.

3 , .

' .

,

is :d' .

-

'

Jaccomplishe'd.for.cadwelds and that inspection results were acceptabl Follow-up item 86-13-2 ' , .

f. CAP-D-C-2-19 (Batch Plant Certification)

This, checklist required CAP reviewers to ascertain that NRMC inspection checklists were completed and were' signed by a registered a engineer every two years.-

,

The~11censee's review of the seven batch plant checklists was' examine Each checklisti was intended to be valid for two_ years. Four had been filled out1for the main plant and three for the standby-plant. One.of the actu'a11 checklists was reviewed in detail. It reflected acceptable data with the' exception of hot water heating capacity. That had been resolved during additional review. Also,'this did net show up as low temperatures in delivered concrete during the review under licensee

.

checklist 02 (delivery).

>

Neither th'is, nor other CAP concrete checklists, appeared to call for verification of performance of mixer uniformity tests,

-

daily / weekly / monthly batch plant inspections, truck water meter calibration and evaluation of records of disposition of identified discrepant conditions arising from such inspections. These elements

_ provide the assurance that batch plant operations were uniformly-controlled so as to represent a continuous process for which production statistical sampling approaches would;be valid. The CAP manager' stated that a new checklist CAP-C-D-02-30 (see paragraph 8.a) would address weighing and dispensing equipment in the batch plant,' including water

,

-

meters and admixture; dispenser The inspector examined files of daily / weekly / monthly quality control

, inspections;and noted extensive records of monitoring by the batch plant contractor quality control inspector and quality control personnel _of the independent testing contractor. A minor discrepancy F was 'notedain the manner-of disposition of a' discrepancy in the calibration of batch plant ^ temper water. scales in the 1978-79 perio '

This appeared to be~of no consequence,.but was directed to the

,

attention of the CAP _ reviewers.' Also, the NCR categorization analysis showed multiple NCRs relating to batch plant operations (status indicated as, closed);' final conclusions regarding batch plant adequacy may need to consider an analysis of the implications of that data base -

3 (e.g. repeated significant findings). See paragraph 12.c for l- additional detail regarding~this item.

I CAP-D-C-02 20 (Cement Storage)

The_ licensee check of the sample of 60 cement storage inspections was t reviewed. The sample reflected a broad range of dates for the -two

,

'

plants and did not disclose deficiency information, t

I' '

l

!

i

,

7 , n

-

-- -- n

_

,

'

. . a', ,

'

,e -

-

7 27

'

,

.- 'gg:- ,

'

"

-

y: y ,

'

• * '

,

, ' -

, ; .

'

" '2, -

. ._ .3 _

- CAP-D-C-02~21.(Field Walkdown)

. . , '

'

. . . - The team performing. field walkdowns was observed for approximately two hours. . The work noted during the course of this. observation appeared

,

to be acceptabl '

i. . CAP-D-C-02-25 (Concr$te Inplace Strength Tests)

.

,

The licensee's work on this checklist.was incomplete at the time of this inspection. The rebound hammer investigation had been completed and the report thereof was examined. The inspector, identified aspects e of additional interest to the NRC reviews as follows: (Follow-up Ite )

. (1) The standard deviation numbers listed for hammer rebound results were not translated into meaningful information such as an indicated strength range to supplement the average indicated strengt ~

(2) The compression test results were not correl'ted a with the 28 day

cylinder breaks for the concrete actually cored in Unit'5. This would have established a strength increase factor which could be used for the indicated results at Unit 3 to correlate the rebound and 28 day cylinder dat (3) The consultants report did not list all report data called for in ASTM C80 (4) The competence of the consultant and machine should have been

' established by correlating the rebound data with the compressiv tests for the cores. The range of certainly of the results should be establishe CAP-D-02-26 (Concrete Block Construction)

.This instruction requires review of safety related wo'rk which 'was originally' designated as Quality' Class II. It calls for review of documents "if they are available. Absence of records is not a defined criteria for a negative finding." With such an ambiguous criteria, it is not clear how a statistically valid conclusion of acceptability or reject can be accomplished. Follow-up item 86-13-2 . NRC Records Examination in Addition to Licensee CAP Effort The inspectors examined the following' records and made examinations which

~ had not been addressed by the CAP Concrete Module Checklists: Concrete Pump Correlation Data The file of concrete pump correlation data for temperature, slump and entrained air was reviewed. The data was found acceptable and found s properly used in the, delivery QC checklists for several placement i

.

.

. ,

, Mix Designs Records were examined to verify the licensee's use of the specified mix design in batching concrete. Mix design 5103.75 was selected because if appeared frequently in'the sample of 60 placements selected for several of the licensee checklists. . A review of eight individual batch tickets distributed over an 18 month period showed consistent use of the designed mix proportions with the exception of the air entraining admixture. This was reduced in later batches but did not lead to unacceptable entrained air in the delivered concret Aggregate Gradation Records were spot checked to assess the gradation controls or both fine and coarse aggregate. The records generally show conformance with specification requirements. The deviations noticed were primarily in 1978 and, in each case, were dispositioned by satisfactory retests of a new sample. The specification conformance improved in 1979 and continued good through 198 . Management Meeting At the conclusion of the inspection a meeting was held with WPPSS management and staff at WNP- Personnel in attendance are noted (*) in Paragraph 1 of this report. The inspector summarized the findings described within this repor The WPPSS representatives reminded the inspector that the Readiness Review Program was a voluntary program by WPPSS and was intended to provide additional assurance of construction quality, over and above the confidence achieved from the existence of a quality assurance program during WNP-3 constructio The inspector acknowledged the WPPSS point and noted that WPPSS had ,

requested NRC participation in the program and had presented the CAP as a basis for NRC acceptance of completed work. The inspector advised that his in process observations during this inspection indicated that the resulting future CAP concrete report has the potential to be of insufficient depth to serve as the sole basis for NRC acceptance of completed work. The inspector noted that NRC may be able to conclude that completed concrete work was acceptable, however, such a conclusion would consider past NRC

routine inspection efforts and findings and may also require significant additional NRC efforts to supplement the WPPSS findings, particularly in the area of analysis of the implications of prior problem Subsequent to the management meeting, the Supply System mailed copies of additional records to the inspector, relating to the reinforcing steel clearances (paragraph 12.f) and grout damage (paragraph 13). Telephone conferences were conducted January 8, 12, 13, 1987 with participation of the EBASCO quality assurance and engineering representatives. The inspector reaffirmed his view that regulatory requirements had not been satisfied with regard to these items.

l

!

i