IR 05000498/1987056

Insp Repts 50-498/87-56 & 50-499/87-56 on 870831-0904.No Violations or Deviations Noted.Major Areas Inspected: Allegations Received by NRC That Affect Plant

ML20235T561
Person / Time
Site:	South Texas
Issue date:	09/25/1987
From:	Constable G, Renee Taylor NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20235T550	List: IR 05000498/1987056 ML20235T545
References
50-498-87-56, 50-499-87-56, NUDOCS 8710130130
Download: ML20235T561 (9)

Text

_

_

.,

1 ,

' APPENDIX U.S. NUCLEAR REGULATORY COMMISSION

REGION IV

NRC Inspection Report: 50-498/87-56- Construction Permits: CPPR-128-50-499/87-56 CPPR-129 License: NPF-71 Dockets: 50-498 "'

50-499 Licensee: Houston Lighting & Power. Company (HL&P) i P. 0.- Box 1700 Houston, Texas 77001 Facility Name: South Texas Project, Units 1 and 2 Inspection At: Bay City, Texas Inspection Conducted: August 31 through September 4, 1987 Inspector: "

& -4dhrf

.

. G~. Taylor, Reactor Inspector, Project Section Qhtd '

Reactor @rojects Branch

.

Approved: / Lonstable,~ Chief, Project Section C Y d7 Ddte '

Reactor Projects Branch l

l Inspection Summary l

'

Inspection Conducted August 31 through September 4, 1987 (Report 50-498/87-56; 50-499/87-56 Areas Inspected: Routine, announced inspection of allegations received by the

,

NRC that affect the South Texas Projec Results: Within the area inspected, no violations or deviations were identifie ,

C $b l

m_______.__ _ - _ _ . - - - - - - - . - -

a ,

DETAILS'

1. Persons Contacted Licensee Personnel R. Whitley, QA Specialist M. Whittaker, Supervisor, Metrology Laboratory B. Hall, Deputy Manager, SAFETEAM Other Personnel F. Miller, Welding Superintendent, Ebasco W. Lear, NDE Level III, Ebasco j The NRC inspector also interviewed other licensee and Ebasco personnel i during the course of the inspection.

l 2. Followup on Allegations l (Technically Closed) Allegation RIV-87-A-066 l

l The alleger stated that an allegation response by the South Texas Project (STP) SAFETEAM was not adequate regarding a calibration

'

l concer The alleger stated that a deficiency report (DR) involving l the metrology laboratory calibration program was issued, but the alleger was doubtful that the DR was sufficien Followup-communication between Region IV personnel and.the' alleger provided l information that the DR largely concerned the work of. one. laboratory calibration technician. The inadequacy in.the DR was'in regard to one calibrated tool for which'there may have been two files or that there may have been two tools with the same number and, therefore,~only one

, record. A portion of a number was provided to the' Region IV followup I contac Via a computerized cataloging system, SAFETEAM was able to provide the NRC inspector with a number of concerns they had received and investigated concerning calibration activities. By scanning the ,

SAFETEAM files on these concerns, the NRC inspector was able to 'l isolate one concernee case with a strong similarity to the allegation l stated above with the exception of the information provided to-  !

Region IV during the followup contact. The SAFETEAM file indicated the concern originated with a letter received in the mail during the latter part of January 1987,. from a person who had to have been familiar with work in the metrology laboratory. The principal thrust of the letter was that the supervisors and leads in the laboratory needed to learn how to exercise their authority more effectively. Th concernee stated that if the supervisors and leads had done the job, the DR (unnumbered and unnamed) would never have been necessar . - _ _ _ _ _ _ - _ - - _

._

,

-

'

.

3- 3 During an interview with~the,SAFETEAM investigator'who had worked on ,

the concern, the NRC inspector was informed that an undocumented

~

!

contact with the concernee had been made in~ order to gain more ,

.

information.- The investigator stated that the concernee had stated that they didn't want to discuss the matter further and that as a result SAFETEAM had worked entirely ~from the information contained in the letter. .The SAFETEAM file contained a copy of a DR initially issued in late December 1986 which required the examination of the work of one, then former, calibration technician based on severa identified instances where the technician had failed to follow detailed calibration procedures. The DR documents a review of over 1000 calibration actions (number stated but individual actions not listed) with de' tailed findings related to 140 actions that were in:

some way deficient. The DR does not indicate any duplication o ,

numbers although there are listingc of identical devices, each with  !

its own unique number. In many instances, the total number consisting of 10 digits, changes in only one position to reflect a different device or a different type of device. Review of.th" listed deficiencies indicates that the majority were in the following-categories:

Calibration to a full-scale tolerance rather than a specified tolerance at the instantaneous reading point. In most. instances j the specification. sheet for the device was inccrrect and the calibration had been done correctly even though at odds with the requirement.

, Failure to list the calibration standard on the record for In

!

most instances, only one standard was available and device recalibration to that standard indicated the device was within tolerance except for three instances where the recalibration indicated the devices were not within tolerance and required I further evaluation by the users of the device '

Miscellaneous failure to complete the pre printed calibration ,

record farm in accordance with procedure ?

The DR alt.o states that a random sampling of approximatel '

1000 calibration actions by other calibration technicians was made l which in61cated a lower error rate (50 errors identified) but were '(

comparable to the type of errors made by the technician discussed abov {

l The DR disposition indicates that training was provided to the technicians on the need to pay attention to the detail requirements in an effort to reduce the error rate. The SAFETEAM file indicated that-their investigator interviewed the technicians in regard to their 1 perceptions of supervision. A review of interview notes by the NRC inspector indicated that 12 technicians were generally satisfied with

, their supervisor and/or that supervision was improving to the point of being satisfactory. Two were annoyed by the attitude of supervision.

'

_.___________ ___-._ -____ - _ -

- _ _ _ ___ _ - - _ . . , . _ _ _ . - - - .-.

,

1:

. . .

.

}

but did note improvement. Two.~others were not. satisfied with .I supervision or' working conditio'ns although'one of these did

'

acknowledge some recent improvement. The interviews were conducted-about 6 weeks after the concernee's letter was received by SAFETEA The NRC inspector toured the metro' logy laboratory during.the course of-the inspection in the company of a supervisor.'.,During the tour, the NRC inspector interviewed the supervisor, one lead technician,- and one calibration technician. The supervisor informed the NRC inspector that the technician, who was the primary subject of the earlier; discussed DR, was a contract employee with no nuclear experience but with considerable prior experience'in the testing land calibration of equipment used to measure. vibratory motion'as in rotating equipmen It was. understood that this particular technician,was.the only one available who had knowledge of setting up~the-components.used as standards, establishing the procedures, and calibrating the-field-use measuring device During the latter part of 1986, the' licensee decided that it would be necessary for .long-term operations 'to have, li only licensee personnel in the _ laboratory. Some of the contract employees accepted. direct employment while others did not; among the latter was the above discussed technicia Another technician was employed with knowledge of the vibratory motion equipment who then detected and reported'the errors of.his predecessor which in turn resulted in the DR as it'was originally issued. In an interview with the "new" technician, the NRC inspector.was informed'

that there were errors in the preparation of specification .

]

requirements, in the detailed calibration' procedures'.and;in subsequent application of the procedures. .Thei"new" technician said that the'

errors, when taken individual _ly, were not-significant but could not be allowed to endure, particularly'where the.' specification and procedure were involved. It appears that. laboratory supervision'have been substantially dependent'on the skill ..and integrity of both technicians in regard to the vibratory motion equipment, a situation'not considered unusual in a diversified metrology laborator The NRC inspector was provided a demonstration of a computerized equipment inventory control' system which is used to assign equipment serial numbers. The computer rejects any attempt:to reuse any given number of previously. input numbers which should' prevent two items of ,

equipment from having the same number. There is,'however, a manua :

operation involved in transferring the computer control. led number'on

~

!

to the equipment and another in preparation of the permanent equipment-calibration record. Either manual operation _ could give rise to an occasional error. The NRC inspector was informed that there are approximately 7500 items in the laboratory control syste '

During a subsequent and unaccompanied interview'with a clerk in the- '4 laboratory office where the records are filed, the NRC inspector was '

shown the record filing system. The clerk.seemed convinced that anything but a short-term error of two items of equipment being under ,

- _ _ . _ - . l4 .<(

- * .

a

'

5' {

i l l* (

.one record.or of there being two records'for'one item of equipment '

would not go undetected.g The NRC inspector judged that the clerk was s a l

probably correct after reviewing selected files and the system of

'

.I file ( j

< , f

'

In regard to the ' partial serial number provided the NRC, the NRC'

inspector learned during the interview with the clerk that the portion ,

available was only the precursor identifier of one of over 6500 items

,

L of construction-related equipment such as torque wrenches and wire i terminal lug crimping tools in the calibration procram and that )

without the final four numerals of the total identifier, it would be  !

difficult to locate the situation indicated by the allege Near the close of the inspection, the metrology laboratory supervisor contacted the NRC inspector to report that a tool had been identified during the previous evening's calibration work that' duplicated the number assigned to another identical tool. The tool's were field use measurement caliper The NRC inspector examined the tools and found l

'

that a number etched into the tool blade with a vibro-tool had been etched out and a slightly different number added. -By holding the tool i at a proper angle to light, it was possible to read the original 4 number through the over etch. The original number was apparently to I have been a 10-digit number but had been. etched.% with 2 zero's missing, each in one field of the number. The final digit in the l number which would have distinguished it from the otherwise identical- i tool was 2. When the full 10-digit number was etched in, the zero's  !

had been added and the final digit had become a 3, the same as the i final digit on the cther too The duplication had been detected by a file clerk attempting to file a record for a recent calibration on one {j of the tools and found that the new record duplicated another action >

on the other tool by a different calibrato The NRC inspector would note that construction-related tools discussed above use an 8-digit

,

number of entirely different construction that used on tools used by l all others which are of 10-digit type. The calipers were used by licensee startup and maintenance personnel rather than construction  :

personne )

L In conclusion: 3 j

SAFETEAM concluded thec the allegation as they received it in the letter was not substantiated, apparently based on their review of'

the interviews carried out by their investigator. The NRC .

inspector judged that there was sufficient substance to the  !

allegation to consider it substantiated since the interviews took place several' weeks after the alleger apparently left the sit A majority of the interviews while not expressing much concern with supervision at the time of the_ interviews, none-the-less did i indicate that supervision had recently improved with the ,

implication that at an earlier time, there may have been a supervisory problem. It also appears that the DR could have possibly been prevented had the supervisors or' lead hersons been i

,

l h

. _ _ _ _ . _ . -

. _ - .

,

, . , o q

..,

<

l .( .!

!

< s i

. more expert in the setup and calibration of, the vibratory motion-equipment and thus prevented the errors of the original 9 technician before they became prevalen .f 1

Without more information,.the~NRC inspector was unable to l

determine in what way the DR was thought to be ' insufficient since- '] i

it' appears to have been a through treatment of the problem ofi j!

l detecting and correcting the carelessness of some of the calibration technician !

The NRC inspector.was not able.to substantiate that two a construction type tools as distinguished by the alleger were duplicated either in serial number or record.. It was established J that the possibility existed where any device in the program l

could be duplicated in serial number or in the record system if i an error was made by the person responsible for applying either

'. the number to the device or.to the record. However, such errors, should they occur, would be quickly identified and resolved as a !

( result of established administrative controls.

I (Technically Closed) Allegation RIV-87-A-0068 This allegation involved a concern that NDE Level II inspection of 'I welds prior to February 1984, may not have been accurate. The system for finding welds was " changed" to require the check out of controlled drawings to find inspection areas. Discussion between Region IV personnel and the alleger indicated that the application of liquid penetrant TPT) or magnetic particle (MT) nondestructive examinations (NDE) by the Level II examining personnel may have been _i to the wrong welds. It was further understood that this concern i applied to ASME Section III pipe welding activitit.s since the alleger 1 stated'that the N-5 certificate forms might be inaccurate. The N-5 l form is an- ASME Section III form used by the " installer" to certify 1 his wor l The NRC inspector took the statement "that the system had been changed" I to mean that the procedures in effect up to February 1984 had been !

changed from whatever method had been used by the NDE personnel to i

'

locato selds to some other method involving use of controlled drad ng The NRC inspector found that relatively new procedures in the Site Standard Procedures (SSP) series had been issued, but generally 1 referenced the older procedures being replaced. Using this information, ')

the NRC inspector was able to review archieved procedures until he located Ebasco Site Procedure CSP-96, Revision 0, " Request For NDE."

Revision 0 to this procedure outlined the method by which construction l craft personnel communicated the need for NDE personnel within the '

Ebasco Quality Control (QC) organization to perform their examination ;

of given weld. The procedure required that the craft'make out a  :

" Request For NDE" form which provided for the weld identification and '

the location of weld which could have taken several forms such as by ,

building an.1 floor elevation or by room identification. The procedure b

to

. ~<< X7

,1 j, J 4

, ,

,

e .

p 4

/(

.

'

'

e ' '

L^1, i

p V

7 .

-

/ <

,

y y e

. . L\ dk i( I

.

'

alscqrequired the craft tcfattach an kPTlldentificatW/ Location" form on avnear the weld vid to'&lso flag thtNeld by tying'aiolo' red N +

ribbon around the pje near the weld. T k attachmwt-of)the form and (

,,. n of ts , ribbon only applied 6, 20se jo i the applicr @S T) and byJsilence, dia'not appiy to radiography Mr.L9Upon pQ,ints requiring inquiry, the licensee proVided a revision to csp-96 accomplished in i

< June?p84 by an "14terim Change To A Procedure." The change required thit fn the fvture,"all welds gold hye the identification form'and  ;

"

thefibbonfidgattachedifthefrequiredNDE. The 4tated reason fo .

the change on the IrMerim Chhuje reccM was to " resolve Comment No 3 '

?

j to Kdmper (ANI) 1984'dudit and QCR Hi.' I to EQS 3847." The originator j , of the change was then and is nov the Ebasco Welding Superintendent. '

The NRC inspector briefly interdewed this~ person who vo7unteeredz to produce the documents. reference #on the change document. dmete are s1 )

produced along with several related documents. The ( Nsi. docums t.was N a Kemper Group Program Monitoring Report dated February'll, M B4, frc,( *z \ 'a the ANI (Authorized Nuclear Insprctor) to the ther, Ebasco Qulity /^ ?

Project n%gv with several attachments reference attachments had three handwritten " list of observations." One Observa of'the , biot No. 2 stated 'il feel the procedtges use:1 for NDE examiners to idensity

.

l- ' weld is lacking. The exa:ainer coald not prove to me that the weld he t

examined was in fact the weld i:e yad b.een requestet to examin Everytime I go on a NDE hold print, the marking of ine welds is i differen A pro:edure addressing how NDE examiners ' k ate proper

"

i welds needs to be generated." Observation'No. 3 on t6e Aame

. / attachment indicated the writers dissatisf action with, the . fact that sofficial controlled copies of NDE procedures were not A su'ed to the

, llDE examiners but rather only " Info" copies. Other a tLchments to the

focument package indicated that Ebasco had initially rebuffed the ANI by stating that their_ procedures were adequate and by saying that Ebasco was not contemplating any~ changes until objective evidence was available to 4ng that an incorrect weld had been NDE examined. The Kemper Group kWnded to this statement by a letter from thylr regional offise itating that Comment No.' 3, of Program Monitoring Report 3659, would require corrective action. .The Ebasco response was to state the CSP-96 vpuld be revised to correct.the situatio The Kemper regional office. letter, however, characterized as Comment No. 3,

.

thecommenttht,had[eennumber2intheFebruary. report.

i

'

ESQ-384 which way also referenced'on the Interim Change to' CSP-96 l related to Ebasco abdit findings that the craft were mot properly l completing the " Request For liDE" forms. - The revision to CSP-96 included a requirement that both the craft. supervision and the welding QC persennel had to approve the form to assure accuracy and compirpnes During the interview with the Welding Superintendent,' the NRC inspector learned that the ANI individual, who wrote the February 1984 Program Monitoring Report, was no longer on site and had not been since November or December 1984. The NRC inspector then sought out and interviewed the present lead site Af!I ,tssigned

to Ebasco in an'

- .

l r

- -

,

_ _

s,5.~. .,

,

,

,

i 8 i effort to get a further understanding. ;This person stated that he.had ~

become the lead ANI in January 1985,.but professed to have no knowledge of the earlier ANI's concern. He then researched his files of program monitoring reports until he found No. 3659 which had most j of the correspondence attached previously provided by Ebasco. In-addition, was a handwritten note to Ebasco QC management indicating acceptance of the revision to CSP-96,-but also stating that until: the-ANI's became satisfied that the welds examined prior to the revision I were the welds that were supposed to have been examined, the ANI's would not sign any acceptance documents. including the N-5 forms. The j l

NRC inspector was aware that the N-5 forms for STP Unit 1 piping systems had been accepted by the ANI in all instances since 1985. The present ANI did not appear to be concerned with the fact that his j

'

predecessor had taken a contrary positio To gain a better perspective of the situation, the NRC inspector '3 interviewed a mmber of licensee and Ebasco personnel who had been i employed in the ASME pipe installation an NDE activities in time j earlier that the beginning of 1984. It was learned that prior to 1 mid-1985, field welds (those made to~ install prefabricated piping ,

assemblies) carried no other identification.than the identification of 1 the welde In mid-1985, because of a problem involving the radiographing of an incorrect weld, it was decided to "hard-mark" the pipe adjacent to a field weld via steel stamp _ impression or by vibro-etching in the weld number. The ASME Code allows that all -

welding traceability information can be maintained by documentation as ,

an alternative to marking the pipe itself. The documentation method I

,

was that selected by Ebasc '

l The NRC inspector then sought out and interviewed _three Ebasco NDE personnel who, by their records, were employed in the same capacity

'

during 1983 and 198 They each stated that all welds in that era were marked or flagged by the craft with marker pens, the identification tags and generally by the ribbon flags and each stated that when the welds were not so identified, they did not perform any examination but rather gave the request form back to the craft. One did say, however, that they would on occasion seek out the craft pipefitter or welder to show them where a weld was in the event that the marking and/or flagging had not been done. Two of the three i stated they were not aware that prior to mid-1984, the craft had only been required to tag and flag welds scheduled for RT since common practice had been tag and flag all welds scheduled for NDE regardless of the type. These three interviewees all expressed confidence that the welds which required NDE had all been examine At the request of the NRC inspector, the Ebasco NDE Level III performed a limited statistical study of all Code class welds made between January 1 and June 30, 1984. His study showed that in this period,1863 welds were examined by either PT or MT with a combined -

rejection rate of under 4 percent. He stated that his records indicated nearly all were accepted after mild surface conditioning of

___

• e, .

i 9 i q

the welds'which removes small anomolies that provide indications in either metho Very few required.any rewelding to become' acceptabl Generally welds made in the piping materials used in nuclear facilities with typical welding processes will experience low rates of rejection when examined by either PT or MT since both are essentially capable of only detecting flaws at, or very near, the surface of wel ~

Most welders will experience their greatest difficulty with a particular weld in the root zone and hence most weld defects 'are in the root area on the.inside of the pipe.' .The code requires that.only the accessible surfaces of the weld be examined which generally restricts the examination only to the~ exterior face of the weld except for a few welds in very large piping such as that in the reactor loop.-

The NRC has conducted a minimum of two inspection of the welding and i NDE activities during each year of the construction of the South Texas Project. Two of the inspections have_ included major efforts by the NRC NDE van personnel. The first " van" inspection was conducted in January 1984, before Procedure CSP-96 was revisedi This inspection involved over 600 inspector hours and was specifically directed at reviewing and reverifying all aspects of the NDE activities. A second

" van" inspection was accceplished in October 1986, which again covered much the same activities except that the revised procedures were in place and included the hard marking of the welds. In October and November 1985, the NRC Construction Appraisal Team also performed a significant inspection which encompassed a programmatic review and a reverification inspections of the NDE process. None of.these inspections have provided evidence that welds that were required to be, NDE examined were no The NRC inspector concluded that there was and is a, possibility that a weld other than a field weld.could be mistakenly examined through some error in identificatio In all likehood, had this error occurred during the era prior to the hard marking:of the' welds, the inspection report for the examination would have reflected either the information on the Request For NDE Examination form or the tag attached near the wel The only method of detecting this type of error would be for the weld to be shown to be defective by a subsequent reexaminatio None of the above NRC inspections have detected defective welds by reexamination using the original methods nor given the low rejection rates during the original examinations, is it likely they would hav In conclusion, the allegation could not be conclusively substantiated for reasons given above, but had it been substantiated, it would have had little safety significance since there is no.. indication that the welds are in fact defectiv . Exit Interview No formal exit was held relative to this inspectio The licensee QA Specialist was informed relative to the NRC inspector's observations and salient facts learned during the inspection.

- _ _ _ _ _ _ _ _ _ _ - - _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ . - - -. - _ _ _ _ - -