ML20128N778
| ML20128N778 | |
| Person / Time | |
|---|---|
| Issue date: | 01/31/1993 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| NUREG-0040, NUREG-0040-V16-N04, NUREG-40, NUREG-40-V16-N4, NUDOCS 9302230396 | |
| Download: ML20128N778 (198) | |
Text
_
NUREG-0040 Vol.16, No. 4 i
l Licensee Contrac~or anc1 Venc or Inspection Status Repor':
Quarterly Report October-December 1992 U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation 1
a " " %,
a
?
1mf s
4,,.Y ;/
j j
l
Available from Superintendent of Documents U.S. Government Printing Office Pcrt Office Box 37082 Wast, igton, D.C. 20013 7082 A year's subscription consists of 4 issues for this publication.
Single copies of this publication are available from National Technical Information Service, Springfield, VA 22161
]
l l
l
NUREG-0040 Vol.16, No. 4 Licensee c,oir:rac:or e
anc Vencor Inspec: ion Sta:us Report Quarterly Report October-December 1992 U.S. Nuclear Regulatory Commission Office of Nuclear Iteactor llegulation l
l t ""'Go S
b#
s, 9302230396 930131 PDR NUREG PDR 0040 R
i l
l l
l l
Available from Superintendent of Documents U.S. Government Printing Office Post Office Box 37082 Washington, D.C. 20013 7082 A year's subscription consists of 4 issues for this pubhcation.
Single copies of this publication are available from National Technical information Service. Springfield, VA 22161 1
a
,~
(
)
~
+
- ',, J M
i Voli 16, No :4 '
y sp f
v Licensee Contractor 4y and: Vendor Inspection-
[
Status Report
'8; a-
$)J Quarterly Report
}
October-December 1992 F
4 f-1 Manuscript Completed: January 1993 Date Published: January 1993 i
'f i
t Division of Reactor Inspection and Licerdee Performance 4g Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission 1-Washington, DC 20555 j-_
4 4
};
e".
j
)
?I 7
h b
5.-
A h-llp i,
.i
- i. -
0
- s
. o-
^}
_i s
s I
ABSTRACT-w:
3
-This-periodical covers the results:of-inspections; perform'ed by; the NRC's Vendor: Inspection Branch that have been distributed =to'-?
the1 inspectediorganizations during-the period from-October-'_1992.:
through Decembery1992.
~
i l'
l i
I l
l' I
l
-t 7
111 l'
TABLE OF. CONTENTS-a PAGE-Abstract,................................_
,.......,.......,__11.1-Preface vil_.
' Inspection Reports.......................,,........,,..,,.
ix-Index.................................................,.....
. x ij Selected-Bulletins, Generic Letters, and Information'dotices Concerning Adequacy of Vendor Audits and_ Quality of--
Vendor Products....................................,......
179-Correspondence Related To Vendor Issues 180 E
O W
i i-V
=
. PREFACE
- i
- A fundamental 3 premise of the Nuclear; Regulatory. commission'(NRC) licensing and inspection-program is that licensees 1are; responsible-forithe-proper construction and safe and' efficient operation of-their.. nuclear power-plants.
TheMtotal' government-industry? system-for.the-inspection of commercial nuclear-4 facilities has been designed to provide-for multiple l levels:of--
H inspection-and-verification.
Licensees, contractors, and, vendors 1 each participate in e cuality verification process-in compliance-with requirements-prescribed by the NRC's' rules and regulations:
)
(Title 10 Code of Federal Regulations). - The NRC performs 1an overview of the commercial nuclear industry by inspecti'on to=
determine whether its requirements are being met by licensees'and-their contractors, while the major inspection effort is performed by-the industry within the framework of-ongoing.aualit) verification programs.
The licensee is responsible for developing and maintaining a detailed quality assurance (QA) plan with implementing procedures pursuant to 10 CFR'50.
Through a system of planned and periodic.
audits and inspections, the licensee is' responsible for assuring that suppliers, contractors and vendors also have suitable and appropriate quality programs that meet NRC requirements,-qaides, codes and standards.
The Vendor. Inspection Branch (VIB) reviews and inspects nuclear steam system suppliers (Nssss), architect engineering- (AE) firms,-
suppliers of products and services,-independent testing.
laboratories performing equipment qualification tests, and holders of NRC licenses (construction permit holders'and-operating licenses) -in vendor-related areas.
Thece inspections are performed to assure-that the root causes of reported vendor-related problems are determined and. appropriate corrective actions are developed.
The inspections also review thelvendors'-
conformance with applicable'NRC and industry quality.
requirements, the adequacy of licensees' oversight of their vendors, and that adequate interfaces exist between licensees and vendors.
The VIB inspection emphasis is placed on the quality and suitability of vendor products, licensee-vendor interface,.
environmental qualification of equipment, and review of equipment problems found during operation and their corrective action.
When nonconformances-with NRC requirements and regulations are-found, the inspected organization is required to take appropriate-corrective action and to institute preventive measures to preclude recurrence.
When generic implications are' identified,1 NRC assures that affected licensees are informed through vendor reporting or by NRC generic correspondence such as information-notices and bulletins.
This periodical -(White Book) is published quarterly and'contains
~
copies of all vendor inspection reports' issued during the calendar quarter for which it is published.
Each vendor
^
v11 r--
2
/ nspection report lists.the nuclear l facilities-i to which:the:
i results are= applicable-.thereby-informing licensues~and vendors of
-potential' problems.
In addition, the'affected RegionalJ0ffices are notified of any significant problem areas-thatfmay require:
specialiattention. ~
The White' Book also contains a list of selected bulletins and!
information notices involving vendor issues.; copiesfof.other
-pertinent-correspondence involving vendor Issues are also:
2ncluded in this White Book issue.
Correspondence with contractors and vendors relative to inspection-data contained in the White. Book is placed in tl.e 10SNRC Public Document Room,- located-in Washington, D.C. _
-l l'
l-viii t
~.......
~..
,_ tr',-
-r.
4-
+
2
'^-
- ~_
m 3
s:-
c 4I
',?
m i
.t
.c L
INSPECTION REPORTS:
I 4
A.
E c
T
~
f i.
T 1
)
a 4
'I
- ,(
s 7
I 4
d 1
+
a s
[ -;
-c l.
M
~.:,
y ix i'w'9-e v 14 at
-4
=---t-ey et vv bww.--
4-m.-
,w.,g-w.w-e-e yvww
- Sew,
--,-4'-4
+-+yr.--#
1
'fwt - Y-
-ew.e
- I'NDEX
'I PAGE FACILITY:
-REPORT NUMBER?
=ABB Combustion Engineering-199900401/92-01; in Nuclear Services Windsor, Connecticut Bechtel Corporation 99900519/92-01' 16-Gaithersburg,-Maryland Coltec Industries 99900300/92 28
-Fairbanks Morse Engine Division Beloit, Wisconsin Connex Pipe. Systems, Inc.
99901255/92-01 45
-Marietta, Ohio Exide Electronics 99900366/92-01.
<6 4 -
Raleigh, North Carolina ~
Iowa Electric Light 05000331/92-201 76) and Power Company
' Cedar Rapids, Iowa Lisega GmbH 99901235/92-01--
101 Zeven, Germany Philadelphia Electric Company
.05000352/92-201
'114 Wayne, Pennsylvania
-05000353/92-201-Rotork Controls Limited-
_99901066/92-01L 147:
g
. Bath, England 1:1
_ estinghouse Electric Corporation 99900404/92-02; 159:
W Nuclear and Advanced Technology-Division Pittsburgh, Pennsylvania l-xi
f ** %,
t' i
UNITED STATES f,('frAd [!
3 i
NUCLEAR REGULATORY COMMISSION.
W ASHINGTON. D'.C. 20666 -
\\q ' %.
f
/
e...*-
November-20. 1992.
Docket No. 199900401 Mr. StevenLA. Toelle, Manager Nuclear Licensing ABB Combustion Engineering Nuclear Services 1000 Prospect Hill Road Post Office Box 500 Windsor, Connecticut 06095-0500.
Dear Mr. Toelle:
SUBJECT:
NOTICE OF NONCONFORMANCE (NRC INSPECTION REPORT 99900401/92-01)
This letter addresses the inspection of your facility at Windsor,-
Connecticut, conducted by R. C. Wilson and S. D. Alexander of.
this office on September 28 - October 2, 1992, and the discussion of their findings with you and members of your staff on October 2, 1992.
The_ purpose of-the inspection was to review-your dedication of commercial grade components for nuclear safety-related applications.
The inspectors reviewed three groups within-ABB Combustion Engineering Nuclear Services (CENS): -Nuclear Spare Parts, Instrumentation and Control (I&C) Engineering, and Mechanical' Engineering.
Areas examined during the NRC inspection and our.
findings are discussed in the enclosed report.- This inspection consisted of an examination of procedures and records, interviews with personnel, and observations by the inspectors.
The inspectors fout.d that the implementation of your.QA program failed to meet certain U.S. Nuclear Regulatory Commission (NRC) requirements. _Specifically, CENS failed to adequately specify and implement design and procurement requirements for packing l
assemblies supplied-for' safety-related use in a' charging pump supplied for a commercial nuclear power plant.
CENS allowed-l shipment of parts-containing natorial which the licensee's purchase order specifically prohibited.
CENS also accepted improper certifications from commercial grade suppliers containing obvious errors, such as common cure dates and ten-year-shelf lives for a variety of plastic and-metallic materials, and the same batch number for aramid and polyethylene plastics, g
The specific findings and references to.the pertinent l
requirements for the above nonconformance are identified in the I
enclosed Notice of Nonconformance (Notice).
! L
Mr. Steven A. Toello The inspectors also found that your dedication-of various commer-cial-grade items was not clearly documented, largely because the dedication path was not clearly specified.- However, in most cases you were able-to satisfy questions by providing additional information not present or referenced in the original files.
The response requested by the enclosed Notice'is not subject to the clearance procedures of the Office of Management and Budget as required by the Paperwork Reduction Act of 1980, Public Law No.96-511.
In accordance with 10 CFR Part 2.790 of the NRC's " Rules of Practice," a copy of this letter and its enclosures'will be placed in the NRC's Public Document Room.
Sinc rely, V Lei J
orrholm, Chief-Vendor Inspection-Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation
Enclosures:
1.
Notice of Nonconformance 2.
Inspection Report 99900401/92-01 -.
-~
- l J
ENCLOSURE 1 e
NOTICE OF NONCONFORMANCE ABB-Combustion Engineering Docket No.
99900401/92-01 Nuclear Services Windsor, Connecticut Based on the results of an inspection conducted on September ~28 -
October 2, 1992, it appears that certain of your activities were not conducted in accordance with NRC requirements.
a.
Criterion III, " Design Control," of Appendix B to 10 CFR Part 50 requires, in part, that measures be established to assure that the design basis for safety-related components-is correctly translated into specifications, and that design control measures be applied to certain items including the compatibili"y of materials.
Paragraph 2.3.1 of System 3,
" Design Control," of QAM-100, "ABB Combustion Engineering Nuclear Services Quality Assurance Manual," Third Edition, Revision 2, dated-July 27, 1992, requires, in part, that the cognizant-engineer,.using the design input as a basis, shall select and prescribe appropriate materials essential to the. function of the component.
Contrary to the above, CENS failed to ensure that design output documents (specifically, CENS Purchase Order (PO)
No. 9206246, recently issued to.UTEX Industries, Inc.) for plunger packing assemblies for charging pumps for the Waterford Steam Electric Station, Unit 3, correctly incorporated the requirements of the design input (specifically, Entergy Op6 rations, Inc., PO No. WPO46223).
The Entergy PO excluded cartain materials, including tin and lead.
Although the CENS PO to UTEX repeated this requirement, it also specified UTEX Drawing B-2870-C, which shows the step bushing in the-packing assembly to be made of l
a bronze alloy that contains tin and lead.
b.
Criterion VII, " Control of Purchased Material, Equipment, and Services," of Appendix B to 10 CFR Part 50 requires;'in part, that measures be established to assure that purchased material conforms.taa the procurement documents and that l-these measures include provisions, as appropriate, for objective evidence of quality--furnished.by the contractor or subcontractor.
Paragraph 2.1.2.b of System 7,
" Control of Purchased Items and Services," of QAM-100 requires.that procurement activi-ties be conducted in a manner which controls evaluation of' -
- -. ~ _
-.. - =.
.. ~
~
objective evidence of quality furnished by-_the supplier.-
Paragraph 2.10.6.c of System 7 of QAM-100 requires that prior to acceptance of a commercial grade; item,_-Quality control shall verify that documentation applicable to-the item was received and_is acceptable.
Contrary to the.above, CENS_ failed to control the" evaluation of objective evidence of quality furnished by the supplier, and failed to verify that' documentation appilcable to the item was adequate.
Specifically, the CENS review of'UTEX-certificates of conformance (upon which CENS relied'for acceptance of the packings). failed to identify that certain of the certifications (1) were inapplicable to the parts for which the certifications were made or'for which required data were provided, and (2) lacked a documented validation of the qualifications of the supplier to make the certifications.
Please provide a written statement or explanation to the U.S.
Nuclear Regulatory Commission, ATTN:
Document Control Desk, Washington, D.C.
20555 with a copy to the Chief, Vendor Inspec-tion Branch, Division of Reacter Inspection and Licensee Perform-ance, Office of Nuclear Reactor Regulation, within 30 days of tho' date of the letter transmitting this Notice of Nonconformance.
This reply should be clearly marked as a " Reply to a Notice of Nonconformance" and should include for1each nonconformance: (1) a-description of steps that have been or will be taken to correct these items; (2) a description of steps that have been or will be-taken to prevent. recurrence; and (3) the dates your corrective actions and preventive measures'were or will be completed.
l Dated _at Rockville, Maryland this @ Ni day of f { j (lQ^
1992. l L
1 ORGANIZATION:
ABB COMBUSTION ENGINEERING' NUCLEAR SERVICES WINDSOR, CONNECTICUT REPORT NO.:
99900401/92-01 CORRESPONDENCE Mr.~ Steven A.
Toollo, Manager ADDRESS:
Nuclear-Licer91ng ABB Combustion Engineering-Nuclear Services.
1000 Prospect Hill Road Windsor, Connecticut 06095-0500 ORGANIZATIONAL Gary S.'Bloomquist, Acting-Manager, Nuclear CONTACT:
Quality,- 203/285-5590 NUCLEAR INDUSTRY Support activities by a Nuclear Steam Supply-ACTIVITY:
System supplier for commercial nuclear power-plants.
INSPECTION September 28-October 2, 1992 CONDUCTED:
~
~
-Id!3O[SL TEAM LEADER:
+n Richard C.
Wilson, Senior Engineer
' Date Reactive Inspection Section 2 (RIS2)-
Vendor Inspection Branch (VIB)
OTHER INSPECTOR:
Stephen D.
Alexander,'RIS2, VIB APPROVED:
5
// 9 /-
Grego# f. Cwalina, Chief Date f
Reactive Inspection'Section 2-Vendor Inspection Branch INSPECTION BASES:
10 'FR Part 21 and 10 CFR Part 50, Appendix B l
INSPECTION SCOPE:
To review the dedication of commercial grade components for nuclear safety-related appli-cations by ABB Combustion Engineering Nuclear Services (CENS)
PLANT SITE Numerous APPLICABILITY:
'1 I I N S P E C T I O N S U M M A R Y ;
III ~Nonconforma(Lg_q
. Contrary :to Criteria-III and VII: ofi Appondix, B toe 107 CFi Partl50 -
and Systems 3 and17.of ABB-CENS's manual.~QAM-100, Edition'3~,
Revision 2, CONS failed toLadequately specify-and-implement' design <and-procurement requirements:and-control'the evaluatioi ot objective evidence of-quality forJpacking assembliestsupplied for safety-related-use in a charging-pump for:a commercial nuclear power' plant (Nonconformance 99900401/92-01-01; see Section;3.4.1' of.this report).
~
2 STATUS OF PREVIOUS = INSPECTION FINDINGS'
-There were no open findings from previous NRC inspections.of_this' facility.
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Entrance and Exit Meetinas In the entrance meeting on September 28, 1992, the NRC) inspectors-discussed the scope of the: inspection, outlined areas to be.
inspected,_and established interfaces with CENS' management-and staff.
In the exit meeting on October 2,
'992, the inspectors--
discussed their findings and concerns with-CENS management and-staff.
3.2 Insocetion Scope ABB Combustion Engineering Nuclear Services (CENS)lprovides Combustion Engineering's support (other than fuel)-for operating commercial nuclear power plants.
Three areas of_CENS: provide replacement parts that may involve _the dedication of' commercial--
grade items for nuclear safety-related_ applications:. -Nuc Aear.
Spare Parts, Instrumentation and Controls (I&C) Engineering and Mechanical Engineering groups.
Nuclear Spare Parts' supplies:
replacement parts (like-for-like or evaluated equivalent),Jwhile the other two groups supply systems or components' involving design activity.
Quality Assurance Manual-QAM-100,-Third' Edition, defines QA policy for CENS.
The Nuclear Spare _ Parts and I&C. groups supplemented QAM-100 with their own QALprogram descripti'ons;(QAM-300 and -400 respectively), and Mechanical Engineering ldid not.:
Each group had its own quality assurancu-procedures'andcoperating The NRC. inspectors r' viewed-selected: procedures and procedures.
e dedication' examples _for-all three groups. -The NRC: inspectors 2.,
~
T k
E
'also' reviewed CENS's program forl conforming toJthe; requirements
'of 10ECFR Part 21.
~
The Spare' Parts and Mechanical: groups had dedicatedicommerciall
~
grade' items!for-safety-related applications,'and the HRC inspectors reviewed several dedication efforts byLthosesgroups'l
- In contrast,'the-I&C group recently had been involved:>only11n.
dedication.of newly developedLsystems.-_HInLthis case,Ethe-inspectors reviewed the qualification of_atmanufacturing71ot of-Potter & Brumfield relays.
3.3 Nuclear Soare Parts 3.3.1 Program 1 Comments h
The NRC inspectors found that the CENS dedication'activitiesidid l
not fully define or consistently' document:(1) the' safety functions of spare and replacement parts or.thair' failure 1 modes-adverse to safety, (2) the rationale for. deriving theicritical;.
characteristics for performing those safety functions under-all; plant application conditions, and-~(3) thefspecific verification methods and acceptance criteria needed to-demonstrate 1thetcrit-ical characteristics.
Thus, it was often difficult'to determine what verification processes and documents were actually being:
relied upon as the basis for' demonstrating suitability for the safety-related application. _ Specific. examples of these weaknesses are provided on the following sections.
The inspectors also_noted that CENS attempted to procure safety-grade components where feasible, particularly where complexity might be a factor (however,. Mechanical ~ Engineering-personnel;_
stated that occasionally time and' availability might cause them to instead-procure and: dedicate commercial-grade materials or items).. Conservatively, CENS had not_yet performed any-dedications based solely on commercial grado vendor. surveys.
Most of the Spare Parts procurements were safety-grade,.and most' of the group's dedication efforts were parts-or relatively_ simple ~
components or_ services.
The' Replacement: Item Evaluations:
. documented differences in current and original catalog numbers.
~
Spare Parts procedures did not clearly require documenting an adequate' basis for dedication,-and-the= files'as: presented to the inspectors did not always contain.the basis,.but additional.
documents and explanation generally completed a; satisfactory.
basis.
The NPC inspectors had a concern regarding'theLCENS Nuclear Spare Parts Quality Assurance Manual QAM-300:
System 3, Paragraph 14.2,
" Replacement Item Equivalency. Evaluations," did notDaddress how,';
in reviewing spare parts assumed-to be identical to' original items, CENS' established that changes _to design, materials,-.or 3
l y
W g
y rm a-y-7r
- T??
-+v w-'Drr
manufacturing processes were in' fact reflected by part number; changes.. CENS personnel explained that their practice (although.
not written procedure) was to compare current manufacturer's:
published product information with the original lpart'and/ortits.
catalog description.
The inspectors noted that this check, although fundamental and necessary, would-not be sufficient to
-identify changes in design, materials, or processes thaticould' impact performance under certain design basis conditions:that.are-not reflected in outward form, fit, or functional descriptions; for normal operating conditions.
Review of selected procurement:
and dedication packages as described below confirmed-the=
inspectors' concern with this practice.
The inrpectors also identified a concern regarding-QAM-300 with respect to the role of commercial grade vendor certifications in the dedication process.
However, Spare Parts personnel: stated-that commercial grade vendor certifications-had never been' relied upon in the dedication process, and Spare Parts has not performed any dedications be. sed on surveying commercial.grrde vendors.
The certifications are used for other purposes.
3.3.2 0-Ring Dedication Review' The NRC inspectors reviewed an 0-Ring dedication by CENS Spare Parts that did not completely document the dedication activities.
A licensee purchase order (PO) to CENS ordered 0-Rings made of Parker Compound E515-80 for Target Rock Corporation solenoid' operated va)ves.
A recent CENS commercial grado PO to Target Rock specified Parker Compound E515-80 0-Rings..
The CENS PO specified that the O-Rings were to be~ dedicated =in accordance with Target Rock Dedication Procedure 5041.
CENS' personnel stated that they had reviewed that procedure,_but'could not document the review.
No source surveillance / verification or independent tests and audits were performed.
Ia response to the NRC inspectors' questions, CENS produced a-telephone memorandum stating-its customer's agreement that the 0-Rings could be dedicated based primarily.on part-number'and..
material $dentification markings.
The inspectors confirmed that-the CENS JAceipt inspection verified proper markings according'to a specification referenced by Parker in another telephone memorandum, as well as part numbers, cure dates, and a certification from Parker.
The inspectors were-concerned that the package did not show a logical path of verifying critical characteristics derived from safety. functions, and some reliance was placed on unvalidated information furnished by the commercial grade suppliers.
However, CENS staff provided additional documentation, not contained or referenced in the file, to satisfy the inspectors.
4 l,
l
3.3.3 Gear Reducer Dedication Review
- The NRC inspectors also reviewed a Spare Parts dedication file for a gear reducer rebuilding effort that;did not fully document the dedication activities.
Specifically, the file did not indicate that CENS had performed a source audit or surveillance of the commercial grade supplier.
Further, the-9.044:1 gear-ratio of the rebuilt unit war verified only by visually. observing; that manually rotating the high speed shaft nine turns caused the low epeed shaft to rotate approximately one full turn.
Although not included or referenced in the file,'CENS performed and documented surveillance of manufacture and testing of new-units.
The functional tecta of the new units showed that they:
had an acceptable ratio, and the rebuilding did not change parts
~
that would affect the gear ratio.
CENS also-did not document-that the same process was used for new and repair work, but the inspectors concluded that the CENS surveillance report provided sufficient basis for accepting the rebuilt gear reducer.
3.3.4 Electrical Connector Dedication Review The NRC inspectors also reviewed an electrical connector dedication by CENS Spare Parts that did not fully address possible parts changes not reflected _by changed catalog numbers.
CENS checked the outer thre:.de with a thread gage,.and checked-materials with a magnet, in addition to visual and dimensional checks.
Part numbers were verified against documentation for the plant system.
The connectors were a MIL specification type.
CENS showed that the parts received were the parts ordered, but did not convincingly document that the connectors were the same as the original parts.
Because of the simple nature of the components, after discussion with CENS personnel the NRC inspectors considered the dedication adequate.
3.3.5 Flow Indicator Dedication Review l
The NRC inspectors also reviewed a CENS Spare Parts dedication'of
)
panel-mounted flow indicators that did not fully address possible parts changes not reflected by changed catalog numbers.
- However, CENS's activities also included partially disassembling the-dedicated indicators to compare internal part and serial numbers with the original technicul manual.
CENS also exercised engineering judgement based on reviewing a qualification test report for similar meters previously supplied to another customer.
CENS's diligence in establishing similarity was also illustrated by identifying that limit switches had been added to the vendor *s current design, and evaluating the effect of that change.
The NRC inspectors concluded that the dedication was adequate.
5
_9
7-
.-._ _.~.- - -. -.
c 4
=
/
~
The customer!s PO for'thofflow(indicators specifiedLa modelj__. -
2 numberidesignating-_a' center-zeroidisplayiscale:forTanDindication
- ranging only' from zero: upward. -Thecvendor'stresponsomto'CENS's:
request <fortquotation:-caught 1this discrepancy.- TCENS_then generated threeiseparate documents correctingjtho' file?
' information,1butione pagejof1the: Dedicat' ion' Work OrderEstillJwas"
._ uncorrected.
~
3 3.6-Power: Supply Dedication Review The NRC.incpectors also reviewed a_ power 7 supply dedication by a
CENS Spare Parts;that did not fully address possible-partsf changes not reflected by changed catalogLnumbers.- ;Although the' functional testing performed.on_the power supplies was' adequate:
for normal conditions, the units were not' weighed or.aeismically?
y analyzed, and only a. catalog specification _ comparison-(currentDtoi j
original)=vas made to determine: continued suitability.forithe-nuclear application._ _ The inspectors considered the-lack f of
. attention to detail.regaiding the seismic qualification as'a
- weakness $n the dedication program.
3.3.7 Other Dedication Reviews The NRC inspectors reviewed two other Spare Parts dedicationi 3
files involving limited work applied'to safety-grade components by commercial grade vendors.
In one case, power supplies that-q had been rendered safety-grade-by a: system vendor-experienced:
blown transistors and were repaired under warranty'by the-commercial grade manufacturer.
CENS verified >-serial numbers and performed functional-tests of the repaired" units.< Inithe'7ther case, a safety-grade. stainless stecl retaining ringEfor a -
vibration pulse dampener passed through a-commercial'gradeivendor:-
who molded a rubber extrusion to it, with'CENS QA surveillance.
The NRC inspectors agreed that the dedication ~ activities'were appropriate for these~ limited-scope activities.
l 3.4 Mechanical Encineerina 3.4.1 Pump Packing Assembly Review The NRC inspectors reviewed a Mechanical' Engineering dedication, j
file that demonstrated inadequate: control-of design?and,
^
procurement activities.
Entergy-operations,JInc. PO:No._ WPO46223:
ordered plunger packingLassemblies for high pressure: charging:
pumps for the Waterford-nuclear power-plant-coolantivolume.
control; system.
Entergy required certification of conformance to:
the PO requirements; certification thatfall the_ parts were fabricated without using lead, tin, and'certain other elements;:
and documentation of shelf life, shelf lifeLexpiration, cure / manufacturing date, batch number, and storage informationfas' appropriate.
The Entergy PO also invoked nuclear QA requirements-and-the reporting provisions of110 CFR Part 21~.
6 N
,Ce vi w e y
m er ay--
DV
- - - - '-g ime-ey.y yg "7T1 T
i CENS1 classified the assemblies _as quality classf1D--commercial
')
l grade-items: intended:for a safety-related application--and procured themLin the corresponding: procurement Class 2.
CENS PO th). 9206246 dated June-19, 1992,"to UTEX Industries,_ Inc.,
Houston,cTexas, invoked-all-the technical and documentation requirements,'but without invoking nuclear QA requirements or 10 CFR Part 21 criteria. _The technical description of the pack-3
-ing assemblies specified_UTEX drawing-no..B-2870-C,1 Revision C.
In planning _the dedication of the_ commercial grade packing p
assemblies, CENS's Commercial Grade-Item Evaluation Form dated' August 4',
1992, documented that the'criticalfcharacteristics were-
" physical characteristics" to.be addressed by materials -
documentation review, dimensions, and " visual."
As acceptance _
methods, the form required-verifying that_ dimensions were within tolerance.. Dedication Work Order (DWO) C-Mech-ER-004,.
Revision 00, stated that the dedication procesc for_these_ items would consist of visual inspection, dimensional checks, and documentation reviews.
CENS did not document the specific safety function (s) of the packing-assemblies in support of the safety function of the pump,_or the failure modes adverse to safety;,
therefore, it was not clear t. hat all of the applicable critical characteristics had been_ derived and adequately identified.
For example, the free lengths of the two stainless steel compression springs in each packing assembly were required to be verified, but not the spring rates (nor were they).
The compression of the springs is not adjustable once they are installed, so the rate would be expected to be-as critical as_the length.
Also',-the material exclusions explicitly stated in the Pos were not listed) as critical.
In the actual dedJeation performed in accordance with the DWO cited above, CENO verified dimensions and performed visual-4 inspections, but relied on the documentation provided by_UTEX to verify that other requirements were met.
In addition _to the dedication planning concerns cited _above, the 4.
NRC inspectors identified the following concerns with the veritication process:
(1)
CENS had not audited or surveyed.UTEX to. qualify them to certify-meeting PO requirements or to issue certificates of chemical analyses, or that CENS had validated the certifications.
l (2) The UTEX certifications were signed by a person titled "QC Tech."
This level of nonsupervisory signature had not been verified to be designated in the UTEX quality assurance program,-
~
as far as the documentation reviewed. indicated.
(3) The apparently computer-printed general l certification simply restated.the PO requirements for each individual part (listed by-
- i part number _and description), addressing the requested '
7 l-I l
l information, but without regard for the applicability of the information to the part.
UTEX drawing B-2870-C, Revision C, showed that each ccmplete packing assembly (UTEX part no. 8500-25-147-00) consisted of primary and secondary polyethylene spacers, primary and secondary stainless steel compression springs, primary and secondary packings made of braided aramid
("Nomex"), and a bronze step bushing.
However, the certification assigned each part (regardless of its material) a batch number, a cure date of July 1992, a 10-year shelf life, a resultant shelf life expiration of July 2002, and the requirement for storage between 40 and 95
'P, not exposed to ultraviolet light or ozone, and wrapped in plastic.
In addition, the primary Homex packing and the primary polyethylene packing both had the same batch number (79224).
Thus, much of the certificate appeared to be meaningless, and it was apparent to the inspectors that CENS had not properly reviewed it.
(4) Finally, the NRC inspectors noted that the UTEX certification of the absence of certain elements (including tin and lead) was erroneous because the step bushing (UTEX part no. 1435-40-0222-
- 02) as shown on UTEX drawing B-2870-C, Revision C, was made of SAE 660 bronze (ASTM 13584 alloy, UNS spec. C93200).
Not only does bronze consist of copper and tin, but according to the CENS engineer, this particular alloy also contains lead.
In response, CENS ommitted to promptly inform affected customers of these concerns.
CENS also began an evaluation addressing the compatibility of the bronze with the pump materials, the wear and leachability of the tin and lead from the bronze, and the possible paths and transport mechanisms by which the excluded substances could enter the primary system.
CENS also stated it will take appropriate corrective action and/or advise affected licensacs accordingly.
From a programmatic standpoint, CENS personnel stated that they intended to determine why this deviation was not detected by their dedication process, and that their overall program improvemer.ts would prevent recurrence of this kind of deviation.
Finally, CENS committed to issue revised de,ication procedures for Mechanical Engineering by January 31, 19CJ (Nonconformance 99900401/92-01-01).
3.4.2 Shot Peening Dedication Review The NRC inspectors reviewed a Mechanical Engineering file for commercial grade shot peening services in which CENS relied on unvalidated documentation and certifications furnished by commercial grade suppliers.
CENS required vendor certification of meeting PO requirements, and subvendor chemical and physical test reports for materials used in the shot peening process.
Although none of the commercial grade suppliers had been audited by CENS, their personnel witnessed performance of the shot peening process and of an Almen surface intensity test, thus providing adequate verification of the process.
8 3.4.3 Other Dedication Reviews I
The NRC inspectors reviewed two other Me dedication files with only one concern. chanical Engineering related charging pump plunger.316 stainless steel was procured and ded
~
The CENS engineer stated that commercial grade material was specified because the plunger is not a pressure boundary comp onent and was thus not-subject. to the American Society of Mechanical Engineers (ASME). Boiler and Pressure Vessel Code. -The NRC inspectors pointed out that'the safety-related designation relates to nuclear safety as defined by the Code of Federal Regulations, and not simply to ASMF code.
CENS personnel agreed that in such cases safety-related material procurements would be employed where possible.
s did not evaluate the adequacy of the CENS dedication for this The inspectors item.
'The NRC inspectors also noted that an " Integrated Project Quality provide e useful control document for the overall effort. Plan" 3.5 IDptrumentation and Controls (I&C) Enaineering Group" approach that uses dedicated project teams for spe ng efforts.
The NRC inspectors reviewed QAM-400 and examined a c
MDR series relays for the Waterford Steam Electric Stationreq work included environmental qualification testing of sam The delivered relay.
Two relays failed during environmental qualification testing each case the specimen was returned to Potter & Brumfield and the In cause of failure was determined.
input commands at 137 One relay failed to respond-to
- F; shaft end play caused primarily by an oversized coil.the cause was found to be i relay responded sluggishly to input ccamands at 40 The other with rotor motion. uncured epoxy deposited on the stator assembly surface interfered
- F beca use with Potter & Brumfield to initiate manufacturing changesCENS I&C En correcting both problems,
{
appropriately reworked.
and the relays supplied.to CENS were concerns by CE Tech Note No.CENS notified its customers of these 52-05 dated September 4, 1993, which states that Potter & Brumfield indicated it planned to no may have been similarly affected.
I The I&C group directly supplied some relays to the customer Others, designated as spares, were supplied through the CINS Spare Parts group.
For the spares.
" Dedication Report" based primarily on bench testing,I&C Engineering prepared a and also on 9.
certification from-Potter & Brumfield that the items conformed to and that-substantiating CENS's specifications and drawings, The visits to Potter &
records were available for inspection.Brumfield by CENS personnel provide lot The NRC inspectors concluded that the the certification.
qualification and dedication activities. performed by:I&C-Engin relays.
3.6 10 CFR Part 21 Activities The NRC inspectors reviewed the CENS Administration Manual NS-1 procedure " Reporting of Defects and Noncompliance," January
- 1992, The current issue of Part 21 was requirements of 10 CFR Part 21.important terms were correctly defined, an approp
- invoked, and the instructions for notice was posted on bulletin boards, employee and management ectivities appeared to be adequate.
The NRC inspectors discussed CENS's recent Part 21 activities-In 1992 CENS had with the manager of nuclear licensing.
Two were found to identified three possibly reportable concerns.but the concerns were considered not require evaluation by CENS, and CENS did advise the to be of possible interest to the NRC, Evaluation of the third concern by the CENS Nuc but all NRC of them.
Safety Committee determined that it was not reportable,affected The NRC inspectors had no concerns in this area.
Group.
4 PERSONNEL CONTACTED ABB CENS:
A. Toelle, Manager, Nuclear Licensing Acting Manager, Nuclear Quality S.
+
G.
S. Bloomquist, G. F. Caruthers, Director, I&C Engineering
& Elec. Sys., I&C Engrg.
Ryan, Mgr., Prot.
+
M. P. Fritz, Supervisor, I&C Engineering
+
L. A. Bradshaw, Manager, Mechanical Engineering R. W. Mcdonald, Manager, Plant Structures M.
S. Burger, Manager, Reactor' Mechanical Systems J. M.
Pump Services Supervisor, C. Gimbrone, Supervisor, Nuclear Spare Parts Engrg.
+
T.
S. Bernard, Supervisor, Nuclear Spare Parts QC.
+
M. W.
Stewart,Engineer, Licensing
- Jarriel,
+
L. Buckholz, Engineer, Nuclear Spare Parts J.K. Neumann, Engineer, Nuclear Spare Parts M. Catwell, Engineer, Nuclear Spare Parts W. E. Mack, Senior Project Manager, I&C R. R. Senechal, Senior Engineer, I&C K. Tomany, Senior Engineer, I&C Quality Assurance
+
10 -
3.4.3 Other Dedication Reviews:
l The NRC inspectors reviewed two other Mechanical Engineering dedication fi13s with only one concern.
Commercial grade type 316 stainless steel was procured and dedicated for r safety-related charging pump plunger.
The CENS engineer stated that commercial grade material was specified because the plunger is not a pressure boundary component and was'thus not subject to the American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code.
The NRC inspectors pointed out that the safety-related designation relates to nuclear safety as defined by the Code of-Federal Regulations, and not simply to ASME code.
CENS personnel agreed that in such cases safety-related materials procurements would be employed where possible.
The inspectors did not evaluate the adequacy of the CENS dedication for this item.
The NRC inspectors also noted that an " Integrated Project Quality Plan" used in the charging pump plunger dedication appeared to provide a useful control document for the overall effort.
3.5 Instrumentation and Controls (I&C) Encineerina CENS I&C Engineering QA Manual QAM-400 defines a " System Staging Group" approach that uses dedicated project teams for specific efforts.
The NRC inspectors reviewed QAM-400 and examined a requalification/ rededication effort with Potter & Brumfield Co.
MDR series relays for the Waterford Steam Electric Station.
The work included environmental qualification testing of sample.
relays and extensive functional performance testing of every delivered relay.
Two relays failed during environmental qualification testing.
In each case the specimen was returned to Potter & Brunfield and the cause of failure was determined.
One relay failed to respond to input commands at 137 *F; the cause was found to be insufficient shaft end-play caused primarily by an oversized coil.
The other relay responded sluggishly to input commands at 40 'F because uncured epoxy deposited on the stator assembly surface interfered with rotor motion.
CENS I&C Engineering and QA personnel worked with Potter & Brumfield to initiate manufacturing changes correcting both problems, and the relays supplied to CENS were appropriately reworked.
CENS notified its custoters of these concerns by CE Tech Note No. 92-05 dated September 4, 1992, which states that Potter & Brumfield indicated it planned to notify other customers if there was evidence that other delivered relays may have been similarly affected.
The I&C group directly supplied some relays to the customer.
Others, designated as spares, were supplied through the CENS Spare Parts group.
For the spares, I&C Engineering prepared a
" Dedication Report" based primarily on bench testing, and also on 9 -__-..
~
cortificationEfroa Potter & Brumfield that ths. items conform d to-
-CENS's specifications ^and drawings,;and.that-substantiating-records were available'for inspection.: The visits to Potter?&-
'Brumfield by CENS-personnel provided adequate basis for-accepting' the. certification.-_ The NRCiinspectors concluded that:the lot qualification and dedication _ activities performed by'I&C; Engineering resulted in the shipment of adequate safety-grade-relays.
3.6 10 CFR Part 21 Activities The NRC inspectors reviewed the CENS Administration Manual NS-1 procedure " Reporting of_ Defects and Noncompliance," January 1, 1992, which provided instructions for-responding to the reporting requirements of 10 CFR Part 21.
The current. issue of Part 21 was=
invoked, important terms were correctly defined,-an1 appropriate notice was posted on bulletin boards, and the instructions for-employee and management activities appeared to be adequate..
The NRC inspectors discussed CENS's recent Part 21 activities with the manager of nuclear licensing.
In 1992 CENS had identified three'possibly reportable concerns.
Two were'found'to not require ecaluation by CENS,.but the concernsfwere. considered to be of possible interest to the NRC, and CENS did advise the NRC of them.
Evaluation of the third concern by the CENS Nuclear.
Safety Committee determined that it was not reportable, but all-affected CE operating plants were notified through the Owners' Group.
The NRC inspectors had no concerns in this area.
j 4
PERSONNEL CONTACTED ABB CENS:
S.
A. Toelle, Manager, Nuclear Licensing
+'
G.
S. Bloomquist, Acting Manager, Nuclear Quality G.
F.
Caruthers, Director, I&C Engineering
+
M. P.
Ryan, Mgr., Prot. & Elec. Sys., I&C Engrg.-
+
L. A. Fritz, Supervisor, I&C Engineering l
R. W. Bradshaw, Manager, Mechanical Engineering M.
S.
Mcdonald, Manager, Plant Structures J. M. Burger, Manager, Reactor Mechanical Systeme
?
C. Gimbrone, Supervisor, Pump' Services
+
T.
S. Bernard, Supervisor, Nuclear Spare Parts Engrg.
+
M. W.
Stewart, Supervisor, Nuclear Spare Parts QC-
+
L. Jarriel, Engineer, Licensing J. Buckholz, Engineer, Nuclear Spare Parts K. Neumann, Engineer, Nuclear Spare Parts H.
Catwell, Engineer, Nuclear Spare Parts W.
E. Mack, Senior Project Manager, I&C R. R.
Senechal, Senior Engineer, I&C
+
K. Tomany, Senior Engineer, I&C Quality Assurance 10 -
T. Marx,.Sonior Enginsor (mschenical)
Viray, Project Engincar (m chanical):
B
,D-.
C. D. - Blanchard, Senior. Engineer - (mechanical).
D. Byerly, Quality Control Rep. (mechanical) 1110:
G.
C.
Cvalina, Section Chief, RIS2/VIB/DIR/NRR
+
Attended the entrance meeting on September 28, 1992 Attended the exit meeting on October 2, 1992 m
\\
11
...s e l' s
<[%
+
UNITED $TATES NUCLE AR REGULATORY COMMISSION w AssmotoN,0. c. 20665
%, *..e
,o#
N0!! 3 UU Docket No. 99900519 Mr.
L.
R. Ruhland, Sr. Vice President Regional Office Manager Bechtel Corporation 9801 Washingtonian Divd.
Gaithersburg, Maryland 20877-5356
Dear Mr. Ruhland:
SUDJECT:
NOTICE OF NONCONFORMANCE (NRC INSPECTION REPORT No.
99900519/92-01)
This letter addresses the inspection of your facility at Gaithersburg, Maryland, conducted by Messrs. B. H. Rogers, J. J. Petrosino, and R. L. Pettis of this office on September 3 and October 1, 1992, and the discussions of their findings with members of your staff on October 14, 1992.
The purpose of this inspection was to evaluate your method of qualifying Five Star Products, Inc., as a supplier of safety-related products and to determine the scope of those products supplied to nuclear licensees.
Areas examined during the inspection and our findings are discussed in the enclosed inspection report.
This inspection consisted of an examination of procedures and representative records and interviews with personnel.
During the inspection it was found that the inplementation of your quality assurance program failed to meet certain NRC requirements.
Specifically, when the Bechtel Corporation surveyed Five Star Products, Inc., in April and June of 1992, it did not verify the existence of objective evidence of quality and did not determine if Five Star Products, Inc., had established an adequate quality assurance program-and was properly implementing such a program.
The specific findings and references to the pertinent requirements are identified in the enclosures of this letter.
Please previde u5 within 30 days from the date of this letter, a written statemene,tn accordance with the instructions specified in the enclosed Notice of Nonconformance (Notice).
We will consider extending the response time if you can show good cause for us to do so.
i Mr.
L.
R. Ruhland The response requested by this letter and the enclosed Notice are not subject to the clearances procedures of the Office of Management and Budget as required by the Paperwork Reduction Act of 1980, Public Law 96-511.
In accordance with 10 CPR 2.790 of the NRC's " Rules of Practice,"
a copy of this letter and its enclosures will be placed in the
!!RC Public Document Room.
If you have any questions about this inspec' ion, we will be pleased to discuss them with you.
Sincerely,
/
-L tead Leif rrholm, Chief Vendor Inspection Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation
Enclosures:
1.
Notice of Nonconformance 2.
Inspection Report 99900519/92-01 l
ENCLOSURE 1 NOTICE OF NONCONFORMANCE Bechtel Corporation Docket No. 99900519/92-01 Gaithersburg, Maryland Based on the results of an NRC inspection conducted on September 3 and October 1, 1992, it appears that certain of your activities were not conducted in accordance with NRC requirements.
Criterion VII, " Control of Purchased Material, Parts, and Components," of Appendix B to 10 CFR Part 50 requires in part that measures shall be established to assure that purchased i
material, equipment, and services, whether purchased directly or.
through contractors or subcon' tractors, conform to the procurement documents.
These measures shall include provisions, as 1
appropriate, forcsource evaluation and-selection, objective evidence of quality furnished by_the contractor or subcontractor, inspection at the contractor or subcontractor, and examination of the products upon delivery.
Paragraphs 4.3.2.2 and 4.3.2.3 of Section 4, " Audit Implementation," of ANSI N45.2.12-1977, conditionally endorsed by-HRC Regulatory Guide 1.144, September 1980, " Auditing of Quality Assurance Programs for Nuclear Power Plants," as committed to by Bechtel Corporation in the Bechtel Topical Report, Revision-4A, February 1988, require that objective evidence shallfbe examined for compliance with quality assurance program requirements and that selected elements of the quality assurance program shall be audited to the depth.necessary to determine whether or not they.
are being implemented' effectively.
Paragraph 2.1.4.1 of Section 2.1 of the Dechtel Corporation Procurement Supplier Quality Manual, eighth edition, August 10, 1990, requires that the Bechtel auditor examine the implementation of the quality assurance program in effect at the time for compliance with supplier procedures, while performing an.
in-depth review as part of a supplier survey.
Contrary to the above, in surveying Five Star Products, Inc. in April and June of 1992 the Bechtel Corporation did not verify the existence of objective evidence of quality and did not determine if Five' Star Products, Inc., had established an adequate quality-assurance progra'a and was properly implementing the program.
(99900519/92-01-01) -
i
~
-. ~.-.-
.=.. --
Please provide a written statement or explanation to the U.S.
Nuclear Regulatory Commission, ATTH:
Document Control Desk, Washington, D.C.
20555 with a copy to the Chief, Vendor Inspection Branch, Division of Reactor inspection and Licensee Performance, Office of Nuclear Reactor Regulation, within 30 days of the date of the letter transmitting this Notice of Nonconformance.
This reply should be clearly marked as a " Reply to a Notice of Nonconformance" and should include for each nonconformances (1) a description of steps that have been or will be taken to correct these items; (2) a description of steps.
that have been or will be taken to prevent recurrence; and (3) the dates your corrective actions and preventive measures were or will be completed.
a Dated at Rockville, Maryland this J1% day of.tlogemb ec, 1992.
. ~ _. - - - - -. _.
ORGANIZATION:
Bachtel Corporation REPORT NO.
99900519/92-01 CORRESPONDENCE Mr. L. R. Ruhland, Sr. Vice President
- ADDRESS '
Regional-Offico Managor-Bochtel Corporation 9801 Washingtonian Blvd.
Gaithersburg, MD 20877-5356 ORGANIZATIONAL Mr. Harry Hollinghaus, Manager of CONTACTS:
Quality Assuranco NUCLEAR INDUSTRY-Provides safety-related products ACTIVITY:
and services for commercial nucioar power plants.
INSPECTION September 3 and October 1, 1992 CONDUCTED:
3311, Kl(eru.
a,tsi.ru Billy II. Roged[, Lead Inspector Dato g
Reactive Inspection Section No. 2 (RIS.2) i Vendor Inspection Branch (VIB)
OTHER INSPECTORS:
Joseph J. Potrosino, VID Robert L. Pattis, VIB
/>
/
APPROVED:
N'oce
// 1 /s
/
Gregory'fCg' Cwalina, Chief,
Date RIS 2, VID INSPECTION BASES:
10 CFR Part 21 and Appendix B.to 10 CFR Part 50 INSPECTION SCOPE:
To evaluato selected portions of-tho implementation of Bechtel Corporation's quality assuranco program, surveys, audits, annual evaluations, and inspections as related to purchases from Fivo Star Products, Inc.
PLANT SITE Numerous APPLICABILITY:
l l L
n
~.
. =..
...... _ -. ~. _. - - - - -.. - -. - - -
l i
i 4
l'8 Adici V MMARY
/.', 'Df orma n ce s
, 2 i
i~;r.. ; 6 y to' Critorion VII, " Control of Purchased Haterial, Parts, e
and Components," of Appendix B to.10 CFR Part 50 (Appendix B) the l
Dechtel Corporation (Bochtel), in conducting a survey of Five Ste
-ts, Inc. (Five Star) in April and June of 1992 did
~
- he existence of objective evidence of' quality and-did not i
not de
.une if Five Star had established and-was properly l
implemu..cing an adequate quality assurance (QA) program, t
(Nonconformance 99900519/92-01-01) l 2
STATUS OF PREVIOUS INSPECTION FINDINGS The inspection of Bechtel was of limited scope and the NRC inspectors did not follow up on any previous. inspection findings.
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Entrance and Exit Meetingg During the entrance meeting on September 3, 1992, the NRC inspectors discussed the scope of the inspection and the areas.of-concern and established the persons to contact within Bechtel's management and staff.
During the exit meeting on October 14, 1992, the NRC inspectors discussed their findings and concerns with Bechtel management and staff.
3.2 Inspectign and ScoD2 3.2.1 DAckground On August 18 and 19, 1992, NRC inspectors attempted to perform-_an I
inspection of Five' Star at Fairfleid, Connecticut, pursuant to the provisions of Section 21.41, " Inspection,"-of Title'10 of the i
C2d2 21 Federal Reculations (10 CFR 21. 41).
The NRC inspectors were denied the access needed to inspect Construction' Products Research (CPR) test laboratory and test records on1the Five Star premises and were told on August 19, 1992, that they would not-be allowed to continue the inspection after the end of:the day even though the NRC inspectors indicated that they had not completed the inspection.
Upon being denied access, the NRC could not examine the extent and adequacy of products produced by Five Star and, therefore, did not have reasonable assurance that tho' products produced were adequate for use in safety-related application in nuclear power plants. _.The staff documented this information-in Information Notice 92-66, " Access Denied to NRC.
.?
W W
l :
w--..
Inspectors at Five Star Products,oInc. and Construction Products Research, Fairfield, Connecticut."- The NRC inspectors noted that Bechtel had performed activities in-April and June of 1992 to evaluate and approve Five Star as a supplier of safety-related products.
Since the limited NRC inspection of Five Star had identified areas of concern, Bechtel was-inspected to examine-their basis for approval of Five Star as a supplier of safety-related products.
3.2.2 Review of the TVA Purchase Order; Bechtel's Procedures.
QA Tonical, and commitments on April 23, 1992, ITA placed a blanket purchase order.to Five Star, with Bechtel acting as agent, for the purchase of grout.
The PO stated that the grout was to be QA. Level 1'(the TVA quality level for items purchased as Appendix B safety-related),.
manufactured and supplied under the QA program evaluated and-accepted by TVA/Bechtel to be in accordance with American Nuclear-Society Institute (ANSI) Standard N45.2, 1971, " Quality-Assurance Program Requirements for Nuclear Facilities," and that 10 CFR Part 21 apolied.
Five Star was-not on either'Bechtel's Evaluated Supplier List or TVA's approved supplier list for '8 owns Ferry, as a qualified supplier of safety-related grout, when TVA placed the blanket purchase order.
Therefore Bechtel: assigned an auditor to perform a supplier survey, as documented on the Assignment Transmittal (Bechtel Form PSQ-211) of April 20, 1992.
The NRC inspectors determined that the Assignment Transmittal instructed the Bechtel auditor to perform a supplier survey in accordance with Section 2.1 of-the Bechtel Procurement. Supplier Quality Manual (PSQM) and a quality program evaluation-in-accordance with the PSQM Section 2.2.
-The Special Instructions section indicated that the assignment was for an in-depth _ survey-including reviewing the QA manual.
Paragraph 2.1.4.1 of Section 2.1 of the PSQM definedLan in-depth review as including the following steps:
" Review the quality- _
program and complete the appropriate survey forms to. include the 1.
PSQ Program Evaluation Checklists.
Examine the' implementation-of-the program in effect at the time _for compliance with supplier-procedures."
In Bechtel Topical Report, revloion 4A, dated February 1988, Bechtel documented its commitment to. Regulatory. Guide (RG).1.144, Revision-1, September 1980, " Auditing of Quality: Assurance Programs for Nuclear Power Plants," in which the NpC
. conditionally endorsed ANSI N45.2.12-1977, " Requirements for Auditing of Quality Assurance-Programs-for NuclearJPower Plants,"-
and RG 1.123, Revision 1, July 1977, " Quality Assurance Requirements-for Control of Procurement offItems and Services for_,
---.+..,.-.4-w...-
--,---.-v.--e
Nuclear Power Plants," in which the NRC conditionally endorsed ANSI N45.2.13-1976, " Quality Assuranco Requirements for Control of Procuromont of Items and Services for Nuclear Power Plants."
In RG 1.144, paragraph C.b.(2) the NRC stated that the purchaser should audit olements of a supplier's quality assuranco program on a trionnial basis with the audit-performed in accordanco with ANa. N45.2.12-1977 and that the trionnial period should begin with performance of an audit when the supplier has begun sufficient work to demonstrate that the organization is implementing a QA program having the required scopo for purchases during the trionnial period.
In RG 1.144, the NRC also stated that a survey conducted before the award may sorvo as the first triennial audit if the supplier is implomonting the samo QA program for other customers that he proposes to uso on the auditing party's contract and if this first survey is conducted 1
in accordance with Section 4 of ANSI N45.2.12-1977.
Paragraph 4.3.2.2 of Section 4,
" Audit Implomontation," of ANSI N45.2.12-1977 states that objectivo evidence shall be examined for compliance with quality assuranco program requirements, and paragraph 4.3.2.3 states that selected olomonts of the quality assurance program shall be audited to the depth necessary to determine whether or not they are being implomonted offectively.
Bechtel personnel stated that the activity-performed in April and June of 1992, resulting in Fivo Star being placed on Bochtol's Evaluated Supplier List, was a survey performed in accordance with section 2.1 of the Bechtel Supplier Quality Manual, and was.
not considered an audit.
Notwithstanding the fact that Bochtel termed the review a survey vice an audit, the NRC inspectors determined that Bochtel was required to moet the requirements of Section C.b.(2) in RG-1.144.
The inspectors also datormined that the survey was the only activity performed by Bochtel to qualify Fivo Star as an Appendix B supplier beforo Bochtel issued the purchase order to release the grout to TVA for uso in the Browns Ferry Nuclear Plant._
Therefore the inspectors concluded.that the 3
survey was required to substantiate that the QA-program was adequate and effectively implementing the control of safety-related materials.
The NRC inspectors reviewed the above documents and Bechtol's commitments and determined that the Bochtel auditor was required to verify that Five Star adequately implomonted its QA program to onsure it had adequate controls in place while producing safety-related products...
~
u 2
3.2.3 Review of Bechtel's Anril/ June 1992-Evaluation of Five Star and Interview with the Bechtel Auditor The NRC inspectors interviewed the Bechtel Senior Supply Quality--
Representative-(the auditor) who performed the surveys-in April and Juno of 1992 which established Five Star as qualified to provide safety-related products to nuclear. licensees in accordance with the requirsments of Appendix B and 10 CFR Part 21.
Bechtel Projects had supplied a July 23,--1991 copy of the Five-Star QA manual to the Bechtel auditor before,the survey.
On April 23 and 24, 1992, the Bechtel auditor performed.an-initial survoy at the Shamokin Filler Company, Inc.~,
(Shamokin) in Shamokin, Pennsylvania, and at the Five Star facility in.
Fairfield, Connecticut.
The Bechtel auditor determined that Shamokin had calibrated its scales, established a nuclear hold area, coded the concreto and grout bags with thc=dato for daily I
lots, and used the same mixtures for both nuclear and non-nuclear-naterial.
Since material was actually shipped from Shamokin and l
not from Five Star, the material was reportedly maintained in the nuclear hold area until the lot had been successfully-tested, which typically took 28 days.
The Bochtel auditor datormined that Shamokin was not a part of Fivo Star even though Five Star stated that the Five Star QA manual applied to Shamokin.. The Bechtel auditor also determined that Shamokin did not'have a copy of the Five Star QA manual on site.
The Dechtol: auditor did not observe sampics being taken and did not know how they;were
-i shipped.
Based on discussions with the Bechtel auditor and Five Star employees, the NRC inspectors preliminarily concluded that-Shamokin was a commercial grado filler plant and did not maintain verifiable or significant controls on the " nuclear hold area."-
The Bechtel auditor discussed the Five Star QA manual with the l
Five Star QA manager and indicated that it was unsatisfactory, partly because Five Star stated in the QA manual that Shamokin was a part of Five Star while Shamokin was actually a sub-.
- P contractor manufacturing a commercial grade product for Five Star.
The Five Star QA managet stated to'the Bochtel auditor.that.Fivo-Star had denied the Nuclear Utilities. Procurement Issues Committoo (NUPIC) access to the " internal"~tostilaboratory at the Fairfield, Connecticut location.
The Bechtel auditor asked whether safety-related teeting was performed in the " internal" -
t laboratory and the Five Star QA manager responded that'it was'not and that safety-related testing was contracted to CPR, an Appendix B laboratory on the supplier list approved by'Five Star.-
The Fivo Star QA manager did not~ inform the Bechtel auditor that u -
2 w
1--.
u...
CPR vng, in fact, the " internal" lab and a subsidiary of Five Star.
While conducting the April survey, the Bechtel auditor determined that the Five Star QA manual was misleading about manufacturing being performed by subsuppliers; organizational charts did not represent the company structurc; all suppliers had not been properly evaluated; and the QA program failed to adequately address test control, calibration, nonconformances, corrective actions, QA records, and internal audits.
Therefore, the Bechtel auditor found Five Star to be unacceptable as a supplier of safety-related products.
The Bechtel auditor did not complete a Bechtel checklist due to the numerous deficiencies of the Five Star QA manual.
The Bechtel auditor informed the Five Star QA manager of the deficiencies only in conversation and no written record was generated.
On June 2, 1992, the Bechtel auditor returned to Five Star at Fairfield, Connecticut, to review the new revision of the-Five Star QA manual and perform an in-depth survey.
The Bechtel auditor determined the revised Five Star QA manual to be acceptable, verified the reporting lines of authority, reviewed the suppliers list approved by Five Star, and reviewed purchase orders for laboratory work from Five Star to CPR.
The Bechtel auditor also reviewed the Five Star audit of CPR.
The Bechtel auditor determined that the Five Star audit of CPR was acceptable but indicated to the Five Star QA manager that a new audit of CPR was required because the original audit form contained questions requiring only an " acceptable" or "not acceptable" response.
The Bechtel auditor determined that the April survey, the June survey, and the June review of the revised Five Star QA manual constituted an acceptable basis for approving rive Star as a supplier of safety-related products in accordance-with Appendix B.
Thus, Bechtel subsequently placed Five Star on the Evaluated Supplier Lict.
The NRC inspectors reviewed the Bechtel survey and determined that it did not clearly document the work performed by the Bechtel auditor.
The Audit Checklist divided each of the Appendix B criteria into very detailed sections and included guidelines for verifying that the applicable criterion was implemented.
Each of the 74 sections were indicated as " sat" (satisfactory) or "H/A" (not applicabic), with 68 indicated as
" sat" and 6 indicated as "N/A."
However, in the summary which followed each criterion, Dechtel did not sufficiently describe the reasons for the ratings chosen.
The audit checklist, on which most of the sections had been marked " sat" with little
~
clarification in the summaries, implied that most of-these guidelines had been followed.
The NRC determined during a discussion with the Bechtel auditor that most of the implementation guidelines for each section had not been followed.
The Bechtel survey did not identify any weakness in the extent of knowledge the Five Star QA' manager maintained in the area-of nuclear quality assurance, although the NRC inspectors determined through a review of Five Star documents and discussions with-the Five Star QA manager that he had little knowledge in the area of nuclear quality assurance.
The Five Star QA manager stated to the NRC that he had a minimal knowledge of Appendix B, 10 CFR Part 21, and other NRC requirements and in addition stated that his job was primarily contract negotiation with filler plants and not quality assurance.
The NRC inspectors determined that the Bechtel survey was insufficient to establish Five Star as a qualified supplier of Appendix B safety-related products in that the survey did not verify the existence of objective evidence of quality and did not determine the adequacy and implementation of the Five Star QA program as indicated in the following examples:
(1)
Bechtel's survey was not sufficient to-determine that Five Star adequately verified control-of testing of safety-related items.
The Bechtel auditor used a review of'the-Five Star audit of CPR as a basis on which to determine if I
Five Star's QA program was acceptable to control purchased materials, equipment, and services in accordanco with Criterion VII of Appendix B and.to control-testing in.
accordance with criterion XI of Appendix B.
The NRC has reviewed the Five Star audit of CPR and found.that the audit form is a checklist of elements marked as acceptable with a.
single acceptability evaluation consisting of one. sentence and does not contain any documented objective evidence of test control.
The NRC concluded-that-the documentation of the Five Star audit of CPR was insufficient to verify that CPR had established and was adequately-implementing a QA program in accordance with Appendix B and therefore Bechtel's review of the Five Star audit of CPR was-an insufficient basis in the establishment of Five Star's qualification as an Appendix B supplier.
Five Star's method of test control was particularly.
important because Five Star's process for producing concrete and grout products consisted of three major components:
(1) the administrative procedures for controlling the product and verifying quality by maintaining-lot homogeneity, batch control, and traceability to test samples;.(2) the-manufacturing process at commercial grado w u
= - -
-a
filler plants contracted by Five Star, currently Shemokin; and (3) testing performed by CPR, a subsidiary of Five Star, and one of only two companies on:Five Star's: approved i
supplier list.
In addition, the Five Star QA manual clearly indicated that the only difference between commercial grade and safety-rclated concrete and grout was the testing to which each is subjected.
(2)
The Bechtel survey was not sufficient to determine that the QA program at Five Star did not adequately address identification and control of materials.
Both the Five Star QA manual und the Dochtel audit report include statements
.i that the safety-related Fivo Star product is difierentiated i
from commercial grade product by testing, which results in.
Five Star performing what constitutes a dedication.
The Five Star QA program does not include sufficient controls to verify quality by ensuring that the commercial grade
-manufacturer maintains lot homogeneity, batch control, and-traceability to test samples.
The NRC inspectors identified Bechtel's failure to verify the existence of objective evidence of quality and determine the adequacy and implementation of the Five Star QA program, as described above, as a nonconformance (99900519/92-01-01) to Criterion VII of Appendix D.
4 PERSONNEL CONTACTED L.
Ruhland, Senior Vice President, Regional Office Manager
+
- h. Hollinghaus, Manager of Quality Assurance D. Kansal, Manager of Quality Assurance
- +
J.
Salasky, supplier Quality Manager
- +
T._Sarma, Project QA Manager
- +
F.
Dykstra, Chief Construction Quality Control Engineer T.
Richardson, Manager of Engineering T.
Parks, Manager of Field Procurement W.-Lower, Project-QA Manager
+
M. Zeiger, Procurement Operations Manager
+
D.
Quattrociocch, Engineering Manager N.
Isbel,. Senior Supply quality Representative Attended the entrance meeting on Saptember 3, 1992-
+
Attended the exit meeting on October 14, 1992 - _.
y[,* ~eg\\,
UNITED STATES e
.#a NUCL EAR REGULATORY COMMISSION f
wAsumotow,0. c. tous r
s OCT 2 319n Docket No. 99900300 Mr. Michael S.
Horinka, Manager Quality Assurance, Nuclear and Military Coltec Industries, Fairbanks Horse Engine Division 701 Lawton Avenue Beloit, Wisconsin 53511 Dear Mr. Horinka
SUBJECT:
INSPECTION OF T!!t. CLASSIFICATION AND COMMERCIAL GRADE DEDICATION PROGRAnS FOR DIESEL ENGINE PARTS AT COLTEC/FMED (REPORT No. 99900300/9?-01)
This letter refers to the-inspection of your facility at Beloit, Wisconsin conducted by Messrs. Walter Haass, Steven Matthews, Ronald Frahm, Jr., and Christopher Regan of this office on August 25-27, 1992, and the discussion of their findings with affected personnel in your company at the conclusion of the inspection.
The purpose of this inspection was to review the new Coltec Industries, Fairbanks Morse Engine Division (Coltec/FMED) program developed in response to the nonconformance identified in the Nuclear Regulatory Commission (NRC) inspection conducted on March 19-22, 1990.
The nonconformance addressed the need for irproved controls for the dedication of commercial grade apare and replacement parts for safety related use in diesel engines that provide emergency power at nuclear power plant facilities.
The inspection also entailed review of the implementation of the new program by evaluating procurement packages, and the review of licensee audits of the new program, In general, the results of the inspection indicate that the new program, as described in Standard Practice 750.00, provides an acceptable approach for the classification and dedication of-spare and replacement parts for the repair and maintenance of safety related diesel engines at nuclear power plants.
This assessment includes the effects of the progran changes arising-from the resolution of the findings identified by several licensees during their audits of the new progt'm, and the observations identified during this inspection.
However, the inspectors' review of procurement packages, to determine whether implementation of the new program was proper, revealed that in several areas there was a need to modify / upgrade the process.
These observations were not found to adverrely reflect on the formulation of the new program, but to rather result from the l
Hr. Michael S. Horinka "
carly and incomplete stages of its development.
The observations identified are discussed in the enclosed inspection report.
Please note that further discussion of HRC requirements for the dedication of commercial grado parts for safety-related service will take place at an upcoming workshop on procurement involving NRC and nuclear industry representatives planned for early 1993.
In accordance with 10 CFR 2.790 of the NRC's " Rules of Practice,"
a copy of this letter and its enclosure will be placed in the NRC's Public Document Room.
Should you have any questions concerning this inspection, we are available to discuss them with you.
~5 sincerely,
)
l
[-
V g3p m.__
Leif
. Norrholm, Chief Vendor Inspection Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation
Enclosure:
Inspection Report No. 99900300/92-01
. i
Enclosure ORGANIZATION:
Coltoc Industries, Fairbanks Morse Engine Division (Coltec/FMED) 701 Lawton Avenuo Boloit, Wisconsin 53511 REPORT NO.
99900300/92-01 CONTACT:
Michael S.
Horinka, Manager Quality Assurance, Nuclear and Military Telephone: 608-364-8314 NUCLEAR INDUSTRY Manufacturer of diesel engines used as an ACTIVITY:
emergency source of electrical power in nuclear power plants; also provides safety-related sparo and replacement parts, and maintenanco, repair and training services for these engines.
INSPECTIr*
August 25-27, 1992 CONDUCTED.
N'
/'
"i v. <~
INSPECTORS:
Walter P.
Naasa, Team Leador Date Reactive Inspection Section No. 2 Vendor Inspection Branch S. M. Matthews, QA Specialist Vendor Inspection Branch R.
K. Frahm, Jr., QA Engineer Vendor Inspection Branch C.
M. Regan, Intern Direct r's Offi e APPROVED BY:
< /rf#G1/
N
/
1 dregoryht//Cwalina, Chief Date Reactive inspection Section No. 2 Vendor Inspection Branch INSPECTION BASIS:
10 CFR Part 50, Appendix B to Part 50, and 10 CFR Part 21 INSPECTION SCOPE:
Announced inspection to review the vendor's new program, Standard Practice 750.00, and its implementation for the classification and dedication of commercial grade spare and replacement parts for use in safety related emergency diesel engines at nuclear power plants.
PLANT SITE APPLICABILITY:
All plants with Coltec/FMED emergency diesel engines..
1 i
1 INSPECTION
SUMMARY
l 1.1 YinlaLinna None identified during this inspection.
A 1.2 ILoliconfaraanqDE None identified during this inspection.
1.3 Unresolved Items None identified during this inspection.- One item remains unresolved from a previous inspection.
This11 tem will be reviewed during a future inspection.-
2 PREVIOUS INSPECTION' FINDINGS AND OTHER COMMENTS 2.1 S10tus of Previous Insoection P[nclingg Nonconformance:
A nonconformance (90-01-01) was identified-regarding the adequacy of the vendor's program for classifying and dedicating commercial grade spare and replacement parts-for safety related service in emergency 1 diesel engines.
In response.to this nonconformance, the vendor developed Standard Practice 750.00 which was assessed by the NRC inspectors during this inspection.
The inspectors concluded that the new program provides an acceptable approach for the classification and dedication of a
commercial grade spare and replacement parts for emergency diesel engines.
Several deficiencies were noted inLthe formulation and implementation of the new program; these were attributed to the developmental-nature of the program and the early stages of implementation.
Nonconformance 90-01-01 is considered closed.
Unresolved item:
An unresolved item-(90-01-02) was identified regarding the-need to update the QA manual to i
reflect tho' new Standard. Practice 750.00 issued on: March 13, 1990.
This updating was accomplished in Revision 25 of the QA manual dated August 15, 1990.
Unresolved itec 90-01-02 is considered closed.
Unresolved items An unresolved item (90-01-03). was identified regarding the need to adequately control the shelf life of engine parts subject to aging such as gaskets and seals.
The: vendor is currently considering a means for lot control, but has not decided on a final resolution.
This unresolved item remains open.,
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Entrance and Exit Meetinua An entrance meeting was held on August 25, 1992, in which the NRC inspectors addressed the purpose and scope of the planned inspection with the Quality Assurance Manager and other members of the Coltec/FMED staff.
An exit mooting was held on August 27, 1992, to summarize the areas reviewed and the conclusions drawn during the inspection.
3.2 Eackcround During the previous inspection of Coltec/FMED conducted on March 19-22, 1990, a nonconformance was identified regarding the vendor's process for supplying commercial grade spare and replacement parts for use by licenso2s in safety related emergency diesel engines at nuclear power plants (see NRC Inspection Report No. 99900300/90-01).
As a response to that nonconformance, the vendor developed a now program as described in Standard Practice 750.00 entitled " Nuclear Spare / Replacement Parts Program," and initiated its inplementation on March 13, 1990.
Copies of the new standard practice were distributed to licenseo customers by letter dated October 10, 1991.
Refinement of the program and full implomontation was expected by April 1991.
Audits / reviews of the now prograz by licensees indicated that of the 22 licensco customers, 21 have found the program acceptabic and one has determined it to be conditionally acceptable.
This inspection focused on the review of the edequacy of the now process, assessment of its implementation, and a review of licensees' audits of the new process.
3.3 Review of Standard Practice 750.00 Star.dard Practico 750.00 (SP-750) describes the vendor's program for the classification and dedication of commercial grade diesel engine parts that are intended for use in safety related service at nuclear power plants.
The program is subject to the provisions of 10 CFR Part 21 and the requirements of 10 CFR 50, Appendix B through the use of the vondor's Quality Assurance Mantal which satisfies NRC requirements as determined as a result of a prior inspection (NRC Inspection Report No. 99900300/90-01).
The process was developed using EPRI HP-5652, " Guideline for the Utilization of Commercial Grade Items in Nuclear Safety Related Applications (NCIG-07)," as a guide.
The program addresses the various modes through which the vendor obtains parts for supply to utilities.
These moAos include: _ _ _ _ _ _ _ _.
i 4
o In-houco manufacturing with matericle procured end manufacturing performed under the auspices of the~QA Manual.
These parts, as manufactured, are suitable for safety related service.
Manufacturing by an outside firm to meet ASME III (Section III), Class 1E, or seismic requirements, with or without an Appendix B QA program.
If the wanufacturer utilizes an approved Appendix B QA 1
program, the parts are suitable for safety related service.
)
Procurement from a-commercial grade supplier with the design of the part controlled by Coltec/FMED.
Procurement from a commercial grade suppl'ier with the design of the part not controlled by Coltec/FMED.
For the initial-mode and the case of an outside manufacturer with an approved Appendix-B QA program, no further classification ar.alysis or dedication activity need be performed.
For the latter three modes, however, procedures and forms have been developed as part of SP-750 that require the following factors to be considered and implemented to assure that a qualified part is supplied to the licensee:
o Classification: Does failure of the--part preclude the diesel engine from performing its safety related function?
Commercial grado: Can the part be procured from a supplier as a commercial grade item or does it have requirements specific to the nuclear 1 industry?
critical characteristics:'If the part is classified as e
safety related, what are its critical characteristics?
Seismic equivalency: Is the new-part identical to the replaced part with regard to part number, dimensions, and material of construction, or are the differences sufficiently insignificant to assure its seismic equivalency?
Assembly evaluation: Can the critical characteristics of an assembly of parts be evaluated without disassembly?
Dedication capability: Is sufficient design information e
available to permit proper dedication by Coltec/FMED or must a third party dedicator be utilized?, -
-2
SP-750 providra guidalin o for follure inpact review cnd safety function classification,. selection of critical characteristics, ovaluation for seismic equivalency, and selection of critical characteristics for assemblies.
A procedure for resolving nonconformance reports is also provided.. Appropriate forms are provided to permit documentation of the classification and dedication processes.
The inspectors' review of SP-750 resulted in the following observations:
The defin' tion of critical characteristics in paragraph 4.3 does not include the pertinent design, material, and performance characteristics-of the part relevant to its safety function; rather, it simply assures "that the item specified is the item received."
For the auditing of potential suppliers that maintain a 10 CFR 50 Appendix B QA-program or an ASME Section-III program, paragraph 6.3.5 addresses a programmatic-review but does not require a verification of the i
proper implementation of the program.
A similar comment applies to the auditing of third party dedicators. addressed in paragraph'6.5.21 the audit-does 1
not require a verification of implementation effectiveness.
This observation was also-noted in the Coltec/FMED Quality Engineering Instruction Serial No. 0111, " Approved Supplier (Supplier Rating ~ System),"
used to evaluate suppliers' quality programs.
For the procurement of parts from non-Appendix B suppliers, paragraph 6.5.3.4 does not addrous the need for a full dedication process unless a-commercial grade survey or source verification provides assurance that a critical characteristic item is properly provided by the supplier and a certificate of conformance is received attesting to that fact, Similar to the above observation, Appendix D, o
" Guideline for Critical Characteristics of Assemblics,"
does not address the approach for verifying the internal critical characteristics of an assembly that may not be reasonably disassembled for detailed examination.
While functional testing of the part would attest to acceptability on a short term basis, subtle deficiencies such as use of incorrect materials or off-dimensions will likely result in long tore failure.
Auditing of-the manufacturer with special' emphasis on-assuring proper. manufacturing controls for the critical characteristic features would provide the necessary assurance.
6 Porcgrcphs 6.3.13. 6.4.14, cnd 6.5.4 do not occuro th0t l
l'icensees are clearly informed regarding the classliicationfof the part_and the status of its dedication at the time of its delivery.
Based on'the' inspectors' review and pending resolution of the observations listed above, it was concluded that SP-750 describen an acceptable process for classifying and dedicacing cummercial grade spare and replacement parts for use in safety related service on emergency diesel generators.
Further review of your final program will be the subject of a future NRC inspection.
3,4 RgdEgff Procuremont: Packnoes To avnivate the offectiveness of the vendor's' implementation of SP-750,, the encpectors reviewed the engineering and purchasiat/ porti4tc of 19 procurement p9ckages documenting the supply of sanoty related diesel engine parts to licensees.
The peita reviewed included engin, governors, heat exchangers, C-ring 2, air filter / regulators, capscrews, solenoid valves, fused disconnect switch, relays, gasket, fuel injector nozzle, fuel injector assembly, spring for a,
fuel injector assembly, and a fuel oil pump.
The pertinent aspects of those part reviews are described below:
Enaine GovernnrR!
Two procurement packages dated 6
January and February 1992, respectively, were reviewed; one involved the repair of a UGB-10 Woodward governor for the North Anna Station, and the other involved the purchase of a new UGB-10 Woodward governor _for the Limerick Generating Station.
Both governors were evaluated to perform a safety related function.
The procurements were contracted to Woodward' Governor which is a ccamercial grade manufacturer of governors for diesel engines.
To qualify the manufacturer, Coltec/FMED had previously performed a commercial grade survey in January 1991.. The inspectors' review of the survey results indicated that there were several weaknesses.. They included the lack of adequate documentation of:
all of the governor models covered by the survey; the evidence to support the conclusions of the survey; and the specific design, material, and performance characteristics, and assessment of their QA controle, relevant to the safety functions of each of the governor models.
In accordance with SP-750, Form BF5378 was utilized to document the critical characteristics of the governors for verification.
In each case, however,-the critical "
-.=
chorectoriotica id:ntified w:ro difforcnt.
For tho initial case, the critical characteristics were as follows:-
visual inspect for damage e
e Woodward's test per TSP-453 Woodward's design per Drawing No. 8240-719 mounting holen e
e serrations on drive shaft e
serrations on the terminal shafts e
shaft extension from the governor case pin connector e
For the second case, the list of critical characteris-tics-did not include the last five interface items but-did include verification of nameplate markings and documents.
The vendor agreed that the set of critical cheracteristics for the second governor was deficient and attributed that occurrenc9 to their. system _of assigning different part numbers to:the same part because of a different customer.: -In addition, a difierent engineer was likely assigned to identify.the critical characteristics for the second governor.
Deficiencies were also noted'in the program for testing the governors.
Woodward's test procedure, TSP-453, only addressed the set-up and calibration of the gover-I nor; it did not address the full range of operating' conditions the governor would see during the perform-ance of its safety function.
Further,-the inspectors noted inconsistencies between-the two test reports provided by Woodward; the droop setting.was 7-for one of the governors and marked N/A for.the other, and the high speed was set at 1115 rpm for one governor and 1100 rpm for the other.
Heat Exchancet_Agnembly _ A water jacket coulant heat exchanger consisting of three separate shell-and-tube units with integral piping connected by_ mechanical joints was supplied to the vogtle nuclear station.
The-assembly was manufactured as commercial grado by ITT-Standard and supplied by Coltec/FMED as'a. safety related part.
The critical characteristics identified for this assembly were:
_ verification of the location and dimensions of the 7
mounting brackets verification of the nameplate markings on each.
e heat exchanger. unit -W
+
Tha coicmic equivalency of tho accccbly vac ovaluntcd b:ccd on tho fcct that tha iten was ordarcd'froa tha same manufacturer as the original.hoat exchanger assembly.
Hovover, the audit of the manufacturer.did not verify that measures for the control of design, process, and material _ changes were being adequately implemented, and, therefore, the inspectors could not conclude that seismic equivalency was demonstrated.
e 0-Rinas:
A purchase order for 31 0-rings was pro-consed by Coltec/FMED for the Wolf Crook nuclear station to be used to double seal the threaded plug installed in the push rod mechanism for each diesel engino cylinder.
The 0-ring was determined.to have a safety related function since failure could degrado the load bearing capability of the diosol engine if-frag-montation and relocation of 0-ring material blocked lubo oil flow in small passagoways.
The critical characteristics of the 0-rings wore identified to be:
visual examination for damage and defects e
inner and outer diamotor and cross-section e
material characteristics by hardness measurement e
verify material as buna-N by specific gravity e
measuroment stretch test to verify flexibility e
Verification of the critical characteristics was documented on Form DFS378B as roquired; only a minor deficiency was noted in that no direct measurement of the cross-sectional dimension was mado.
An area of concern, however, was identified in that.tho 0-rings were ootained from the warehouse where the vendor indicated that lot control was not exercised, i.e.,
0-rings had L9en precured from different-suppliers and were stored in tha same bin.
The vendor examined a sample of three 0-rings from-the batch of 31 to assess the acceptability of the entire batch. -The inspectors concluded that a sample sizo of 3 may not be representative for a total-batch-sizo of 31 due-to the mixing of similar parts from differont suppliers.
Air _P11ter/Reculator:
A purchaso order for eight filtor/ regulators van processed by the vendor for the Peach Bottom and Limerick nuclear stations.
While the function of the filtor/ regulator was datormined to be non-safety related, its failure could result.in declaring the dionel generator inoperable.. The vondor therefore decided this part-would be treated as safety related.
The-filter / regulator was'datormined to be 1 1 -)
l
colcaically cquivcicnt to previously cupplicd perto.
The critical characteristics were identified as visual examination for damage and defects o
visual examination of subparts e
e supplier's part number performance characteristics to verify air flow e
capacity and pressure drop e
leak test The insrectors indicated that the leak test did not specify an acceptance criterion and a required test duration.
Coltec/FMED agrsed to modify the applicable Form BF5378B to include these requirements.
The vendor verified the critical characteristics on a sample of 2 filter / regulators; as in the case of the 0-rings, the inspectors expressed a concern regarding the small sample size considering the lack of a lot control program.
Capscrews:
A purchase order for 15'capscrews was processed by Coltec/FMED for the Wolf Creek nuclear station.
The capscrews.were determined to be safety i
related since they are used to-secure the cover on the air manifold.
The capscrews were-procured commercial grade and were determined to be seismically equivalent to the previously_used items.
The critical characteristics of the capocrews were identified to be the following visual examination for damage, defects, and e
markings e
material hardness e
thread tolerance length Coltec/FMED verified the critical characteristics for a sample of 3 of the 15 capscrews.
As above, the lack of a lot control program made the sample size suspect..
Coltec/FMED provided 2 solenoid valves to Georgia Power. Company for.the Vogtle nuclear station for use in the air starting system for the diesel engine.
Because of their safety function, the valves were classified as safety related.
They were procured from a commercial grade supplier and assessed to be seismically equivalent to the_previously supplied-part., -.
i The critical characteristics woro identified on Form BF5378B to be:
i visual examination for damage, defects, and model e
number o
size and location of conduit connections and mounting slots
-material suitability by functional test to verify air flow and valvo position, onorgized and de-energized The inspectors made the observation that material suitability / acceptability may not be adequately demon-strated by a functional test, especially over long term operation.
No evidence was provided that auditing of the manufacturor had been performed to verify adequate control of the manufacturing process.
Fused Disqqnnect Switght Coltoc/FMED provided a fused disconnect switch, a device to hold fuses to protect various parts of the control circuitry, to a licensee for a diesel engine.
The switch was procured as a commercial grado item, and was determined to be seismically equivalent to the previously supplied part.
The fused disconnect. switch was evaluated to perform a safety related function.
The critical characteristics of the switch were identified to be as follows:
visual examination for damago and defects continuity check with fuses installed e
continuity check of switch in "on" position e
discontinuity check of switch in "off" position e
The inspectors concluded that the dedication activity for the fused disconnect switch was acceptable.
Rolava:
Three different procurement packages involving relays for diesel engino control circuitry for the Calvert Cliffs nuclear station were reviewed.
The relays performed reverse power, undervoltage, and overvoltage functions and were procured from Westinghouse as commercial grado items.
The relays were determined to perform safety functions and therefore were classified as safety related.
Soismic equivalency was indicated due to their similarity to previously provided relays.
The listings of critical characteristics for the relays included the following items:
9-rym w..-
--,.7-s-,
9.r,e
-+
e
,---r"-
---+--v-
"-r
+
er--
o Vorify canufccturcr'o identification o
Vicus11y incp;ct for dncega cnd dafccto Verify dimensions of width, height, and depth e
Verify material of construction for the caso-and the cover check number of terminals, verify electrical e
continuity /open circuits between terminals, check for grounds Verify proper ganket and offective seal for cover.
e Perform functional test to' verify operation with proper. voltage The inspectors noted that no evidence was provided to preclude the need for examination /varification of the relay internals regarding-materials of construction or dimensions.
No audits'of Westinghouse, the relay manufacturer, were available to assure proper controls were implemented during manufacture.
Completed.BF5378B forms and manufacturer's CoCs were included in the files.
Lube Oil System Casket:
A gaskot for the engine lubo oil system was provided to Florida Power Corporation's crystal River nuclear station.
The gasket was procured as a commercial grado item, but its function to provont a breach in the lube oil system pressure boundary was deemed sufficiently important to classify the item as safety related.
The critical characteristics of the gasket wero identified to be the following:
visual inspection for damage and defects e
verify compressibility of the material e
e dimensional check The gasket compressibility was measured using a Coliec/FMED device and found acceptable.'A CoC was Issued to the licensee.
The inspectors concluded that this dedication' activity was satisfactory.
Fuel Iniector Nozz3g:
Tho; inspectors reviewed the dedication package for a fuel injector nozzlo.- The part was determined to be safety related sinco its failure could prevent the engine from sustaining its rated capacity.- The nozzle was procored by Coltec/FMED as a commercial grado item.
The critical characteristics of the nozzle were identified as:
visual inspection for damage and defects e,,
' functional tost'for parformanco o
Testing was accomplished using a written procedure and a 11artridge Nozzle Analyzer to pressure _ test the -nozzle at 2200-2300 psi for zero leakage and no after-drip.
i The nozzle was also checked-for chatter over<tho. full operational velocity range with requirements of a minimum velocity of 1 stroke /sec with thorough.
)
atomization and dischargo through all orifices.
Materials of construction, dimensions, and soismic equivalency were not listed as critical characteristics, nor was evidence of an audit of the manufacturer conducted by Coltoc/FMED mado available to demonstrato proper manufacturing controls to assure conformance to design requirements.
Fuel Iniector Ascombiv Fuel injector-assemblies were purchased by Georgia Power Company and Philadelphia Electric Company (different nozzle holder). for their respective nuclear stations.
The fuel injector assembly consists of 14 individual parts, seven of which are manufactured by Coltec/FMED under. its QA program which meets the requirements =of Appendix B while the remaining seven parts are procured on a commercial grado basis,from an.outside=manatacturer.
The assemblies were considered to be safety-related, t
The critical characteristics identified for verification were:
visual inspection for damage and.defecta verification of part name, manufacturer's logo, e
and assembly part number verification of operability by conduct of-a functional test 9
Functional testing was accomplished by written
(
procedure and included slow and fast. lift. tests and spray pattern checks of stream and cono flow.
Cocs were issued tc the licensees.
11owever, natorials of construction and dimensions were not identified as critical characteristics (with the exception oflthe spring discussed below), nor was there evidence that.
Coltec/FMED performed an_ audit of the manufacturer of the soven commercial grade. parts to assure that proper _
design and manufacturing. controls were established and implemented, Sorins for Fuel Inicetor Assembiv:
The spring is one; of the seven purchased parts for the fuel = injector I occembly.
It wao procured from a comm3rcial greda nanufacturer and is determined to be safety related.
The critical characteristics of the spring were identified to be the following:
verification of part number and tagging e
dimensions including jnside diameter, wire e
diameter, and numbetlof coils spring constant e
a material of construction Verification of the material was accomplished by an outside contractor using destructive testing techni-ques.
The inspectors questioned the method of obtaining a representative sample since the springs are stored in onc bin in the warehouse regardless of manu-facturer, material lot, or order number.
Therefore, it is uncertain whether the springs delivered to the licensee are adequately represented by the sample-tested.
No other dir,repancies were noted for this dedication activity.
Eue) Oil Pumn The inspectors reviewed the dedication activity for a fuel oil pump-which is an electric pump used to prime the engine fuel oil headers and provide a l
backup to the engine-driven pumps.
Its failure modes i
resulted in a determination that the pump is safety related, but-it was procured from the supplier as a commercial grade item.
The critical characteristics of the pump were identi-fled to be:
visual inspection for damage and defects e
verification of part number and model e
verification of inlet and outlet port locations e
transverse mounting hole spacing e
distance from the discharge opening to the pump e
end mounting holes e
functional test Functional testing was performed in accordance with a written procedure to verify adequate pumping capacity.
Materials of construction were not identified as 4
critical characteristics, and no-evidence of an audit by Coltec/FMED was available to demonstrate adequate controls during the design and manufacturing processes to assure conformance to design requirements.
No other.
discrepancies were noted.
l _,
Based on the reviews of the procuremant packageo, ths inspectors concluded that several deficiencies oxisted in the implementation of SP-750 that require correction to assure properly qualified parts.
These observations were believed attributed to the early stages of development of the new program and could be properly resolved through further exporlonce with implementation of the now program.
In the above examples, the inspectors concluded that, although sore degree of dedication was done, weaknesses were rvident.
The inspectors did not considor that the weaknesses resultad in the use of parts that could cause safety problems, however, they were unable to assure that this was the case.
Accordingly, for those.
identified and similar items, Coltec should review the dedication to assure that all parts meet the current dedication program requirements.
If problems are identified, Coltoc should take the necessary stops to notify all affected customers in accordance with 10 CFR Part 21.
3.5 Licensee Audits The inspectors reviewed the results of the various audits of the Coltoc/FMED program conducted by nuclear plant licensees for the supply of spara and replacement parts for emergoney diesel generators.
The audits reviewed were conducted over the period of late-1989 to mid-1992 which is the time during which the new standard practice was under development and initially implemented by Coltec/FMED.
The licensee audits reviewed included the following:
ERC Environmental and Energy Services Co. conducted for GPU Nuclear Corp. and Iowa Electric Light and Power Co.
on October 3-5, 1989 Baltimore Gas and Electric Co. on September 10-14,.1990 South Carolina Electric and Gas Co. on October 15-16, 1991 Southern Nuclear Operating Co. conducted for Alabama Power Co. on November 20-21, 1991 NUPIC Joint Utility Audit conducted by Detroit Edison Co.,
Baltimore Gas and Electric Co., and Florida ?ower Corp. on January 27-31, 1992 Union Electric Co. and Northern States Power Co. on July 23, 1992 Several of the licensees' findings resulted in significant modifications to the Coltec/FMED program for dedicating commercial grado parts for safety related service.
The major modifications included the following:
Analysis of the supplied part for seismic equivalence to the previously supplied part - -. - -
~
Id nt'fication of_tha!clecaification bf tho part i
o
- supplied with the_ delivery"of the part order.
o.
? Analysis of the design requirements, _ industry usage',
.and-manufacturing controls for a;part to determine whether11tfis a commercial gradeiltem Documentation of the! rationale for: evaluating 1the-o safety function of:a part:to-determine its classification.'
Inadequate-failure-modes and effects analyses resulting in the need for additlanal personnel training.
Other findings-identified were related primarily to1 1mple--
= mentation-difficulties _ involving documentation and shop' problems.
File records indicated that Coltec/FMED had-developed appropriate corrective actions for essentially;all.
of-the licensee findings.
The inspectorsiconcluded that-the licensee audits and findings resulted.in significant contri-butions to the effectiveness of the Coltec/FMED program for the dedication of commercial' grade parts.for safet'; related~
service.
4 COLTEC/FMED PERSONNEL CONTACTED The following Coltec/FMED personnel were contacted during the course of_the inspection:
+
- Mary Armfield Senior Quality Assurance-Engineer
+
- Jeanne Eskes Buyer
+
- Todd Magnuson Warehouse Manager
+
- Bruce Schoenike Chief Inspector
+
- Ted Stevenson Parts and Service Representative-
- Warren Martin Director, Sales and-Marketing
+
- George Ferriter-Engineer
+
- William Satek Vice President,: Quality Assurance
+
Michael Horinka Manager, Quality Assurance, Nuclears
'and Military-
- Norman Traeger Supervisor, Mechanical Application Engineering
+
- Scott Fratianne
_ uality Assurance Engineer-Q
+ Attended the entrance meeting on August 25,-1992
- Attended the exit meeting on August 27, 1992 l l o
(
u
.[so mek,Ig -
UNITED STATES g.,, f j NUCLE AR REGULATORY COMMISSION k
WASHINCTON. D. C. 20555 :
- b
'[
'\\,hg}/
November 4, 1992 Docket No.
99901255 Mr. G.'Gardner, Chief Executive Officer Connex Pipe Systems, Inc.
1115 Gilman Stree.t Marietta, Ohio 45750
Dear Mr. Gardner:
SUBJECT:
NOTICE OF PONCONFORMANCE (NRC INSPECTION REPORT NO. 99901255/92-01)
This letter addresses the inspection-of your facility at Marietta, Ohio, conducted by Mr. L. L. Campbell of this office on-October 5 through 9,.1992,.and the discussions of his findings with members of your-staff at the conclusion of the inspection.
The performance based inspection was conducted to evaluate Connex.
Pipe Systems, Inc.'s.(Connex's). quality program and its implementation in selected area such as (1) purchased material and services and controls for subsuppliers, (2) control of work processes,-(3) inspection and tests,-and (4) Connex's ASME Code certification program.
Areas examined during the U.S. Nuclear Regulatory Commission l
(NRC) inspection and our findings are ' discussed in the enclosed inspection report.
This inspection consisted of an examination l
of procedures and representative' records, interviews with personnel, and observations by the inspector.
During the inspection it was found that the implementation of your quality assurance program = failed to meet certain NRC requirements.
The inspection identified instances in which Connex failed to follow its Nuclear Quality Assurance Manual and applicable implementing procedure requirements for preparing certificates of Compliance and for processing purchase orders.
The inspection also identified that the completed. checklists for the 1990 and 1991 Connex audits of Ram Forge & Steel,-Inc.,
contained no documented objective evidence that special processes such as heat treating and forming, and certain ASME Code product.
inspections such as wall thickness measurements of forged products, had been reviewed.
The specific findings and.
references to the pertinent requirements are identified in the enclosures of this letter.
Mr.
G.
Gardner Please provide us within 30-days from the date of this letter a written statement in accordance with the instructions specified in the enclosed _ Notice of Nonconformance.
We will consider extending the response time if you can show good cause for us to do so.
Additionally,-during the inspection, the NRC inspector expressed a concern that Connex's practice for_ determining when welds-are inaccesible for nondestructive examination - (NDE) may result in not performing required ASME Code NDE.
Please provide us a discussion on how Connex's practice for determining when welds are inaccessible for NDE, as discussed in the enclosed inspection report, will not impact the performance of any-required ASME Code NDE that is accessible for NDE at some point during the fabrication process.
The responses requested by this letter and the enclosed Notice are not subject to the clearance procedures of the Office of Management and Budget as required by the Paperwork Reduction Act of 1980, Public Law No.96-511.
In accordance with 10 CFR 2.790 of the NRC's " Rules of Practice," a copy of this letter and its enclosures will be placed in the NRC Public Document Room.
If you have any questions concerning this inspection, we will be pleased to discuss them with you.
Sincerely, 1r J"' Leif Norrholm, Chief Vendor Inspection Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation
Enclosures:
1.
Notice of Nonconformance 2.
Inspection Report No. 99901255/92-01 l
l Encic;uro 1 NOTICE OF NONCONFORMANCE Connex Pipe Systems, Inc.
Docket No.:
99901255/92-01 Marietta, Ohio During an inspection conducted on October 5 through 9, 1992, it appears that certain of your activities were not conducted in accordance with NRC requirements.
A.
Criterion VII, " Control of Purchased Material,_ Equipment, and Services," of Appendix B to Title 10 of the code of Eederal Reculations (10 CFR) Part 50 requires, in part, that measures be established to assure that purchased material conforms to procurement documents and that the effectiveness of the control of quality by the supplier of material shall-be assessed by the purchaser.
Criterion XVIII, " Audits," of Appendix B to 10 CFR Part 50 requires, in_part, that a comprehensive system of planned and periodic audits be carried out to verify compliance with all aspects of the quality assurance (QA) program and to determine the effectiveness of the program.
Section 17.3.0,
" External Audits and Surveys (By Connex Personnel)," of the Connex Pipe Systems, Inc. (Connex)
Nuclear Quality Assurance Manual (NQAM), Revision 20,-dated March 13, 1992, requires, in part, that objective evidence be assessed to verify the implementation of a supplier's QA l
program and documented on the audit checklist.
Contrary to the above, there was no documented objective evidence in Section IX, "Special Processes," of the 1990 and 1991 Connex audit checklists of Ram Forge & Steel, Inc.,
(Ram Forge) to substantiate that requirements were in place and being effectively implemented for the control of: the forming and heat treatment processes.
Additionally,.there was no documented objective evidence in Section X,
" Inspection," or in Section XIV, " Inspection, Test, and Operating Status," of these audit checklists that Ram Forge had requirements in place which were being effectively implemented for the control of material vall thickness required by applicable sections of the ASME Code (Nonconformance 99901255/92-01-01).
B.
Criterion V,
" Instructions, Procedures, and Drawings," of Appendix B to 10 CFR Part 50 and Section 5,
" Instructions, 4" P Procedures, and Drawings," of the Connex NQAM requires, in.
part, that activities affecting quality be accomplished in accordance with instructions, procedures, and drawings.
The following are examples of quality related activities that were not accomplished in accordance with applicable Connex requirements (99901255/92-01-02):
1.
Sections 4.2.0, "Haterial/ Service Requirements," and 4.3.0, " Purchase Order and/or Material Transfer (Other than Weld Material)," of the Connex HQAM, Revision 20, dated March 13, 1992, and Connex Procedure No. GP-15, "ASME Order Entry and Customer Document Control,"
Revisions 5 and 6, dated August 27, 1991, and October 1, 1992, respectively, require, in part, that when documents are received, they will be reviewed by Project Engineering to determine their impact on work in process and on existing Connex documents.
Contrary to the above, the assigned project engineer for Gulf States Utilities Company (GSU) Purchase Order (PO) No. 91-J-73927, Revision 0, 3ated September 10, 1991, failed to correctly translate the GSU PO technical requirement for the 8-hour heat treatment of the forging material into a revision to Connex PO No. E8634-1, dated September 17, 1991, issued to Ram Forge for the manufacture of a safe end and its test specimen.
2.
Connex Procedure No. STD-GP-32, " Instructions For the Preparation of Certificates of Compliance," Revision 0, dated April 3, 1989, requires, in part, that the assigned QA engineer obtain the technical specification referenced in the customer's PO, review Connex's subsuppliers Certified Material Test Report and/or certification for conformance to the customer's requirements, and if acceptable prepare the Connex certification.
Contrary to the above, the assigned QA engineer issued a Certificate of Compliance, dated March 12, 1992, certifying that the requirements of GSU PO No. 91-J-73927, Revision 1, dated February 24, 1992, had been met, however the GSU PO requirement that safe end test specimens be subjected to an 8-hour heat treatment had not been met.
Also, Connex incorrectly certified on May 29, 1992, that another set of test specimens had been subjected to a minimum 1150*F heat treatment, as required by Revision 1 to GSU PO No. 91-J-73927, when the test data attached to the certification indicated that the specimens had been subjected to a minimum temperature of 1125'F. i
Please provide'a writtenlstatementior explanation-to$the':U,S.'
Nuclear. Regulatory Commission,j ATIN: Document Control Desk, Washington, D,C.J20555,.with-_aLeopy;to_thel Chief,: Vendor? ~ ~ '
Inspection: Branch, Division;of Reactor--InspectionLand Licensee _-
4 Performance, Of fice 'of: NuclearlReactor.: Regulation, withint30, days of the date of:the. letter transmitting this Notice;-of--
Nonconformance.- This reply should be clearly: marked'as~a " Reply!
to a Notice of Nonconformance"=and should include for:each
~ ~
. nonconformances (1)ia description of: steps _that have been orivill--
be taken to correct:these: items; (2) a' description: of steps =-that :
f have' been or'will be taken to prevent recurrence; _ and -(3) - the -
dates your corrective actions:and; preventive measures-vere'or.
will be completed, Dated at R_ockville, Maryland-day of-k h iLAWb O, 1992.
this UDi t
i-l
- 49-Wv u-
-r----
r--
ENCLOSURE 2 ORGANIZATION:
Connex Pipe Systems, Inc.
Marietta, Ohio REPORT NO.:
99901255/92-01 CORRESPONDENCE ADDRESS:
Mr. G. Gardner, Chief Executive Officer Connex Pipe Systems, Inc.
1115 Gilman Street Marietta, Ohio 45750 ORGANIZATIONAL CONTACT:
K.
A. Welch Quality Assurance Manager NUCLEAR INDUSTRY Hanufactures and supplies ASME Code ACTIVITY:
pipe subassemblies and material for commercial nuclear power plants.
INSPECTION CONDUCTED:
October 5 through 9, 1992 Yr /
Y ll Eh 2
L. L.ff' ampbell T '
Date Reactive Inspection Section No. 1 Vendor Inspection Branch OTHER INSPECTORS:
None II E 7
3 APPROVED:
M,eUlitis Potapovs, Chief 7
Date Reactive Inspection Section No. 1 Vendor Inspection Branch INSPECTION BASES:
10 CFR Part 21 and Part 50, Appendix B INSPECTION SCOPE:
To review and evaluate the Connex Pipe Systems, Inc. (Connex) quality assurance (QA) program and its implementation in selected areas such as (1) control of work processes, (2; purchased material and services, (3) inspection and test, and (4) Connex's American Society of Mcchanical Engineers (ASME) certification program.
PLANT SITE River Bend Station (50-458)
APPLICABILITY
- Hope Creek (50-354)
Other plants using Connex products..
1{ INSPECTION
SUMMARY
~1.1 Nonconformances Contrary-to Criteria VII and~XVIII of Appendix B to. Title 10-of the Code of Federal Reculationg (10 CFR) Part-50iand the Connex-Pipe Systems,-Inc. (Connex) Nuclear Quality _ Assurance Manual (NQAM), there was no documented objective evidence on the'1990-and-1991 Connex audit checklists for Ram Forge & Steel, Inc.f(Ram Forge) to substantiate that Ram Forge had requirements in place which were_being effectively _ implemented for the-control'of the forming and heat treatment processes and forithe control of material wall thickness (Nonconformance 99901255/92-01-01, see Section 3.3 of this report).
Contrary to the requirements of Criterion _V of. Appendix B to 10 CFR Part 50, the Connex HQAM, and the requirements specified in Connex implementing procedures,.connex failed to follow procedure requirements (Nonconformance 99901255/92-01-02, see Section 3.4 of this report).
2 STATUS OF PREVIOUS INSPECTION FINDINGS l
This was the first NRC inspection of Connex.
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Entrance and Exit Meetinas In the entrance meeting on October 5, 1992, the NRC inspector discussed the scope of the_ inspection and established interfaces with Connex management.
During the exit meeting on October 9, 1992, the NRC inspector discussed his findings and concerns with Connex's management and staff.
3.2 10 CFR Part 21 The NRC inspector determined that Connex has maintained _the required 10 CFR Part 21 postings and a procedure for implementing Part 21 requirements, Procedure No. STD GP-1, " Procedure For-Reporting Defects and Failures to Comply In Respect to Regulation 10 CFR 21," Revision 5, dated July 7, 1992.
This' revision of Procedure No. STD GP-1 reflects the revised 10 CFR Part 21 requirements issued by the NRC'that became effective October 29, 1991.
3.2.1 Imolementation of 10 CFR Part 21 Procedures Gulf States Utilities (GSU) issued purchase order.(PO) l No. 91-J-73927, Revision 1, dated February 24, 1991, to Connex !
'for the supplyLof1several items to be installed'inifeedwater
-piping to the reactor pressure-vessel (RPV), includingEPO~ Item No.
5, " Elbow, 12"190 degree, CS, SA-234, WPC,: -long radius, long; tangent'2 1/2. inch:both ends, scheduleHBO, coun'_erbore one'end to schedule _60' nominal', no weld prep, ASME III,fClass.l."
Connex issued'PO E8634-2, dated February 13, 1992,- to Ram Forge.for several items including PO Item _No.'1, the 12 inch elbow-specified by the GSU PO, and invoked 10 CFR Part 21 and-ASME,Section III,'NCA-3800, material requirements on all11tems manufactured by Ram Forge.
The NRC inspector discussed the_GSU-10 CFR Part 21 Report No. RGB-37042, dated June 18, 1992, with Connex and reviewed ~
Connex's 10.CFR Part 21 evaluation for the nonconforming elbow purchased by GSU on PO No. 91-J-73927. JGSU-reported'that~the elbow in question, a replacement for the first elbow off the-River Bend Station, Unit 1, RPV Feedwater' Nozzle N4A, was.
required to have a nominal wall thickness of 0.688_ inch and a code minimum wall thickness of 0.535 inch, but'had a measured minimum wall thickness of 0.378 inch in one local area.
Because GSU discovered this deviation after acceptance at receipt inspection and prior to its installation there was no' operational impact, however GSU noted in its Part 21 report thatL"due to the gross reduction in wall thickness a substantial safety hazard-could have been created-had this defect remained uncorrected-because the elbow is the first elbow off reactor pressure vessel feedwater nozzle N4A, an unisolable portion of the reactor coolant pressure boundary."
After being' informed by GSU of the nonconforming elbow.the:
initial corrective action taken by Connex was to remove Ram Forge from Connex's approved vendor list (AVL) and to document this.
condition on_Nonconformance Report (NCR) E-8634/QA2 and Corrective Action Request (CAR) No.92-007.
In a letter to the NRC, dated-July 8,.1992, ConneX provided additional information concerning the manufacturing of--the elbow by its subsupplier, Ram Forge.
This letter also identified actions taken by Connex to satisfy the evaluation and reporting requirements of 10 CFR Part 21.
During-the inspection the NRC reviewed the infornation and actions addressed in'Connex's July 8, 1992, letter to the NRC for compliance _to 10 CFR Part 21-requirements.
Following the removal of Ram. Forge from its AVL, Connex performed wall thickness measurements on 400 elbows from various subcontractors, with a size range from 4 to 36-inches in diameter-and a-wall thickness of schedule-10 to double extra heavy,- in'its-Marietta, Ohio, warehouse.
Measurements were taken on each elbow -
on the centerline of the outer arc at the 77.1/2 degree, 45 degree, and 22 1/2 degree locations.. The wall-thickness measurements were obtained using Connex Procedure No. ES-UT-6, ;
Revision 3, dated May 1,-1989, and revenicd that-ell _olbows measured contained acceptable wall thickness.
It11s noted that during the-inspection, the NRC inspector-observed Connex measuring _the1 wall thickness of an elbow in accordance:with Procedure No. ES-UT-6, Revision'3, and. concluded that'the-measurements were correctly taken (see Section 3.5.2 of thic report).
In order to determine the root cause and to verify corrective action to prevent recurrence, Connex, accompanied by GSU, went to Ram Fo.Je and determined that this particular elbow required ten heats to form the elbow within the die when normally only two heats are required.
Connex concluded that since the additional heats resulted in an excess of scale, and required substantial-grinding to remove, the elbow was ground below minimum wall thickness during the removal of the scale.
Based on discussions with Connex and review of Connex's Pos to Ram Forge,.the NRC-inspector concluded that Ram Forge did not_ routinely measure the wall thickness of products supplied to Connex unless the Connex PO specified that wall thickness measurements be taken.
In this instance, the Connex PO for the elbow did not specify wall thickness measurements.
According to Connex, Ram Forge did not question the wall thickness because past experience had shown their elbows to have much heavier wall thickness than required.
Ram Forge demonstrated to Connex this heavier wall thickness assumption by taking readings from completed elbows in the shop awaiting shipment to other customers.
Ram Forge informed Coktex that they will check all elbows for wall thickness on future ASME Section III Code orders, and'that the wall thickness measuremente would be taken on a grid over the-entire elbow.
During the NRC inspection it was determined that Connex had not placed Ram Forge back on its AVL and that Connex had not verified if Ram Forge had revised any procedures to require wall thickness measurements of ASME Section III Code fittings.
Connex informed the NRC inspector that Ram Forge would remain off its AVL until Connex verified that Ram Forge had controls in place to ensure that these wall thickness ~
measurements would be taken.
The NRC inspector discussed with Connex several ASME Code requirements that addressed wall thickness and were applicable for the manufacture of the nonconforming elbow.
Code requirements discussed included: (1) Section 12.3.2 and 12.3.4 of SA-234, " Specification for Piping Fittings or Wrought Carbon Steel and Alloy For Moderate and Elevated Temperatures," of the ASME Section II Code, 1992 Edition, (2) Manufacturers Standardization Society (MSS) Standard Practice SP-87-1982,
" Factory Made Butt-Welding Fittings for Class-1 Nuclear Piping Applications," including the SP-87-1982 Errata For paragraph 3.1, and (3) NB-2121, " Permitted Material Specifications," NB-3690,
" Dimensional Requirements For-Piping Products," and NB-4214, "Minitun:Thickn000 of'Fct!ricatcd!'MatOricl," of thi: ASME
'Section III Code.
The NRC inspector emphasized:the importance-of meeting these Code requirements andLthe need_toEverify that subsuppliers had_ controls in place-for assuring that these requirements were-met.
The'NRC inspector then reviewed Connex's
'1990 and 1991 audits of Ram Forge to determine if-these_ audits
-addressed applicable code requirements (see-Section 3.3 of'this report). -
Based on the results of the review at Ram Forge and the measurements taken on elbows in the Connex and Ram Forge shops,__
Connex concluded that the-nonconforming elbow in question was an isolated occurrence.
Although the actions by Connex and the results of their investigation at Ram Forge support-_the Connex conclusion that-this occurrence is isolated to its supply of the elbow to GSU,, valuations by Ram Forge to substantiate that the nonconforming elbow is an isolated manufacturing error has not been reviewed by the NRC inspector.-
3.3 Review of-Connex's Audits of Ram Force The NRC inspector performed a limited-review of Connex's audits of Ram Forge performed on October 9 and 10, 1990, and on October 8 and 9, 1991, in order to determine if Connex adequately verified that Ram Forge was effectively implementing-controls for special processes and inspections.
The revision of Connex's NQAM in effect when these audits were conducted, Revision.18, dated March 21, 1991, failed to addressed certain requirements of NCA-4134.18, " Audits," of_the 1986-Edition of the ASME Section III Code which Connex was centractually obligated to meet for the supply of Code fittings (i.e. the elbow discussed in Section 3.2.1 of this report and a safe end in accordance with GSU PO No. 91-J-73927, Revisions 1 and 2).
NCA 4134.18 requires, in part, that objective-evidence'shall be examined to verify the implementation of the QA program and documented.
It is noted that Section 17.3.2 of the Connex NQAM, Revision 19, dated October 21, 1991, and the current Revision 20,. dated March 13, 1992, include this requirement.
The NRC inspector also identified to Connex that Criterion VII,
" Control of Purchased Material, Equipment, and Services," of Appendix B to 10 CFR Part 50 requires, in part, that measures be established to assure that purchased material conforms to procurement documents and that the effectiveness of the control of quality by the suppliers of material shall.be assessed by the purchaser.. Additionally, Criterion XVIII, " Audits," of' Appendix B to 10'CFR Part 50 was identified as requiring that a i
comprehensive system'of planned and periodic audits be carried out to verify compliance with all aspects of-the QA program and to determine the effectiveness of the program.
- Further, l
Section 17.3.0, " External Audits and Surveys (by Connex Personnel)," of the Connex NQAM, Revisions 19 and 20 (applicable t
p !
-revicions' for th3 GSU PO)' requiros, in-part, that objcctivo-___
evidence be assessed to verify:the implementation of,a. supplier's-QA program and it be documented on the audit checklist.
The NRC inspector. informed'Connox that contrary to'the above, there was no document 9d objective evidence'in Section1IX,
" Control of Special Processes,"Section X,
" Inspection," or Section XIV, " Inspection, Test'and Operating Status," of the 1990 and 1991 Connex audit checxlists of Ram-Forge to substantiate-that requirements were-in place and being effectively' implemented for the control of the forming and heat treatment processes and for the control of material. wall _ thickness required by the ASME Code (Nonconformance 99901255/92-01-01).
Both the 1990 and 1991 audit checklists listed forging, heat treatment, and marking in the summary for checklistLItem~IX, however, there was no objective evidence listed to substantiate that Ram Forge's QA and procedure requirements for the forging and heat treatment processes were being effectively implemented.
Also,Section IX of the checklists did not identify any procedure, forge order, unique instructions or records, or-activity observed during the audits to substeetiate that Ram Forge was controlling these processes.
Simila_ly,Section X and XI of these audit checklists provided no objective evidence that Ram Forge's inspection program required any thickness measurements of the completed forged product.
It is;noted that:
the cover sheets for the 1990 and 1991 Connex audits of JamgForge indicated that the audit checklist attributes applicable for Inspection, Test Control, and Inspection, Test and Operating Status were not applicable for the audit scope.
The NRC inspector discussed the audits of Ram Forge with personnel performing the audits and reviewed back up notes and referenced documents for the various checklist items.
Although.the 1991 auditor's handwritten notes were made available for review and L
listed forge orders and identified heat treatment furnaces, it-I was unclear to the NRC inspector which audit checklist attributes these references supported-or what associated activities were verified by the auditor for the references.
The auditor informed the NRC inspector that although not reflected in the handwritten notes, a review of Ram Forge's administrative controls for the forging process was performed and found satisfactory..
3.4 Reactor Vessel Feedwater Safe End The NRC inspector reviewed Connex's procurement.and certification-activities supporting the supply of a RPV feedwater safe end to GSU, for use at the River Bend Station, Unit 1.
On August 9, 1991, GSU forwarded an advance, " Info only" copy of-a purchase requisition (Request for Quote-No. XX 75054) to Connex for the purchase of a 16 inch long x 9.5 inch inside diameter (ID) x 15.25 inch outside diameter forging manufactured to the -.
rsquirczOnto of ASMELSA-508,.Clc03--1,-Ond cartificd cofASME Section III, Class NB, 1974 Edition through thel 1976 Winter
- Addenda.- On August 11, 1991,-- GSU forwarded to Connex additionali information via facsimile ~which served as the basis for' ordering the forging prior'to the receipt of GSU's' formal PO.-
Based on this information, Connex issued PO'No. E8634-1, dated September 17, 1991, to Ram Forge for the_ supply:of the forging.-
On September 19, 1991,LConnex-received the formal GSU PO-for the safe end forging,-PO No. 91-J-73927, Revision 0, with an authorization signature dated September 12, 1991.
.This PO contained a requirement that the safe end be heat treated to a minimum tempering temperature of 1150*F for a minimum-of 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.
It is noted that the preliminary GSU procurement: documents to Connex did not require or address the extended 8-hour heat treatment of the forging.
The NRC inspector determined that:
contrary to the requirements of Section 4,
" Procurement: Document ~
Control," of the Connex NQAM and Procedure No. GP-15,~"ASME Order Entry and Customer Document Control," Revision 5, dated August 27, 1991, the assigned project engineer _did not adequatelyE review the formal GSU PO for impact on existing procurement documents to Ram Forge or on in-process work and thus failed to incorporate the 8-hour material heat treatment into Connex's PO to Ram Forge (Nonconformance 99901255/92-01-02).
On January 14, 1992, Connex received a preliminary copy _of GSU PO-No. 91-J-73927, Revision 1, dated Decenher 17, 1991, for the procurement of additional ASME Section III, Class 1, fittings such as the elbow discussed in Section 3.2.1 of this report.
On March 16, 1992, Connex received the formal Revision 1 to GSU PO No. 91-J-73927, with an authorization signature date of. Marche 13, 1992, which contained different requirements than the preliminary-version such as applying the.8-hour heat treatment to the test specimen and only a 3-hour heat treatment to the ring forging for the safe end.
The NRC inspector reviewed two Connex Certificates of Compliance (CoC) to Revision 1 of GSU PO 91-J-73927, for the supply of the forging for the safe end and determined that contrary to the requirements of Connex Procedure No. STD-GP-32, " Instructions For-the Preparation of' Certificates of Compliance," Section H,
" Steps For Completion," the assigned QA engineer issued a CoC, dated March 12, 1992, certifying that the material supplied met the requirement of GSU PO No. 91-J-73927, Revision 1,~even though the CoC package contained no objective evidence that the test specimens for the safe end had been subjected to the 8-hour heat treatment.
Also, Connex incorrectly certified on May 29, 1992,.
l that test specimens'had been subjected to a minimum of 1150'F i
heat treatment, as required by Revision 1 to GSU PO No. 91-J-73927, when the-test data attached to the certification i
indicated that the specimens had been subjected to a_ temperature i
of 1125'F (Nonconformance 99901255/92-01-02).-
- - ~-~-
l The NRC inspector reviewed numerous CoC, procurement documents, and written correspondence between Connex, GSU, and Ram Forge-regarding heat treatment ~of test specimens.
Based on the documentation presented to the NRC inspector and explanations by Connex personnel, it was determined that two out of the three test specimens produced to verify properties of the safe end wore unacceptable.
The following paragraphs describe the three attempts to properly produce the test specimens:
1.
The first set of test specimens received the same heat treatment as the forging for the safe end.
The test results were rejected by GSU because the 8-hour heat treatment had not been applied.
2.
On or about May 5, 1992, the second set of test specimens received an additional heat treatment of 1650*F for 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> and were water quenched, followed by a tempering at 1200*F for 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, followed by a heat treatment at 1125'F for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.
Although the documentation was somewhat unclear, Connex informed the NRC inspector that in addition to being improperly heat treated, these test specimens were improperly prepared because the test specimens were taken too close to the heat treated surface.
These test results were rejected by Connex because of-the excessive heat treatment and the improper preparation of the test specimens.
3.
On May 28, 1992, under the surveillance of GSU, test specimens were correctly prepared, the required heat treatment for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> was properly applied, and the safe end test specimens properly tested.
These test results met the requirements of GSU PO No. 91-J-73927, Revision 2, dated May 29, 1992, which reduced the 8-hour heat treatment temperature to 1100*F.
During the conduct of the inspection, Ram Forge provided documentation to Connex which revealed that the forging from which all test specimens were obtained came from a prolongation of one of the cylinders that was used for the manufacture of the safe end.
This prolongation, excess material from the forging process, was subjected to the same heat treatment cycle as the safe end (austenitized at 1650*F for 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> and quenched in water followed by a tempering at 1200*F for 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> and air cooled).
3.5 Reactor Veggel Feedwater Pinina Elbow The following sections of this report discuss the NRC inspector's review of the procurement documents for the nonconforming elbow described in Section 3.2.1 of this report, the repair of this olbow by Connox, cnd cctionc teken by Conn:x wh;n the rcpnired elbow was rejected by GSU and returned to Conn 3x.
3.5.1 Initial Procurement and Receint of the Elhag GSU issued PO No. 91-J-73927, Revision 1, to Connex for the supply of Item 5, the elbow-discussed in Section 3.2.1 of_this report.
Connex issued PO No. E 8634-2, dated February 13, 1992, to Ram Forge for the supply of the elbow.
Based on the review of documents presented to the NRC inspector, it appeared that Connex correctly translated GSU PO technical and quality requirements into the Ram _ Forge PO for the elbow.
Connex informed the NRC inspector that the test specimens for the elbow were taken from a prolongation of the elbow which remained on the elbow throughout all heat treatment.
Following the receipt acceptance of the elbow by GSU, but prior to the installation of the elbow, GSU discovered that the elbow did not have the required minimum design wall thickness.
GSU reported this condition to the NRC, as discussed in Section 3.2.1 of this report, and returned the elbow to Connex for repair.
3.5.2 ReDair of the Elbqy GSU issued PO No. 91-J-73927, Revision 3, dated June 10, 1992, to Connex for the repair of the elbow in accordance with NB 4130,
" Repair of Material," of the 1986 edition of the ASME Section III Code.
Connex informed the NRC inspector that Wold Procedure Specification E 8634-RP-1, Revision 1, dated June 9, 1992, was used to deposit weld filler material on an area of the outer arch of the elbow approximately 12 inches wide x 24 inches long and locally on other portions of the elbow.
After welding, grinding, and performing the required nondestructive examination (NDE),
Connex, with surveillance by GSU, verified that the elbow wall thickness was 0.602 inches or greater.
The wall thickness measurements were obtained using a digital ultrasonic thickness gauge in accordance with Connex Procedure No. ES-UT-6,
" Ultrasonic Thickness Gauge For Thickness Measurement,"
Revision 3, dated May 1, 1989.
The NRC inspector performed a limited review of the elbow repair documentation and found no deficiencies.
The NRC inspector questioned Connex regarding the local structural discontinuities created by such a massive weld repair and what effects, if any, this would have on-the stress indices provided in Tables NB 3685.1-1 and NB 3685.1-2 of the ASME Code, and if Connex had been requested to evaluate the local discontinuities created by the veld repair.
Connex informed the NRC inspector that such evaluations were outside the scope of its PO from GSU.
The NRC inspector did not consider this question to be relevant after being informed by Connex that the elbow would be destroyed in order to determine the exact wall thickness of the areas described in Section 3.5.3 of this report.
3.5.3 Insoection of the Renaired' Elbow Reiected by GSU
'l On August 21, 1992, a copy of GSU Receiving _ Inspection' Report-(RIR) No. 92-RIR-00729, Revision No. 2, dated August 20, 1992, was forwarded to Connex and identified that the repaired elbow had an area'which did not meet the minimum wall thickness requirements of 0.602-inches.
The RIR stated,-in part, that
-using outside calipers, readings of 0.515 and 0.530 inches were obtained at Column T/ Rows 10-11, and that the elbow should be rejected and returned to Connex because the minimum wall thickness, 0.602 inches, was not present.
Connex generated NCR E-8634/I-3, on September 29, 1992, to document the_ condition reported by GSU and"to control the returned elbow.
The NRC inspector observed Connex perform the following inspection activities on the repaired elbow:
1.
Wall thickness measurements were taken using Procedure No. ES-UT-6, Revision 3, in the areas identified by GSU as being below minimum wall and in surrounding areas approximately 2 inches from Column T,-Rows 10-11.
The lowest reading _Connex obtained in these areas using the ultrasonic thickness gauge was 0.619 inches.
2.
Connex cut the elbow into three sections, with one cut adjacent-to the Row 10-11 area.
Connex then performed wall thickness measurements using a micrometer in the same areas that were previously measured using the ultrasonic thickness gauge.
The lowest reading Connex obtained in these areas using the micrometer was 0.609 inches.
The NRC inspector observed Connex performing the measurements with the ultrasonic thickness gauge, the cutting of the elbow, i
and measuring the sectioned elbow with the micrometer.
Connex appeared to have performed all measurements in accordance with applicable procedures and the disposition of NCR E-8634/I-3.
3.6 In-Drocess Work Activities The NRC inspector observed Connex's inspection, examination, and-test activities for two piping subassemblies being fabricated for Public Service Electric and Gas's (PSE&G's) Hope Creek Nuclear Plant.
PSE&G issued Release No. 36, dated September 4, 1992, to PO_NO. B1-360909 for the fabrication of 16 pipe spools and the supply of 8 pieces of material for the Hope Creek reactor building service water system.
The PO invoked the requirements of the 1974 Edition through the 1974 Winter Addenda of the ASME Section III Code and for certain items invoked the 1930 Edition of ASME Section III Code and Code Cases N-454, N-455, N-453-1, and N-439.
Th3 NRC in pictor conductCd c licitcd review of the procedures, documentation, and procurement documents and observed some in-process inspection, examination, and-testing for Pipe Spool' No. 1-EA-009-S05A on Connex Sketch No. E-4269-9, Revision 0, dated August'20, 1992, and-Pipe Spool.Ho. 1-EA-0090-S06 on Connex Sketch No. E-4269-11, Revision 2, dated September.28, 1992.
In cddition to reviewing documentation and observing activities performed by Connex, the NRC inspector performed visual cxaminations of selected welds and internal inspections of pipe cubassemblies, and reviewed the configuration of piping-subassemblies and material wall thickness measurements.
As a result of the NRC inspector's reviews and inspections the following items were discussed in detail with Connex:
1.
Sketch E-4269-11 requires that the fabrication of the pipe spool be performed in accordance with Connex Fabrication Procedure No. E-3974 SP-1, "PSE&G Hope Creek," Revision 4, dated August 9, 1991.
Section 8.1.1.C of Procedure No. E-3974 SP-1 requires a 100 percent liquid penetrant examination of the accessible internal and-100-percent external surfaces of fillet and structural welds and adjacent base metal.
Connex did not perform a liquid penetrant examination of the internal surfaces of Welds M and N, 8 inch diameter x 6 inch long stanchion pipe to 36 inch diameter service water pipe on Sketch No. E-4269-11, full penetration structural welds with a reinforcing fillet weld, because Connex claimed it was inaccessible due to the welding of a plate that capped the end of the stanchions.
Connex informed the NRC that it was their practice to complete all fabrication and then to present the pipe spool for final dimensional inspection and NDE, and that after the completion of the fabrication accessibility is determined.
The NRC inspector questioned this practice since there were no provisions in either the PSE&G PO or Procedure No. E-3974 SP-1 which addressed or permitted this practice.
Following this discussion, Connex contacted a staff engineer in PSE&G's Nuclear Engineering Piping / Materials organization who agreed, via a note on a facsimile, with Connex's practice for determining that these welds were inaccessible for liquid penetrant examination.
The NRC inspector expretsed a concern that Connex's practice for determining when welds are 17 accessible could result in not meeting its customer's requirements, as well as not performing certain Code required nondestructive examinations.
2.
The NRC inspector questioned 15 ground areas, approximately 1 1/2 x 2 inches long, adjacent to Welds A and B on the ID of the 36-inch diameter completed pipe spool on sketch No. E-4269-9.
Connex informed the inspector that these areas were ground to remove mechanically caused impressions _ _ _ - - _ _ _ _ _ _ - _
a and raised metal'resulting fromLthe use.'offa-weld joint fit up'devico. LThe_NRC-inspector requested 1information only; ultrasonic thickness. measurements of.three ground larean adjacent 1to Weld B and' determined that approximately 0.010-inches of base material had been removed in two areas and 0.100 inches removed-in the' third area.- The-nominal vall-thickness, 0.375-inches, was present in all three areas measured with the_ lowest. wall thickness measured being 0.409 inches.
Connex also informed the NRC inspector that these ground areas were not liquid penetrant examined because-Procedure No. 3974 SP-1, Paragraph 2.5, only requires liquid-penetrant. examination following the removal of a defect by grinding, and these depressions are not_ considered defects..
3.
The NRC inspector performed an examination of the outside diameter of Weld B on Sketch No. E-4269-9'and questioned the-acceptance of the amount of weld reinforcement on the outside diameter.
This. examination was independent of any-acceptance inspection of the weld performed by.Connex.
Measurements from the toe of the veld reinforcement on tho' elbow side of Weld B locally exceeded the maximum weld reinforcement of 1/16 inch required by Section 5.10 of Procedure No. E-3974 SP-1 by approximately-1/32 inch.
Connex engineering and quality control personnel indicated that their acceptance of the weld reinforcement was based.on-measuring the height of the reinforcement from the pipe or elbow to the top of the reinforcement excluding any shrinkage in the toe: area of the weld.
The.NRC inspector and Connex discussed the requirements of Section-5.10 for the 1/16 inch maximum reinforcement for piping with a wall thickness over 3/16 to 1/2 inch inclusive and determined that the 1/16 inch value is more restrictive than the ASME Section III Code for this thickness range.
Connex indicated i.
that they would bring this to the attention of PSE&G.
3.7 Observation of NUPIC Audit Activities During the inspection at Connex, the NRC inspector observed portions of a Heclear Utilities. Procurement Issues Committee (NUPIC) audit conducted at Connex on' October 5 through 9, 1992.
The NUPIC audit's scope included reviews, inspections, and observations to determine if thefrequirements of Appendix B to 10 CFR Part 50 were being effectively implemented by Connex.
The NUPIC audit team consisted of four members representing three utilities.
The NRC inspector attended the.NUPIC entrance and exit meetings, and observed the NUPIC auditors reviewing. procurement.
documentation, witnessing hydrostatic testing and inspection activities, and interfacing with Connex.
For the most part, the NUPIC audit was performance based.
The NRC inspector observed' that the NUPIC auditors appeared to verify Connex was effectively,
implementing.their QA program mainly by observing Connex perform; in-process activities, and by reviewing completed quality records.
and associated items.- However, it is_noted that the NUPIC~
auditors did_ perform selected overview inspections such as takingl dimensions-on pipe fittings, checking micrometers for accuracy against calibration blocks, and witnessing the calibration of
_ selected instruments and gauges.
The NUPIC audit checklist was very thorough with respect to addressing the attributes of Appendix B to 10 CFR-Part 50, however it was generic in nature.
For example, there were no audit checklist _ items that addressed essential _ attributes to'be verified to ensure that piping subassembly fabrication activities were correctly being performed.
Subsequent to the inspection, the NRC inspector was informed that the NUPIC audit procedure-requires that, in addition to the checklist items, the technical specialist identify specific areas to be reviewed during the audit.
4 PERSONNEL CONTACTED Connex Pine Systems. Inc.
+*
Gil Gardner, Chief Executive Offitor-
+*
Donald J.
Leininger, Vice President Technical Services
+*'
James H. Seddon, Vice President Estimating
+*
James L.
Sweeney, Manager of Projects
+*
Kevin A. Welch, Manager of Quality Assurance
+*
Nelson R. Brown, Project Manager
+*
Ralph E. Nodes, Chief Inspector
+*
Thomas R. Smith, Shop Superintendent Lawrence M.
Evans, Assistant Purchasing Manager-Jerry Burris, Quality Control Inspector Randall L.
Davis, Chief Metallurgist Ron Goddard, Shop Foreman Mark A. Huck, Quality Assurance Technician Daniel L.
Schaffer, Quality Assurance Technician Dough L. Walker, Quality Control Inspector Public Service Electric and Gas Company Robert Bell, Source Inspector
~+ Attended-the entrance meeting on October 5, 1992
- Attended the exit meeting on October 9, 1992
-g
. a r
,-;9 )
w>
- ~.,
5 g.
Nuclear Utilities' Procurement" Issues: Committee Audit-Teami
-vte t il g
m.:,;
s George Tillsky,-Audit: Team Leader,=
- W u.
Cleveland Electric-Illuminating;-Company--
^] '1
- Wayne-Blazek,JAuditor,;.
Public Servico Electric and< Gas Company; Steve;Koeing,; Auditor, Wolf = Creek: Nuclear' operating Corporation
~
Richard Watson,: Technical-Specialisti Cleveland: Electric. Illuminating l Company _.
y
-4 k
1
~
4 4
=
- t 1
' I r..3
,j
foia,4 af N
umTED STATES
'9 %
/
NUCLE AR REGULATORY COMMISSION i M#-I wasmoton.o c.rosss
., ? f,l NOV ', 3 1932 Docket No.:
99900366/92-01 Mr. A. Schnaidt, Vice President Manufacturing Operations Exide Electronics 3301 Spring forest Road Raleigh, North Carolina 27604 Dear Mr. Schnaidt-
SUBJECT:
NOTICE OF VIOLATION (NRC INSPECTION REPORT NO. 99900366/92-01)
This letter addresses the inspection of your facility at Raleigh, North-Carolina, conducted by K. R. Naidu, J. G. Ibarra, and R. K..frahm, Jr...of this office on September 14-17, 1992, and the discussions of their findings with members of your staff at the conclusion of the inspection.
Areas examined during the inspection and our findings are discussed in the enclosed inspection report. The inspection team evaluated the circumstances surrounding the loss of five Exide uninterruptible power supply units-following a main transformer failure as described in NUREG-1455, " Transformer-failure and Common-Hode loss of Instrument Power at Nine Mile Point Unit 2 on August 13, 1991." The inspection team also reviewed the-implementation of your quality assurance program in selected areas.
Based on the results of this inspection, certain of your activities appeared to be in violation of NRC requirements, as specified in the enclosed Notice of Violation (Notice),
it was found that Exide's procedurcs did not address the requirements of Section 21,21, " Notification," of Title 10 of the Code of Federd Reaulations (10 CFk) as revi. sed on July 31,-1991, and effective on October 29, 1991. These requirements pertain to the 60-day period for evaluating possible defects and failures or filing an interim report.
Your staff informed our inspection team that, in 1991, you discontinued-implementing your quality assurance program meeting the requirements of Appendix B to 10 CFR Part 50, and that you discontinued to accept the
- reporting requirements of 10 CFR Part-21. Even though you no longer accept the requirements of Part 21, you still retain Part 21 responsibility for the inverters and components previously supplied in accordance with that regulation.
You are required to respond to this letter and should follow the instructions specificd in the enclosed Notice when preparing your response.
In your respense, you should document the specific actions taken and any additional act 'ans you plan to prevent recurrence.
Mr."A. Schnaldt -
In addition, NUREG-l_455 identified a discrepancy between the label on a front panel alarm indicator light of the Nine Mile Point-UPS and its _ identification in the associated equipment manual.
In a telephone conversation on November 20, 1992, Mr. Paul Steinman agreed to address this-discrepancy with Exide's response to the Notice of Violation.
~
The responses directed by this letter and the enclosed Notice are not subject-to the clearance procedures of the Office of Management and Budget as: required by the Paperwork Reduction Act of 1980, Public Law No. 96-51).
In accordance with 10 CFR 2.790 of the NRC's ' Rules of Practice,'. a copy of this letter and-the enclosed inspection report will be placed in the'NRC Public Document Room.
If you have any questions about this inspection, we_will be pleased to discuss them with you.
Sincerely, h4h Leif N
holm, Chief Vendor Inspection Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation
Enclosures:
1.
Inspection Report 99900366/92-01 cc:
Mr. B. Saylor Exide Corporation 3301 Spring forest Road Raleigh, North Carolina 27604
. 1
s i
1
! ENCLOSURE 1 2
' NOTICE OF-VIOLATION:
Exide"Electroniis).
Docket No.:-99900366/92-01;:
'Raleigh; North Carolina?
During an" inspection conducted-on September 14-17,;1992,9aiviolationofNRC.
- requirements-was i_dentified.: -In'accordance with the " General Statement ofi Policy and Procedure for NRC Enforcement Actions,":10 CFRlPart_2,EAppendix C,1 the. violation isilisted below:
Section 21.21, " Notification ~ of fa'ilure to compl' l or'existe'nce of defect?
y and'its evaluat' ion,"' of.10 CFR requires,51n part,- that each cor) oration.
subject to tha regulations adopt': appropriate procedures-for eitler.
evaluating deviations and failures. to: comply, or! informing 'the;11_censee?
or purchaser of:the deviation orifailure to comply.
Contrary to the'above,'Las of September 14,E1992 Exide had not revisedL its procedures, required by'10~CFR 21'.21, to address thelsubstantive.
revisions-to 10-CFR Part-21 that became' effective on October 29, 1991-(99900366/92-01-01).
This is a Severity Level _ V_ violation (Supplement VII).-
Pursuant to the provisions of 10lCFR-2,201, Exide;is herebyLrequired cto_ submit.
a written statement: or explanation _ to the U.S. -Nuclear Regulatory Commission, -
ATTN: Document Control: Desk, Washington,- D.C. 20555. with a copy tto the Chief,
. t Vendor. Inspection Branch -- Division _of Reactor Inspection and Licensee'..
-Performance, Office:of Nuclear Reactor Regulation,;within 30 days of;the date of the letter transmitting this Notice'~of Violation. This. reply should'be; clearly marked as a " Reply. to Notice of Violation" and;shculd include-for each.
violation '(1) the reason for;the: violation,-.or if contested, the basis. for _
H disputing the violation;-(2)1the corrective: steps that have been,taken ;and.the
-results achieved;:(3) the corrective steps that will be taken to avoid further:
-violations; and-(4) the date when full-compliance will be achieved. Where'-
~
good cause~is shown, consideration will be given to extendingithe response?
~
- time.
Date at Rockville, Maryland this 2'5 day of Ahd,1992 '
l' -
y y
ENCLOSURE 2--
U. S. NUCLEAR REGULATORY COMMISSION OFFICE 0F NUCLEAR REACTOR REGULATION DIVISION OF REAC10R INSPECTION AND' LICENSEE-PERFORMANCE-ORGANIZATION: -
Exide Electronics-Raleigh, North Carolina REPORT N0,:
99900366/92-01 ORGANIZATIONAL Mr. B. Saylor, Director CONTACT:
. Quality and Reliability (919) 878-2378 CORRESPONDENCE 3301 Spring Forest Road ADDRESS:
Raleigh, North Carolina. 27604 NUCLEAR INDUSTRY Manufactures electrical inverters, ACTIVITY:
uninterruptible power' supply (UPS) units,-
and replacement components.
INSPECTION
-September 14-17, 1992 l-CONDUCTED:
//!)
LEAD INSPECTOR:
f.s r K. R. Naidu, Team Leader-Date:
Reactive Inspection Section No._ 2 (RIS-2)
Vendor Inspection Branch (VIB) i OTHER INSPECTORS:
Ronald K. Frahm, Jr., VIB Jose G. Ibarra, AE00/ROAB APPROVED BY:
(MA
///5'NL--
G. C. Cwalina, Chief,-RIS-2, VIB Da'te Division of Reactor Inspection-and Licensee Performance Office of Nuclear Reactor Regulation INSPECTION BASES:
10 CFR Part 21 and Exide's quality assurance-program i
INSPECTION SCOPE:
The NRC inspection taam evaluated the circumstances surrounding the loss of five Exide i
UPS units following a main transformer failure at the Nine Mile Point Nuclear Station Unit 2.
PLANT SITE All plants using Exide inverters and q
APPLICABILITY:
UPS units.
'1-INSPECTION
SUMMARY
-1.1-11olationa
' Contrary-to the requirements of Part 21 of Title 10 of. the Code of Federal
'Reaulations.(10 CFR Part.21), Exide Electronics (Exide)' Procedure QCP 195, July 1,1985, had not been revised to implement the requirements of.the revision of Section 21.21(a) of 10 CFR Part El effective October 29, 1991.
[ Violation 99900366/92-01-01, see Section 3.5 of this report).
1.2 Nonconformances None identified during this inspection.
1.3 Unresolved items None identified during this inspection.
2 STATUS OF PREVIOUS. INSPECTION FINDINGS 2.1 Deviation 99900366/79-01-01 (Closed)
Deviation 79-01-01 stated, in part, that contrary to Criterion VI of Appendix B'to 10 CFR'Part 50, Exide had not issued a change notice (CN)'for-changes _made to Drawing No. 110603071, Revision A, August-25, 1978. 'Section 3.3 of the current Quality Control Procedure (QCP) 229,-Revision A,_ of January 25,1991, " Quality Assurance Internal Engineering Change Notice' Procedure,"
requires that each drawing that has been marked in accordance with a'" signed-off ECN" will be stamped " Red Line Drawing." The Internal ECN clerk or.
designer will write his or her initials on the stamp mark, and will stamp "NOT -
VALID AFTER" on the drawing.
2.2 Deviation 99900366/79-01-02 (Closed)
Deviation 79-01-02 stated, in part,.that contrary.to Criter' ion XV of Appendix B 'to 10 CFR Part 50, rejected material with an attached " Rejected Materiai" tag was designated as " Accept" (use-as-is) ~ without explicit approval from Engineering.. The current QCP 232, _-" Material. Review Board- (MRB)
Procedure," requires that a =" Rejected Material Tag"-(RMT) be completed and attached to nonconforming material. The nonconforming material will be placed -
in a segregated area pending NRB disposition. Each of the MRB board members, including a representative from Engineering, must be present during the weekly meetings to determine the disposition of such material and. complete items 21 through 29 of the'RMT.
" Accepted" items are so documented on the RMT form indicating approval by the MRB.
3 INSPECTION FINDINGS AND OTHER C0t91ENTS 3.1 Entrante and Exit-Meetinas During the NRC. entrance meeting on September 14, 1992, the inspectors' stated
-j the scope of-the inspection to the Exide staff. During the_ exit meeting on -
l September 17, 1992, the team leader summarized to the Exide management and-
-staff the team's findings.
3.2 Principle of Uninterruptible Power Sucolv (UPS) peeration i
The purpose of the UPS is.to provide continuous uninterruptible power _ to important electrical loads should the UPS lose-its normal ac input power from the in-plant electrical distribution system. This normal ac power supplies input to the ac to de rectifier, which provides regulated de power to the inverter.
If the normal ac power is unavailable to supply the ac to de rectifier, the alternate de power source to the inverter is supplied by an on-site storage battery.
The de to ac inverter provides a reliable, high quality ac output power source from the UPS to the critical bus. An alternate source of UPS power output is an off-site maintenance supply, but_ it is not of equivalent quality to the inverter output.
In earlier vintage nuclear power plants, rectifiers and inverters were installed in Teparate cubicles to provide uninterruptible ac power.
Subsequent designs combined a rectifier and inverter in a single UPS cubicle.
The switch that transfers ac power from the UPS to the alternate source is -
located inside the cabinet and necessitates safety precautions to be taken while performing maintenance to the UPS to prevent injuries to personnel.
Failure reports indicate that the longevity of components inside the UPS units may shorten when operated at abnormal operating conditions.
For instance, the condition of power supply batteries and lubricants (grease in circuit' breakers) installed inside the UPS rapidly degrade at elevated temperatures and voltages.
l 3.3 Backaround Information Under various names, Exide Electronics, as the corporation is currently known, designed and manufactured battery chargers, rectifiers, inverters, and UPS units and supplied them to the nuclear industry.
Exide personnel described the manner in which Exide, under the various names of its predecessors, has developed and improved inverter and UPS designs. Exide-informed the team that if a customer requests a replacement product for a component that had been manufactured by one of Exide's predecessors, Exide makes a conscientious attempt to locate the necessary drawings to assure the replacement is equivalent to the original.
For example, when a customer recently requested Exide to supply a replacement printed circuit board (PCB) for a battery charger, Exide conducted a search, located the drawings, and manufactured the PCB even though it had been supplied by its predecessor.
Exide has changed owners several times since producing the battery chargers and the first inverter. Exide personnel stated that the corporation began as the Electric Battery Storage Company and later became International Nickel of _
.Canadi. The company th:n became the privately owned:Exide Electronics and
_ most recently the publicly owned Exide Electronics.
Records indicate _that Exide supplied inverter Models:(1) _120/25F1,---
(2) 120/9.3F1, and (3)_-240/30F1, and UPS Models- (4)- 2730_ and (5) 575-60T3-120-a 208.to various nu: lear power plants.
Exide personnel stated that items (1),
(2),. and (3) weretof a common design, which included silicon-controlled -
rectifiers-(SCRs),ca-two step' inverter (dc-ac),-and'quasiwave,:and did not include an AC-DC rectifier. The corporation,_ then known as Electric _ Battery 1
Storage Company, first made available the; inverter with these components in-1962. Records indicate that these types of. inverters were installed at the Oconee Nuclear Station.
In the nomenclature of the models, the first number denotes the output ac voltage, the number after the slash denotes the_ output in KVA, and "Fl" signifies that it-is.a single phase unit.
In 1982, Exide introduced item'(4) as a 2000 series UPE which' included transistor-transistor logic (TTL). powered by either rectified de or D-cell batteries in the same configuration-as the one installed at the Nine Mile Point Nuclear Station Unit 2 (NMP-2). However, the rectified ac power from; the inverter output is preferred. Exide discontinued this single phase
,roduct in 1986.
In 1972,.Exide introduced item (5)- as a 3000-. series UPS, and
)
continues to produce them with-an improved logis design.'
In 1985, Exide enhanced the design by eliminating the K5 relay, which was used to select an -
~
alternate power (maintenance power) to the load when the inverter output power was lost, adding dual dc logic-power supplies and adding _an auctioneered configuration to select the source of input power to the logic. - Exide personnel stated that they' periodically inform customers of design enhancements by conducting presentatiens when called upon to perform service on their equipment or by distributing sales literature.
3.4_
D.yality Assurance (OAi Proaram Exide Electronics used a documented Quality Assurance Procedures Manual (QAPM), Revision A, November 12, 1990, which meets the criteria specified in Appendix B to 10 CFR Part 50 to manufacture and supply safety-related power supplies. However, Exide personnel stated that they discontinued implementing the QA program in 1991 and are revising the manual.to meet the requirements of International Organization-for Standardization (150) 9000 for QA program.s.
Exide expects to isst.a revised manuals reflecting the requirements for ISO -
9000 by the end of October 1992.
In a letter of February 24, 1992, Exide informed its nuclear customers that it would no longer accept any' purchase L orders (P0s) or provide any certificates of conformance (C0Cs) referencing conformance to Appendix B to 10 CFR Part 50. The C0Cs will reference only commercial grade requirements.
In the letter, Exide stated that it will continue to give technical and engineering support to all previous-orders.
referencing Appendix B to 10 CFR Part 50.
Exide further stated that current holders of controlled copies of its QAPM will receive revised manuals reflecting its change in QA policy.
The inspection team verified Exide's compliance to selected elements of the existing QA program as described in the following sections. l,
i
3.5-Reviets of 10 CFR Part 21 ProarjLn)
In'1991 Exide had discontinued implementing the QA program which had met the requirements of Appendix B to 10 CFR Part 50. The inspectors reviewed Quality _
Control Procedure (QCP) 195 "10 CFR Part 21 Compliance," July 1,1985,: which addresses _ the implementation of 10 CFR Part-21. The inspection team-found one violation.of 10 CFR Part 21-in this area. The _QCP does not-reflect the requirements of the revision to 10 CFR Part 21 of July 31.-1991 which became-effective October 29, 1991.
The QCP.also does not include a definition for either a deviation or a defect. The NRC inspection team emphasized to Exide that it still retains 10 CFR Part 21 responsibilities to report defects and noncompliances for those ite'us Exide had supplied in accordance with that regul ation. The QA Director stated that he would revise QCP 195 to reflect the current revision of 10 CFR Part 21, 3.6 fleview of Records for UPS Units installed at Nuclear Power Plants
-Exide supplied a list of the various UPS models supplied _ to the nuclear power plants, but was not certain of its accuracy. Records at the NRC reflecting the failures of UPS units indicate that Exide may have supplied safety-related-UPS units to at least the Oconee, NMP-2, and Maine Yankee nuclear power plants. Exide personnel ststed that they may not have accurate information on the actual users of their UPS units because the purchasars, such as architect engineers and constructors, did not state either where tht. UPS units were to be used or if they were installed in safety-related applications.
Furthermore, construction personnel at the site frequently installed and started the UPS units. Therefore, Exide could not maintain accurate records on the various types of UPS units installed at the nuclear power plants.
If a customer awarded a contract for Exide to service its UPS units, Exide field personnel dispatched to the site would exchange technical information and advise the customer of the latest improvements in the UPS design. The inspectors concluded that Exide did not segregate or otherwise track nuclear orders to the extent necessary to trace the different designs that were installed in various nuclear power plants.
Exide personnel were, however, able to provide sufficient information to the inspectors to evaluate the specific problem at NMP-2 as discussed below.
3.7 Review of Records Related to the UPS Units That Failed at NMP-2 On August 13,1991, an internal failure in the main transformer at NMP.-2 caused five, 3-phase, 75 kVA, UPS units to fail. These UPS units were installed in nonsafety-related applications. The NRC investigated the incident and documented its findings in NUREG-1455, " Transformer Failure and Common-Mode Loss of Instrument Power at Nine Mile Point Unit 2 on August 13, 1991." The team reviewed the records available at Exide related to the UPS unus and determined the following.
The Stone and Webster Company, the architect-engineer for NMP-2, issued purchase order (PO) NMP2-E035A of January 6, 1978, to Exide Power Systems Division to furnish and deliver UPS units in accordance with Specification NMP2-E035A, April 6, 1977. The P0 required Exide to ship the complete units to NMP-2. Records indicate that Exide shipped the UPS units to NMP-2 on Occember 5, 1979.
Field Report (fR) N40762, April 2, 1984, stated that Exide personnel performed startup services on January 30 and February 4, 1984 at NMP-2. During the startup, Exide field personnel replaced the control batteries, fuses,-and logic boards.
Exide stated that all tests had been completed and witnessed i
ar,d that all UPS units except No. 5 were operational.
In 1985 Exide improved the d6 sign of the 3000 series UPS by eliminating the I
KS relay, which was used to select an alternate power (maintenance po"er) to the load when the inverter output power was lost, and adding dual dc. logic power supplies.
It also added an auctioneered configuration to select the source of input power to the logic.
Exide stated that it had informed all its customers of the design enhancement in a 1985 letter but could not furnish the team with a copy of the letter. NMP-2 did not purchase this design i
enhancement.
In 1987, NMP-2 requested Exide to furnish re)lacement technical manuals for UPS units purchased in 1979.
Exide stated t1at since it had not retained copies of the NMP-2 manuals, it had to compile an instruction manual and send it to NMP-2.
In this manual, Exide included a design change that had been incorporated in 1987 to increase the reliability of the UPS ur.its, in the new design, the inverter power (instead of the maintenance power) was the preferred power to the logic. Apparently, the NMP-2 site >ersonnel did not recognize that, contrary to the arrangement described in tie new manual, the.
power to the logic in their UPS was connected to the maintenance power supply and not the inverter. This discrepancy remained undetected untti the post event evaluations and the NRC's incident investigation Team (111) verified the configuration af ter the event of August 13, 1991, as discussed in NUREG-1455.
NRC Information Notice (!N) 91-64 " Site Area Emergency Resulting from a loss of Non-Class lE Uninterruptible Power Supplies" was issued to alert licensees of potentially generic problems resulting from the common mode failur1 of the-UPS units.
Supplement I to IN 91-68 was later issued to provide further details on the replacement intervals for the control logic power supply batteries. The inspection team concluded that the consequences f the August 13, 1991, incident at NMP-2 would have been less severe if dere had been better communications between Exide personnel and NMP-2 personnel.
3.8 Review of Desian Documentation Chance Control Since the llT identified discrepancies and concerns with changes to the equipment manuals, the team reviewed and verified the implementation of Exide's design documentation change control program.
Exide has an engineering change process (Engineered Change Notice Procedure, RMG007) implemented by an engineering change notice (ECN) to document changes to controlled engineering documents and drawings.
Exide makes engineering changes to its assembly documents and engineering documents to solve design problems and incorporate.
design enhancements. The Corrective Action Request (CAR, QCP 177) defines.the system for continuously correcting the designs. __
..____._.._.a.,
Exide field engineers can modify the design as they find need.
In such instances of a field request, Exide initiates an ECN after receiving the field request to implement the modification. The customer can also submit a special feature request (SfR) to request a design change, which would enter the formal process through the ECN process.
The field request is utilized to implement the approved ECN at the site. Another method to change the design is by issuing a field service bulletin (fSB).
The field request would reference the FSB when field engineers perform the modification in the field. After installation, the field engineers submit the field request to complete the design change process to assure that the as-built configuration of the UPS units will match the final documentation package. The team concluded that Exide had adequate controls over their design' change process.
3.9 Qbservation of Receivina inspection Activities The team inspected the aren where Exide personnel received and inspected incoming material.
The team observed that "Old Drawings" tags had been I
attached to some material to indicate that the material had been inspected and determined to conform to an earlier revision of the drawings.
Exide will continue to use this material until it is depleted.
In a separate area, material with "New Drawings" tags had been stored to indicate that the-material had been inspected and determined to conform with revised drawings.
t Exide stored in a segregated arca any material that was inspected and determined to be unacceptable. The team determined that the identification of the material-in the receipt inspection area conformed with Exide's Procedure, QCp 190, Revision E, of August 24, 1990, " Receipt inspection."
3.10 [prrective Action Reauest Process The team reviewed the corrective action request (CAR) process used by Exide to find problems and provide solutions. Exide described its CAR system in Quality Control Procedure (QCP) 177, " Corrective Action Procedure, "
l Revision D, february 17, 1992. The team observed that the procedure did not emphasize that corrective actions should be adequate to preclude repetition-in the future.
Exide uses a computerized, paperless system to control its CAR activities. The originator of a CAR assigns a priority level defined by QCP 177 as follows: "To) - A problem causing a product line to be shut down; Accelerated - A pro)lem causing inefficiencies, a detriment to the company and the product; Convenience - A problem causing an inconvenience." In paragraph S.4.1 of QCP 177, Exide stated that " CARS will be monitored against the CAR target date on a weekly basis."
In Paragraph 5.4.2, Exide stated that
" monitoring will be done by way of a weekly report which will identify CARS past their target date and those CARS that have been opened which have not.
been assigned a target date."
In reviewing existing CARS, the NRC inspectors observed numerous CARS that lacked a target completion date and remained open for several months. Many of these delinquent CARS are coded with a priority level of " Top" and " Accelerated." Exide informed the inspection tus that a previous auditor had identified this same concern.during an unrelated audit in-August 1992. The NRC inspectors reviewed the response to the auditor in which Exide committed to review each open CAR by October 12, 1992 and assign it to an individual with a target completion date.
The inspectors found Exide's' actions acceptable. '
3.11 Mdntenante Recomendation for UPS Units Exide performs maintenance on UPS units only upon receiving a service contract. On such occasions, Exide personnel brief their customers on technical enhancements made on specific models to improve the reliability of the UPS units.
Exide personnel stated that they observed that maintenance cannot be readily performed on the UPS units installed at NMP-2 because the units can not be isolated and removed from service and still provide a means to power the loads. The UPS units were designed and installed with no provision to service the UPS without interrupting power to its load.
Exide recommends that customers perform as a minimum, the following annual maintenance on the UPS units:
Shut down the unit and inspect the physical condition of all components and connections.
Determine if any electrolytic capacitors leak.
Operate the UPS with low load or no load and check the logic functions.
Verify that no changes were made to the overload settings which jeopardize the reliability of the UPS.
Verify the wave shapes wherever applicable.
+
Vary the load to verify all the corresponding settings respond appropriately.
Check all the trip functions, fuse failures, and overload conditions.
Verify that the transfer mechanism is operable.
Exide further recommends customers perform minor maintenance on their UPS units at a frequency depending on the environmental conditions in which they operate. Technicians should check and replace the air filters if they are dirty, check for discolored components, such as resistors or capacitors (to indicate that they are overheating), and verify that the D-cell batteries function properly.
3.12 Review of Recent Nuclear licensee Audits of Exide The NRC inspectors reviewed recent QA audits performed by owners of nuclear power plants and their designees.
In March of 1991, the Philadelphia Electric Company (PECO) conducted an audit to determine the implementation of Exide's QA program, which had been written to meet the requirements of Appendix B to 10 CfR Part 50.
PECO used a checklist that had been prepared with the guidelines established by the Nuclear Utilities Procurement Issues Committee (NUPIC).
PEC0 did not identify any unacceptable findings and concluded that Exide was satisfactorily implementing its QAPH, 10th Edition, November 12, _ _ _ _ _ _ _ _ - - _ - _ _ _ _ - _ - _ _ _ _ _ - _ _ _ _ - _ _ _ _ - _ _ _ - _ _ _ _ _ _ - _ _
1990. The team did not review the NUP!C audit checklist because it was not avalloble in the audit folder. Exide stated PECO did not send it a copy of this checklist.
b in May 1988, the Baltimore Gas and Electric Company (BG&E) performed a QA audit in which it made five findings.
In August 1988 Exide submitted a-i response to those findings, in which it discussed its proposed corrective actions.
Exide's audit file did not contain BG&E's acceptance of the proposed corrective actions.
In February of 1991 BG&E scheduled an audit but canceled it before the commencement date. The NRC inspectors did not review the adequacy of Exide's proposed corrective actions to resolve the findings from-BG&E's 1988 audit. The only concern identified by the team in this area was that the audit files did not contain all of the necessary documentation to detail the audit findings and their corrective actions.
4 PERSONNEL CONTAC1ED
[xide Electronics A. Schnaidt, Vice President, Manufacturing
- +
W. Raddi, Senior Vice President K. Agee, Director, Power Systems Group-
+
R. Brewer, Manager, Product Support Engineering V. Freeman Technical Support Supervisor
- +
D. J. liess, Director, Customer Support M. E. Grady, Manager, field Service Administration
+
G. D Jarvis, Manager, Project Engineering
- +
J. C. Lovelace, Manager, Quality Control and Audit R. A. Machilik, Director, Power Systems Group
- +
B. Saylor, Director, Quality Assurance
- +
P. Steinman, Lead Auditor Attended the entrance meeting on September 14, 1992.
+
Attended the exit meeting on September 17 1992, i 4
=. -
y y
y ' t, p'
UNITED STATES NUCLEAR HECULATORY COMMISSION s. %,p[/[
WAsHINo1oN, D.C. 20WA o
December 8, 1992 Docket No. 50-331 lowa Electric Light and Power Company ATIN: Mr. Lee Liu Chairman of the Board and Chief Executive Officer IE Towers p.0. Box 351 Cedar Rapids, IA 52406
Dear Mr. Liu:
SUBJECT:
IN5pECTION OF THE PROCUREMENT AND COMMERCIAL GRADE DEDICATION PROGRAMS AT THE DUANE ARNOLD ENERGY CEN1ER (REPORT N0.
50-331/92-201)
This letter transmits the report of the inspection conducted May 11 through 15. 1992, at Iowa Electric Light and Power (IELP) Company's Duane Arnold Energy Center (DAEC) by R. Pettis, S. Alexander, L. Campbell, and S. Magruder of the Nuclear Regulatory Commission's (NRC's) Vendor Inspection Branch (VIB) and R. Langstaff of NRC Region 111. The inspection was related to activities at the plant site authorized by NRC license DPR-49. At the conclusion of the inspection, we discussed our findings with Mr. J. Franz, Vice President, and the other members of your staff identified in Section 5 of the enclosed inspection report.
The inspection was,enducted to review the implementation of IELP's program for the procurement and dedication of commercial grade items (CGis) used in safety-related applications at DAEC. The. inspection team identified a major.
program strength in that IELP's policy was to purchase safety-related spare and replacement items from original equipment manufacturers with approved 10 CFR Part 50, Appendix B, quality assurance programs when available, thus reducing the number of items that need to be purchased commercial grade and dedicated for use in safety-related applications. Another strength identified was IELP's receipt inspection and testing capabilities for performing material verification which are performed by independent quality control inspectors qualified to ANSI N45.2.6.
The results of the inspection also indicate that some weaknesses existed within IELP's program for the dedication of CGis procured for use in safety-related applications. These weaknesses contributed to the specific findings that were identified durir.g the review of dedication flies for CG!s installed or available for installation in safety-related plant systems. Examples of these findings are that IELP did not fully verify the material of certain CGis,-did not have provisions to fully assure material traceability and lot homogeneity when using statistical sampling plans, did not properly classify l
Mr. Lee Liu ;ertain valve and pump parts when using a generic classification approach, and did not assure the calibration and control of certain test equipment used in the dedication process.
It should be noted that the inspection team selected for review equipment from DAEC's safety-related component list and made no further attempts to assess individual component contribution to the system's safety-related function.
This was the fourth of a series of five pilot inspections being conducted to evaluate the implementation of licensee CGI procurement and dedication programs and to finalize an inspection procedure for future inspections of this type. The other four inspections also identified certain weaknesses in licensee CGI dedication programs and their implementation. The first three inspections classified these weaknesses as deficiencies with the potential for subsequert enforcement actions by the appropriate NRC regional office. During this period, the NP.C also received significant feedback from several utilities and from the Nuclear Management and Resources Council questioning the regulatory basis for the CGI dedication guidance contained in NRC Generic Letter 91-05 as well as specific interpretations of that guidance by the NRC inspection teams.
In order for the NRC to properly respond to these concerns and to resolve major points of contention concerning the CGI dedication process we will be holding public meetings with individual utilities and conducting a public workshop which will address issues concerning the existing dedication guidance and specific interpretations of that guidance.
Prior to the public workshop, the proposed inspection procedure for future inspections will be issued for public comment and discussion at the workshop.
in view of these pending actions, the specific observations concerning your CGI dedication programs and implementation are classified as " findings."
Because of the need for additional discussion of the appropriate guidance for dedication of CGis, no enforcement action will be taken based on any of the five pilot inspections.
It should be recognized, however, that some of these findings may require corrective action based on: (1) their potential impact on the performance of specific equipment, or (2) because of noncompliance with your commitments to specific industry standards or regulatory guides endorsing such standards. We expect that you will evaluate the inspection findings in view of these considerations and take appropriate corrective action as necessary.
In accor.1ance with 10 CFR 2.790 of the NRC's " Rules of Practice," a copy of this letter and the enclosure will be placed in the NRC's Public Document Room.
s
( 1
Hr. Lee Liu Should you have any questions concerning this inspection, we will be pleased to discuss them with you. Thank you for your cooperation in this inspection.
Sincerely, Jack W. Roe, Director D'Ivision of Reactor Projects !!!, IV, V Office of Nuclear Reactor Regulation
Enclosure:
Inspection Report 50-331/92-20) cc w/ enclosure:
see next page l,,
. Mr. lee Liu lowa Electric Light and Power Company
-Duane Arnold Energy Center Jack Newman, Esquire Kathleen H. Shea, Esquire Newman and Holtzinger 1615 L Street. N.'4.
Washington, D.C.
20036 Chairman, Linn County Board of Supervisors Cedar Rapids, Iowa 52406 lowa Electric Light and Power Company ATIN: David L. Wilson Plant Superintendent, Nuclear 3277 DAEC Road Palo, Iowa 52324 Mr. John F. Franz, Jr.
Vice President, Nuclear Duane Arnold Energy Center 3277 DAEC Road Palo, Iowa 52324 Mr. Keith Young Manager, Nuclear Licensing Duane Arnold Energy Center 3277 DAEC Road Palo, Iowa 52324 U.S. Nuclear Regulatory Commission Resident inspector's Office Rural Route #1 Palo, Iowa 52324 Regional Administrator, Region 111 U.S. Nuclear Regulatory Commission 799 Roosevelt Road Glen Ellyn, Illinois 60137 Mr. Stephen N. Brown Utilities Division Iowa Department of Commerce Lucas Office Building, 5th Floor Des Moines, Iowa 50319 -.
U.S. NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION DIVISION OF REACTOR INSPECTION AND LICENSEE PERFORMANCE Report No.:
50-331/92-201 Docket No.:
50-331 DPR-49 License No.:
Licensee:
lowa Electric Light and Power Company lE Towers P.O. Box 351 Cedar Rapids, Iowa 52406 Facility Name:
Duane Arnold Energy Center Inspection Conducted:
May 11 through May 15, 1992 Inspection Team:
S. Alexander, EQ and Test Engineer, VIB S. Hagruder, Reactor Engineer, VIB L. Campbell, Reactor Engineer, VIB R. Langstaff, Reactor inspector, Rlli b
,O
~'
Prepared by:
- tobert L. Pettis, Jr., P.E.
Date Team Leader, VIB N
/
b Reviewed by:
A Leif Jf/fyfrholm, Chief Date t
VendoFirtspection Branch Division of Reactor inspection and Licensee Performance Office of Nuclear Reactor Regulation
[
t!fa [
/1/2/#7 7 rw Approved by:
'^
CEarles E. Rossi, Director
~
Dat4 Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulatio1
TABLE OF_ CONTENTS-DJtf!
EXECUTIVE
SUMMARY
1=
1 INTRODUCTION..............................................-
1 2
COMMERCI AL GRADE DEDICAT ION PROGRAM REVIEW.................
2 2.1 Procedures Review...................................-
2 2.2 Parts Classification................................
3 2.3 Trend i ng o f Suppl i e rs............................... -
74 2.4 Detection of Fraudul ent Materi al.....................
5 2.5 Receipt Inspection.................................._
_5 3
DEDICATION PACKAGE REVIEW FINDINGS........................
6.
4 PROCUREMENT AND DEDICATION TRAINING.......................
15-5 EXIT MEETING.............................................
16
+
_A
EXECUTIVE
SUMMARY
from May 11 through May 15, 1992, staff of the Nuclear Regulatory Comission's (NRC's) Vendor Inspection Branch (VIB) and the Region !!! office inspected lowa Electric Light and Power (IELP) Company's activities related to the pro-curement and dedication of comercial grade items (CGIs) used in safety-related applications at the Duane Arnold Energy Center (DAEC). The inspection team reviewed IELP's procurement and dedication program to assess the licensee's compliance with the quality assurance (QA) requirements of Appendix B to Part 50 of Title 10 of the Code of Federal Regulations (10 CFR Part 50 Appendix B).
On August 24, 1990, the NRC staff forwarded to the Comission SECY-90-304, NUMARC Initiatives on Procurement," in which the staff reported the status of the Nuclear Management and Resources Council's (NUMARC's) initiatives on general procurement practices.
Procurement initiatives as described in NUMARC 90-13, " Nuclear Procurement Program Improvements," October 1990, recomended that licensees assess their procurement programs and take specific action to strengthen inadequate programs. The initiative on the dedication of CGis, which was to be accomplished by January 1, 1990, stated that licensee programs should meet the intent of the guidance provided in the Electric Power Research Institute (EPRI) Final Report NP-5652, " Guideline for the Utilization of Commercial Grade items in Nuclear Safety Related Applications (NCIG-07),"
June 1988. The staff also stated in SECY-90-304 that it would conduct assessments at selected sites to review the licensees
- impicmentation of improved procurement and commercial grade dedication programs, assess improvements made in the areas covered by the NUMARC initiatives, and report the results of those assessments to the Comission.
From February through July 1991, the VIB conducted eight assessments of selected licensees to determine the current status of activities to improve the procurement program related to industry initiatives and NRC requirements. On September 16, 1991, the NRC staff forwarded to the Commission SECY-91-291, " Status of NRC's Procurement Assessments and Resumption of Programmatic Inspection Activity,"
in which the staff reported on the results of its assessments and noted that it was resuming inspection and enforcement activities.
The NRC conducted this inspection, the fourth of the headcuarters pilot inspections in this area, to review IELP's procurement anc dedication programs and their implementation since January 1, 1990 (the effective date of the NUMARC initiative on dedication of CGis). The staff focused its inspection on a review of procedures and representative records including approximately 30 procurement and dedication packages for mechanical and electrical items classified as safety-related by DAEC; interviews with IELP staff; and observations of IELP's activitin. The inspection team's findings were discussed with IELP's representatives and senior management at the exit meeting held on May 15, 1992. The inspection team identified several strengths in IELP's procurement program and its implementation. A major program strength was IELP's policy of procuring safety-related spare and replacement items from original equipment manufacturers (0EHs) with approved Appendix B QA programs when available. This policy reduces the number of items that need to be purchased commercial grade and dedicated for use in
-i- :
__-_m_
_________.m_-__-_.--_-_-_----_---
--.-.u--
safety-related applications. Another program strength identified was lELP's.
receipt inspection and testing capabilities for performing material l
verification which are performed by independent-quality control inspectors.
qualified to American National Standards Institute Standard N45.2.6.
1he inspection team also identified some weaknesses in the generic procurement program and in IELP's im)1ementation.
These weaknesses contributed to the specific findings descri)ed in Section 3 of the report.
Examples of these findings included DAEC's practice of not evaluating the material of certain CGis.
Instead,-DAEC relied upon a functional or performance test (o,g., proof load, hydrostatic test) to demonstrate that the item would perform its intended safety function (e.g., maintain pressure integrity).
Such tests may not be sufficient to demonstrate that the item is suitable to perform its-safety function under the full range of design-basis conditions or to-demonstrate reliability for the time between inspections or tests.
The team also found that in certain applications valve and puma parts were being generically classified as nonsafety-related despite~ t1eir potential for performing safety-related functions.
Other findings identified by the inspection team concerned the use of samp1'ing to demonstrate suitability of certain CGis, including greases,-lubricants, and -
oils used in safety-related and environmental qualification applications.
In general, the DAEC sampling plan was based on a Hil-STD-10SD approach whichf _.
assumes lot or batch control and reasonably homogeneous batches.
However,-the i
l documentation reviewed during the inspection did not support how DAEC had-l determined such supplier controls on product consistency for CGis purchased.
l Additionally, in the case of lubricants, traceability to the OEM was also not demonstrated.
The inspection team also identified several concerns related to' the conduct and application of commercial grade surveys and with the calibration and use of a spectrograahic analyzer during the dedication process.
It should be noted that t1e inspection team selected for review-equipment from DAEC's safety-related component list and made no further attempts to assess individual component contribution to.the system's safety-related function, 6 _ - _ -. - - -. - -, _ _
1 INTRODUCTION During this headquarters pilot inspection, the Nuclear Regulatory Commission (NRC) inspection team from the Vendor Inspection Branch (VIB) of the Division of Reactor Inspection and 1.icensee Performance of the Office of Nuclear React" Regulation reviewed the Iowa Electric Light and Power (IELP) Company's program and its implementatio'. for the procurement of commercial grade items (CGls) used tri safety-related applications in the Duane Arnold Energy Center (DAEC). The team also reviewed the IELP program and its implementation at DAEC for determination or verification of suitability of those CGis for their intended or approved safety-related applications, a process referred to as
" dedication."
Part 21 of Title 10 of the Code of Federal Regulations (10 CFR Part 21) defines dedication as the point at which an item or service becomes a " basic component," essentially, an item or service with a safety-related function.
However,10 CFR Part 21 also deffiies CGis (Section 21.3(a)(4)(a-1)), as distinguished from items procured as basic components. The regulation then allows the procurement of items that tre to become basic components, but that meet its definition of CGis, without invoking 10 CFR Part 21 in the procurement documents.
When CGis are procured for safety-ralr.ted service, their arocurement and dedication constitute activities at fecting quality and,11erefore, these activities must be controlled in accordance with the quality assurance (QA) requirements of Appendix B, " Quality Assurance Requirements for Nuclear Power Plants," to 10 CFR Part 50 (Appendix B).
In particular, Criterion 111
" Design Control," and Criterion Vil, " Control of Purchased Material, Equipment, and Services," of Appendix B are m:st pertinent to procurement and dedication of CGis; therefore, the inspectors reviewed the IELP program governing these activities and the implementation of that program at DAEC for compliance with these (primarily) and other applicable Appendix B criteria, as well as with the requirements of 10 CFR Part 21.
The NRC has provided further guidance to the requirements of Appendix B as they pertain to the procurement and dedication of CGls in NRC Generic letter (GL) 89-02, " Actions to improve the Detection of Counterfeit and fraudulently Marketed Products," on March 21,1989, and GL 91-05, " Licensee Commercial-Grade Procurement and Dedication Programs," on April 9, 1991. Therefore, the IELP CGI procurement and dedication trogi am and its implementation also were evaluated for consistency with the guidance and NRC staf f positions promulgated in these GLs.
Finally, with respect to procurement in general, including procurement and dedication of CG!s, IELP has committed to various industry standards and other publications (as endorsed or conditionally endorsed by NRC regulatory guides, NUREGs, and GLs); as stated in the IELP QA program description as contained in or referenced in the IELP Updated Final Safety Analysis Report for DAEC, and as expressed for the industry by the Nuclear Management and Resources Council (NUMARC) in the NUMARC initiative on tha dedication of CGIs as part of NUMARC 90-13, " Nuclear Procurement Progran Improvements." -
2 COMMERCIAL GRADE DEDICATION PROGRAM REVIEW 2.1 Erm d ures Review lhe IELP program for procurement and dedication of CGis for safety-related applications at DAEC is described and prescribed by a hierarchy of procedural l
documentation, beginning at the IELP Nuclear Generation Division (NCD) level of ILLP with the NGD procedures.
Subordinate to these are the three principal groups of procedures used by the organizations at DAEC that share primary responsibility for the various aspects of CGI procurement and dedication.
i Overall guidance is provided in tae first of these groups of procedures,-the NGDP 100-series procedures of the IELP NGD, The procedures generally describe the pertinent IELP policies and procedures governino DAEC-procurement and dedication activities. Detailed guidance in various procurement-related subjects is provided in the Engineering Department Manual 1200-series procedures, the QA implementing procedures (QAP 1100-series), the DAEC Administrative Manual, the DAEC Procurement Manual (PH 1400-series) procedures, and lower tier instructions and other directives.
The inspection team's review of the DAEC dedication program procedures identified some concerns. The definition of critical characteristics in paragraph 3.6 of DAEC Procedure 1204.14, " Dedication and Upgrade of Commercial Grade items," Revision 3 April 10, 1992, was not consistent with the working definition in Paragraph 6.2.14 and was not consistent with the guidance contained in GL 91-05.
Guidance on determining safety functions and failure modes was considered weak, but it improved in Revision 3 of 1204.14. Nevertheless, the procedures of the Engineering Department Manual governing the plant component safety classification Q-200 data sheet lacked guidance on documenting the technical basis for the classification, safety function, or credible failure mode determination; nor was there a provision on the Q-200 data sheet for such information.
If the CGI application happened to have a plant component identifier or " tag" number, a Q-200 data sheet would be provided, but not-a Classification of Subcomponents/ Materials (CSH) evaluation. The procedure j
concerning the technical evaluation of replacement items called for performing I
this evaluation when the CGI part number has changed and for other situations in which the replacement CGI is considered not to be identical, but it was not clear from the engineering procedures how like-for-like (identical) determinations, i.e., concluding that the replacement CGI is identical, were to be done.
The section of Quality Assurance Procedure (QAP) 1116.5 on commercial grade supplier surveys distinguished between critical characteristics for design (CCDs) and critical characteristics for acceptance (CCAs) as in Electric Power Research Institute (EPRI) NP-6406 (NCIG-ll),
in addition, in this procedure, IELP identified a subset of CCDs called " critical characteristics of the manufacturing process" (CCHs). This was a useful distinction, particularly for the commercial grade survey, source inspection, or source surveillance processes. However, this distinction was not reflected in engineering procedures, in addition, the CCA definition in Paragraph.3.1.3 of QAP 1116.5 as written (i.e., attributes providing reasonable assurance that the item received is the item specified) was not consistent with the guidance contained in GL 91-05; although, this was tempered somewhat by the requirement l.
--- -.A
in Paragraph 5.5 of QAP 1116.5 to include CCDs, CCMs, and CCAs in the survey plan checklist.
A strength was identified in the commercial grade survey procedure in that it.
contained the GL 89-02 restrictions on commercial grade surveys and also some additional guidance on handling the situation in which a distributor is involved. Another programmatic strength was identified in IELP's use of QA Form 1104.4-6a, " Checklist for Evaluating Third Party Audit Reports," which also covered commercial grade surveys, to ensure applicability of the audit or survey to the same part being procured or dedicated (or both) for DAEC as well as to DAECs application requirements. including addressing critical characteristics identified for the DAEC dedication.
Another strength identified by the inspection team was DAEC's policy to procure safety-related saare and replacement items from original equipment manufacturers (OEMs) wit 1 approved Appendix B QA programs, as stated in DAEC Procedure 1204.14.
The QAPs for source inspections and-surveillance called for Engineering.
approval of plans and required that results be sent to Engineering, but it was not clear from engineering procedures if or when these results would be reviewed for technical adequacy. This would be particularly necessary when the results are not clearly satisfactory or unsatisfactory.
In order to assess the effectiveness of the implementation of IELP's commercial grade survey (and source verification) program in support of dedication, the team reviewed a number of completed survey and source inspection reports associated with some of the individual dedication packages reviewed.
Surveys and source inspections evaluated are discussed in Section 3 of this report, in conjunction with the dis:ussion of.the associated j
individual dedication package.
2.2 Parts Classification The inspection team reviewed DAEC Procedure 1204.15, " Classification of Subcomponents/ Materials," Revision 1. November 12, 1991, and discussed the methodology for parts classification With DAEC' procurement and QA personnel.
The methodology and criteria used to determine safety classification of parts includes identifying and documenting information such as the following in a CSM document:
subcomponent/ material data, such as the model/part number, description, and manufacturer parent component data, suct as equipment identification, quality-
+
evaluation, specific plant application (description and function), and seismic and environmental qualification considerations subcomponent/ material credible failure modes and effects on plant e
operation T
y w.
-mw.YvT-T Su L--t-".**
- T*-
Ti a-'
4 e
r--
- A w
-um N'
M M
The inspection team considered the CSM process adequate for classifying
-subcomponents/ materials; however, the team noted that DAEC Procedure 1204.15 Revision 1, does not require the preparation of a CSM if the subcomponent/
material is assumed to have the same safety classification as the parent t
structure, system, or component. DAEC procurement personnel indicated that the intent of Paragraph 6.2.8. of DAEC Procedure !?04,14. Revision 3, was that a CSM would be prepared for each subcomponent/ material being dedicated that' was not generically classified in accordance with DAEC Engineering Guide i
DGC-P100, ' Design Guide Criteria for the Classification, Procurement and Application of Selected items,' Revision 1. April 10, 1992. Review of Procedure 1204.15, and Paragraph 6.2.8 of DAEC Procedure 1204.14, and discussions with DAEC procurement and QA sersonnel resulted in an agreement _to revise both procedures in order to meet tie stated intent..
t The inspection team performnd a limited review of the following generic classification documents included in Engineering Guideline DGC-P100, l
Revision 1:
ED-002, "0-Rings, Gaskets, Packing, and Thread Sealant,"
Revision 1, January 24, 1992; ED-004, " Lubricants and fuel Oils," Revision 1 April 6, 1992; and ED-008, " Pump and Valve Components,* Revision 1, January 2 1992.
t Af ter performing a limited review of ED-002 of Engineering Guidelir.e DGC-P100, the inspection team expressed concerns about the omission of certain 0-rin gaskets, and packing in containment' isolation valves forming part of the gs, primary containment system. DAEC personnel indicated that ED-002 would be' revised to identify such 0-rings, gaskets, and packing as being part of the containment boundary if they form part of the local leak rate test boundary.
ED-004 of Engineering Guideline DGC-P100 requires lubricant and fuel oil to be classified as safety-related except that lubricants which do not directly support the equipment's safety function are classified as nonsafety-related.
The inspection team considered this to be a reasonable approach for classifying lubricants.
Following the inspection team's review of ED-008 of e
Engineering Guideline DGC-P100, it was determined that in certain applications some pump and valve parts such as disc nut set pins, locking keys, and i
bearings were generically classified as nonsafety-related but may perform a safety-related function and, as such, would be considered basic components._
i DAEC indicated that ED-008 would be revised _to1 require that a CSM be prepared-to determine the classification of safety-related pump and_ valve parts, except that pump and valve o-rings, gaskets, and packing should be classified in r
accordance with ED-002.
2.3 Trendina qLSapplien At the time of the inspection, no formal program for determining trends of problems associated with suppliers had been established at DAEC.- Problems identified through either receipt inspection or dedication, which warranted initiation of-a procurement action request, wern considered during annual supplier evaluations performed per DAEC Procedure-1104.4, " Supplier Evaluation."
Since early 1992, for purposes of determining trends, IELP has
_ t 1
been collecting information on all safety-related and nonsafety-related problems identified through receipt inspection, liowever, IELP had not fully _
_..,..~.-. _ __ - _. _. _ _ _ _. _ _ _. _. _ _ _ -... _.. _ _ _.
developed the overall program and software for trending supplier quality.
IELP expected to have its trending pr) gram in place by August I, 1992.
2.4 htection of Fraudulent Miltrhl DAEC has established a program for detecting counterfeit and fraudulent material to ensure that such material is not installed in the plant. DAEC Procedure 1404.2, " Procedures for Receiving Materials, Parts and Components,"
Revis'.on 3C, April 11, 1992,- provides guidance for warehouse storekeepers to use when material is received at the plant. These warehouse employees perform the initial screening of material; discussions with them indicated that they were knowledgeable about what to check for. Checklists describing typically misrepresented items, and methods used to counterfeit them, are posted in the receiving.rea for the workers' reference.
DAEC Procedure 1105.1, " Acceptance of Material, items, Services, and.
Components," Revision 8, April 1,1992, gives detailed guidance to the QC inspectors who perform the receipt inspections on all safety-related and dedicated parts. This procedure contains the DAEC Receiving Inspection Instruction Checklist (QA Form 1105.1-1A) which includes a generic checklist designed specifically to help the QC inspectors detect counterfeit and fraudulent material.
in addition to this generic checklist, DAEC also has detailed checklists for detecting counterfeit valves, fasteners, and circuit breakers.
The OC inspectors who were interviewed during the inspection appeared experienced anti very knowledgeable about the methods used to detect counterfeit and fraudulent material. DAEC's policy is to rotate QC inspectors throughout the plant so that they gain experience in different areas.
This helps the inspectors understand where and how the parts they are inspecting at receiving will be used, enabling them to more easily detect anomalies in new In addition, all the personnel involved in the receipt of parts at the parts.
plant, including warehouse storekeepers, QC inspectors, and QA engineers who write the receiving inspection instructions, have attended training sessions given by outside contractors on the detection of counterfeit and fraudulent material.
2.5 Mceiot Insoection The NRC inspectors identified a strength in DAEC's procurement and dedication program in that DAEC utilizes a Spectrotest-F spectrographic analyzer (SA) to analyze the chemistry (including carbon content) of CGIs and basic components to verify receipt of the specified material. Additionally, all tests and-inspections are performed by independent QC inspectors qualified to American Nation.
Standards Institute (ANSI) Standard N45.2,6.
During the inspection of the DAEC receiving inspection area, a three-way manifold valve was being tested (Dadication Package 092-028) using the SA.
Chapter ll, " Inspection and Testing," Revision 5, of the IELP QA Manual gives requirements for the control of measuring and test equipment (M&TE),
l l -
Weaknesses noted during the review included: no unique M&TE controls assigned to the SA; no prescribed interval or frequency established for the calibration and adjustment of the SA; and no calibration sticker affixed to the SA. Addi-tionally, when the SA was found to be out of calibration during performance of its test, the operator adjusted it using nationally recognizei standards; however, no evaluations were being made and documented concerning the validity of previous test results. Also, it was not clear to the inspection team that a records system had been established and maintained for the SA showing the results of the calibration. During the inspection, DAEC QA personnel issued Revision 1 to Procedure 2161.15 " Chemical Analysis of ferrous and Non-ferrous Materials With the Spectrotest-f Spectrographic Analyzer," which now provides requirements that implement Chapter 11 of the IELP QA Manual and addresses the weaknesses identified by the inspection team.
3 DEDICATION PACKAGE REVIEW FINDINGS To help the NRC review individual dedications, IELP prepared (at the NRC's request) approximatcly 30 files of dedication packages compiled from diverse records, but each pertaining to one dedication, as selected by the team from its review of the DAEC dedication file lists. The review packages were organized by discipline into (1) electrical and instrumentation, (2) mechanical, and (3) materials (including lubricants).
In addition, IELP provided, as applicable, the associated commercial grade survey report if EPRI Method 2 was used.
The team reviewed the available records for the selected dedications; these included receiving reports and receipt inspection reports.
The following 12 examples are items that IELP purchased as commercial grade and either installed or made available for installation in safety-related plant applications without performing a fully specified review for suitability of service or in some cases, a design verification (seismic and environmental.
evaluation).
The NRC inspectors did not consider that the findings documented in the examples resulted in the use of CGis that could cause operability problems, however, they were unable to assure that this was the case.
Accordingly, for those identified and'similar items, DAEC should review the dedications to assure that all parts are suitable for their intended safety-related applications.
(1)
Dedication Package D90-003, Revision 2, May 14, 1991, dedicated a three-quarter-inch, four-and-one-half-pound, cast steel relief valve (model number 3/4 1990C-XDAl), manufactured by Dresser incorporated, which replaced a three-quarter-inch, seven-eighth-pound, bronze relief valve (PSV7333A), manuf actured by Aquatrol, Incorporated, and installed on control building heating, ventilation and air conditioning (HVAC) air system accunulator IV-5-12.
The Q-200 data sheet identified the safety function of the relief valve as forming part of the pressure boundary in a Safety Class 3 system and having only a passive integrity function.
System overpressure protection was not listed as a safety function.
The critical characteristics identified were configuration, pressure integrity, popping point, and assurance that the valve conformed to American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code, Section Vill. Material of construction was not identified as a critical characteristic.
The verification methods for accepting l l
each critical characteristic wore listed as visual inspection, dimensional verification, weighing the valve, pressure testing performed at 88 pounds per square inch gauge (psig) and ASME Coda Section VIII stamp verification.
The engineering evaluation for the replacement relief valve did not address the material substitution and its suitability for use in the intended safety-related application. DAEC's position was that a specific material for the re'ief valve is not required by any design specification; therefore, a replacement relief valve made of bronze, carbon steel, or some other material is acceptable providing it meets the requirements, if any, which are included in the approved specifica-tion. DAEC stated that Engineering is not required to reverify the adequacy of the original design during the procurement process. The inspection team noted that-Sections 1.3, 5.5, and 10.0 of Specification 7884-M-86, " Technical Specification for HVAC Instrument Air Compressor,"
Revision 2, March 27, 1975, required, in part, that the supplier of the units and accessories be responsible for their design; the accessortes include the air receivers (accumulators) that utilize a relief valve design meeting applicable ASME Code requirements. The manufacturer was also required by Section 10.3 of the specification to submit to DAEC design output documents and data for use in the plant design. The inspection team noted that DAEC had committed to ANSI Standard 18.7-1976, " Administrative Controls and Quality Assurance of the Operational Phase of Nuclear Power Plants," and that Section 5.2.13 of ANSI 18.7 requires, in part, that if the requirements of the original item cannot be determined, an Engineering evaluation shall be conducted and documented to establish the requirements. Additionally, Criterion !!! of Appendix B requires, in part, that measures.be established for the selection and review for suitability of application of materials used in safety-related applications. During the inspection, DAEC generated an office memorandum to justify using cast steel material as a replacement for bronze material.
DAEC's position on not identifying the valve's material as a critical characteristic is that the relief valve's only safety function is-to maintain pressure boundary integrity, and that pressure testing is an_
acceptable method of obtaining assurance that the relief valve will-per-form its safety function. The inspection team questioned this position since a one-time-only pressure test may provide assurance.that the component and system do not leak, but does not provide assurance of the material's strength or corrosion resistance. Material may need to be verified to ensure that the item will perform its-intended safety function when subjected to the full range of normal and design-basis operating environments and events.
The Q-200 data sheet did not identify overpressure protection as a safety function.
DAEC's position was that the compressor for the HVAC-instrument air has start /stop controls (pressure switch) that. shut the compressor off when the pressure in the receiver is at approximately 88 psig. Section 5.3.1 of Specificat. ion 7884-M-86, Revision 2, originally required that " Pressure control shall be set to start the L (
compressor at 85 psig and stop at 100 psig. Receiver mounted relief valve shall be set at 125 psig," The inspection team expressed concern-that DAEC was taking credit for system overpressure protection using-overpressure control devices (i.e., a pressure switch stopping a compressor from functioning) that-are not recognized by the ASME Code or by the original design specification.
The ASME Code requires the use of in-line pressure-relief devices for overpressure protection. Although i
not identified on the Q-200 data sheet as a safety function, the valve's
)
overpressure protection capability was verified during the dedication process.
Other concerns identified by the inspection team included: (1) an incorrectly dispositioned procurement action request (S-55728-01, November 20, 1990) that permitted the relief valve to leak at 115 psig, 1
before it could reach its 125 +/-4 psig popping point (Note:
t"v valve was later correctly tested) and (2) during the closecut of the maintenance action request (MAR), the equipment data base for relief l
valve PSV7333B was incorrectly updated. DAEC System Engineering corrected this data base discre)ancy during the inspection. Dedication Package D92-010, Revision 0, Te3ruary 25, 1992, was another example reviewed during the inspection that resulted in similar concerns.
(2)
Dedication Package D90-005, Revision 0, February 14, 1990, dedicated four 1-inch, socket-welded, 600-pound, stainless steel, conventional, port-gate valves manufactured by Velan Valvo, incorporated. The valves were originally purchased as nonsafety-related in 1988 and were dedicated as basic components for use in the emergency diesel generator air start system, The Q-200 data sheet for these four drain valves (V32-0065, -0069, -0073, and -0077) indicated that their safety function is forming part of the air start system pressure boundary and that their draining function is nonsafety-related.
The critical characteristics identified were shell, seat, and backseat leakage; body, bonnet, and gate material; and configuration. The verification methods.for accepting each critical characteristic were listed as reviewing leak test reports sent with the original purchase order (P0); reviewing.
certified material test reports (CMTRs) supplied by the manufacturer to ensure the material conforms to the American Society for Testing and '
Materials (ASTM); and verifying configuration and systerr, identification.
DAEC used a third-party audit conducted by the Nuclear Suppliers Quality Assurance Committee (NSQAC) at the Velan Valve facility in.Williston, Vermont, on May 27-29, 1987 (NSQAC Audit 87-18) as the basis for accepting the documentation supplied by Velan. However, the 1988-purchase order (P0 S42548) was issued to Velan's Montreal, Quebec, facility..The CMTRs and certificate of conformance (CoC) issued for the material and-testing ap) eared to have been generated by the Quebec facility. Therefore, tie use of the audit or survey of the Vermont-facility as a basis for accepting documentation and items from Quebec is not consistent with the guidance contained-in EPRI NP-5652, or in NRC Gl. 89-02 Additionally, a review of the NSQAC audit revealed that the majority of-the activities addressed in the audit.were related to,
. ~
Velan's ASME Code Section 111 quality program with very little documentation supporting the review of its commercial quality program.
DAEC gave the inspection team documentation supplied by Velan that addressed the traceability of the installed valve:.
Each of the four valves had raised cast markings indicating F-316, A-BF, and had-a dye-stamped mark of Velan'600.
The inspection team reviewed the receiving inspection report and documentation supplied by Velan and determined that the heat code for the valve bodies listed on the CoC was A-BF and that the CMTRs supporting the CoC indicated the ~ correct material for the valve body, bonnet _and gate material. However, it is noted that the CoCs and valve marking were produced in accordance with a QA program not reviewed or audited by the licensee.
(3)
Dedication Package D90-016, Revision 0, May 16, 1990, with attached revision control sheets through Revision 6, May 20, 1991, dediccted safety-relatedlouver spare parts such as stainless steel pinions, wing and jam gaskets, rivets, nuts, and bearings. The Q-200 data sheets listed several plant dampers (equipment identification numbers D07514A, B, C) in which these parts could be used and identified their safety function as positioning the ventilation dampers.
The critical characteristics identified were part number, configuration, materials of construction, and dimensions. The verification methods for accepting each critical characteristic consisted of a visual inspection, a review of the CoC, identification of part number, and verification of_ the material specification.
For the pinion, an independent chemical analysis of the material (ASTM A351 Typo 304, CF8) was required to be performed on a sample basis in accordance with HIL-STD-105D, along with a random dimensional check.
The dedication package incorporated the performance of a commercial grade survey in order to ensure that items were designed, procured, and controlled sufficiently to warrant acceptance following a standard receipt inspection along with a review of the supplier's, Construction Specialties, Incorporated (CSI), 000...The six attached revision sheets for D90-016 limited the use of the survey of CSI by requiring that DAEC perform dimensional checks on a sample basis of all louver parts and that a chemical analysis of the pinions also be performed on a sample basis. The dedication package permitted the use of CSI's CoC as the basis of accepting the materials of construction for all of the louver parts except the pinion, even though the survey stated, in part, that
" Certification documents are supplied by the vendors upon request from CSI. CSI prepares its CoC based on these vendor documents. CSI does not verify the correctness of certification documents with regard to material." Therefore the use and validity of the C i from CSI, as the basis for accepting these materials is not considered consistent with the specifications of Section 10.2, " Certificate of Conformance," of __
i ANSI N45.2.13-1976, " Quality Assurance Requirements for the Control of Procurement of items and Services at Nuclear Power Plants."
DAEC analyzed the chemistry of the pinions and checked dimensions on the louver parts.by randomly sampling these items in lot sizes required by
_g..
Hil-STD-105D. However, there was no documented basis to support the i
vendor's lot or batch control process, nor was it evident how the i
supplier controlled product consistency for the purpose of determining lot homogeneity.
The-inspection team also discussed with DAEC procurement personnel the use of sampling, based on statistical methods, when purchasing bulk or comodity items from organizations that do not maintain Appendix B QA programs. DAEC gave the inspection team documentation that indicated all louver parts, dedicated as basic components and available for safety-related service, had been installed in nonsafety-related applications.
s (4)
Dedication Package D92-019, Revision 1, february 7, 1992, dedicated non-rotating, non-lubricated, one-quarter-inch-diameter wire rope (19X7 classification) for the auxiliary hoist used for various-purposes, including refueling.
The critical characteristics identified were part number, configuration, dimensions, and material strength.
The verifica-tion methods for accepting each critical characteristic consisted of a visual inspection to verify configuration, length (105 feet), and diameter, and a source surveillance to verify breaking strength of the cable.
Section 4.6.6.5.3 of General Electric Specification 21A9246, " General Requirements for Refueling Platform," Revision 1, January 13.-1970, required that the cable be unlubricated stainless steel; however, mate-rial was not identified as a critical characteristic. The inspection team considered material (stainless steel application because of the potentially cor)rosive environment.to be important for The source inspection confirmed the breaking strength of the cable as 5550 pounds. However, a carbon steel cable of the same diameter and configuration has a breaking strength of 5540 pounds (exceeding the breaking strength of stainless steel which is 5400 pounds breaking strength alone does not assure that the cable was);made oftherefore, stainless steel. The source inspection report indicated that the CoC, generated by Wire Rope Corporation of America, was reviewed and that the material supplied was stainless steel. The report did not indicate that the supplier and tester of the cable (Yale Hoist) performed any independent overchecks to support the validity of the Coc, as specified by Section 10.2 of ANSI N45.2.13-1976.
(5)
Dedication Package D91-013, February 13,-1991, dedicated a three-eighth-inch Whitey needle valve which replaced an identical valve installed in the plant. The safety function of the valve was maintaining system pressure boundary. The function of the valve was isolating pressure switch PS4556 which provides reactor vessel pressure input to the low-pressure coolant injection loop select logic.
The critical charac-teristics listed were: part number, dimensions as specified in the-
. log, and performance of a hydrostatic test.
The verification 9ds for accepting each critical characteristic consisted of a n adard receiving inspection and a hydrostatic test performed by DAEC's Inst.umentation and control laboratory. The NRC inspection team-identified that the hydrcstatic test was only performed with the valve in the open position; therefore, there was no assurance that the valve -
. ould be capable ef performing itsL intended safety function when the w
valve was closed and-being used to isolate the pressure switch.. The' valve also required testing in the closed position to ensure that the-valve would be capable of maintaining system design pressure.
(6)
Dedication Package D90-105,-Revision 0. December 12, 1990, dedicated from warehouse stock several-one-eighth-inch by one-half-inch Woodruff keys to be used in various models of Fisher control valves.- :The.
critical characteristics identified by DAEC included part: number and dimensions (these were to be verified by standard receipt inspection) and chemical analysis (performed by Mobile Metal Analysis (MMA)
Laboratories, Huntington Beach, California;= an Appendix 8 qualified laboratory). The chemical analysis verified that the two keys supplied to MMA for destructive testing met the material. requirements of ASTM A29 (1035) steel. No documentation was available to sup) ort lot. homogeneity of the eight keys since they were. purchased from Fisaer Controls as commercial grade; therefore, testing two keys had'no direct correlation-to the remaining keys. DAEC installed one key in control. valve CV 4301 (isolation valve for standby gas treatment torus exhaust) and the NRC inspector verified the-remaining five keys in warehouse stock. No keys.
were issued for installation under Revision l~or 2 of. the dedication package. Other applications of these control valves, according to~'the Q-200 data sheet,: included standby gas treatment torus and.drywell exhaust.
The keys are manufactured to commercial standards (ANSI 817.2).
and are used as anti-rotation devices; as such, they must maintain their structural integrity. DAEC did not establish if Fisher had a' program in place to control lot or batch homogeneity.
j y
(7)
Dedication Package D90-021 dedicated a spare Woodward model 8901-037 booster servomotor for the Woodward model EGB-100 governors (tag numbers IG021/ GOV and IG031/ GOV) used on DAEC's emergency diesel generator-(EDG) engines-(1G021/ENG and 1G031/ENG). The item was purchased for potential ~
use in an u) coming outage'under IELP P0 555870 (Revision 0) issued to the Fairban(s-Morse Engine Division (FMED) of Colt Industries on May 17, 1990.
The purchase order identified the item by IELP stock code (Pll8563), the Colt-FMED part number (16200525), and the Woodward part number.(8901-037), and was required to be tagged ~with'the.IELP purchase order number and Colt-FMED part number as a. minimum.-
r The Q-200 data sheet for the booster servomotor's-parent component, the_
governor, listed hydraulic fluid pressure boundary as being a safety-function of the governor as.well as diesel. engine speed control.--
However, the safety functions (as given 'in D90-021) of the booster.
l servomotor (which forms part of that pressure boundary) did not include its being a hydraulic fluid (governor oil) pressure boundary (boosted to 250 psig nominal during start), nor a pressure boundary for the starting air system (also nominally 250.psig).
The critical characteristics identified by DAEC consisted of part number, seal integrity, mounting dimensions, and spring servo travel.
These were to be verified by Woodward under a separate P0 for IELP using Woodward Test Procedure TSP 344. The package indicated that spring
" L_.
s,,.h.
. a _ _ _._.,2c_ _ _. - -
aa - --
um
servo travel was an indirect functional performance requirement related to the response of a special test device used by Woodward as part of -
Woodward's Test Procedure TSP 344. however, it was not documented in the package how ILLP had determined that spring servo travel (of a-specified amount, but within no specified time) would ensure that-the 8901-037 booster servomotor would pressurize the EGB-100 governor upon its pressurization by starting air (at DAEC's minimum allowable pressure) so that the governor would open DAEC's engine' fuel racks the required amount and within the required time to ensure engine startup within the technical specification limit (10 seconds).
It was also not-documented how the performance of Test Procedure TSP 344 would ensure that the booster would release its pressurization of the governor and allow its own-oil pump to take over at the required engine speed or upon the removal of starting air to prevent overshoot, or worse, overspeed shutdown of the engine.
According to the dedication instructions, the critical characteristics-of seal integrity and spring servo travel were to be verified by means-of a source inspection. The objectives of the source inspection were discussed in detail in Design Engineering Memorandum NG-90-1736. The memorandum's source inspection requirements included witnessing Woodward's performance of TSP 344 and verifying no oil leakage:from the vent hole and no air bubMes in the effluent oil, but 'did not mention "no excessive air leakags from the vent hole (per Step 7 of TSP 344).
The source inspector was supposed to verify traceability of the booster by tracking number (P0 number) and part number to-the manufacturer and to test results. However, IELP produced a memorandum from the source inspector that explained that he had marked the booster with his own special marks to preserve traceability since the unit had ni serial number and the source inspection plan failed to require reco. ding the tracking number. The NRC inspector examined the unit and observed the-marks as described-in the memorandum,. indicating that it was the same-unit described by the source inspector, but:the test-and repair records for the unit still did-not reflect this means of identification. The completed test record consisted of a filled out Jage 4 of TSP 344, I
L Revision 0, October 9, 1990. -Step 1 indicated t1at:the unit was to be tested with SAE 10W40 oil, but it was not stated in-the dedication-package whether the unit will use the same oil when installed on a DAEC
- EDG, Finally, Step 5 of TSP 344 called for setting the supply air pressure between 100 and 300 psig and the recorded pressure was 124:
psig.
However, the DAEC starting air pressure is nominally 250 psig.
Therefore, while the unit did move the spring return servo 1.1 inches (at least 1.0 was required), at.a pressure lower than installed, the test did not verify lack of oil or air leakage (i.e., seal integrity). at the pressure to which the unit would be at least nominally exposed when installed.
Although not addressed as part of the dedication, IELP claimed that the pressure integrity of the unit would be checked-(although not necessarily at maximum design pressure) upon startups during-preoperational testing of the EDG on which it would be installed.
(8)
Dedication Package D90-012 dedicated a General Electric (GE) heater element, part number 470518675, to be used in the main steam isolation valve lea (age control system (tag numbers IS122A,1H3713). procured from NSSS-Divesco.
The stated safety function of the heaters was to maintain leakage as vapor and to evaporate leakage condensate. No failure modes were identified, but the critical characteristics listed were heated length, overall length, diameter (to be measuied during a source inspection (SI) at Divesco), wattage rating and serial number (to be verified visually during the SI), GE master parts list number E32-B001, and GE part number (to be verified visually during receipt inspection).
The evaluation addressed seismic and environmental qualification on the basis of establishing that the heater element from Divesco was a like-for-like replacement. Although the stated characteristics were all appropriate attributes for item identification, the NRC inspector was concerned about the omission of verification of functional performance-related attributes for heater elements, such as. direct current (de) resistance, actual heat output, insulation resistance or dielectric strength, and pressure-retention capability (pu ess in a well).
IELP's reply to these concerns was that this procurement was actually of a com--
ponent originally manufactured as a basic component under an Appendix B QA program and that the dedication process was used merely to reesteb-lish traceability to'the original equipment manufacturer. Although this was not an unreasonable a)proach, it did not take into consideration that this heater element 1ad changed hands several times and had been stored and handled under unknown conditions making independent i
verification of certain key attributes prudent.
(9)
Dedication Package D90-034 dedicated Westinghouse, type FH03 through FH94, overload heaters to be used in various Class lE (safety-related) motor control centers at DAEC.
The safety function of the part was stated as "not to cause the overload relay to tri) sconer than desired."
The critical characteristics listed were part num)er and trip time as determined by IELP Procedure GMP-TEST-031 during-post-installation testing, but starting, transient, and full-load hold-in capability were not addressed.
These unlisted attributes were considered to be essential to the stated safety function.
IELP stated that this was an oversight that would be corrected by adding the hold-in test'to GMP-TEST-031 in a future revision, but that credit was taken for operational testing that had not resulted in any premature tripping.
The DAEC evaluation of overcurrent tripping tests (per GMP-TEST-031) stated that "one pole at a time was tested," but a test circuit diagram in GMP-TEST-031 indicated that three poles would be tested in series.
This would not provide conclusive results on each individual heater.
GMP-TEST-031 stated that the acceptance criterion was tripping tire within +/-15 percent of the time corresponding to the test current on the manufacturer's trip curves. However, the DAEC evaluation of the raw data indicated that the 15-percent tolerance was interpreted to be the accuracy of the measured time values as opposed to the tolerance of the measured trip time with respect to the trip curve. The test results as documented were therefore inconclusive with respect to the stated acceptance criterion.
IELP agreed to review the results. l
[
r e-4 9
(10) Dedication Package D90-044 dedicated various AMP Corporation crimp-type electrical connectors (insulated and uninsulated) for general purpose safety-related applications.
The generic safety function was described as maintaining electrical and mechanical integrity of the circuits in which the connectors were used.
Failure modes adverse to safety were not specifically addressed.
The stated critical characteristics for the connectors were part number, conductor material, barrel length, inside diameter and wall thickness, insulation material, and insulation thickness.
The characteristics of part number and dimensions were to be verified during receipt inspection by visual observation and measurement, and material was supposed to be verified visually and by part number.
The conductor material and insulation matcrial were to be certified in a CoC from AMP Corporation and the validity of AMP's certifications was to be determined by a commercial grade survey.
Review of the survey report (V-90-85) indicated that it resembled a broad-based, programmatic, Appendix B-type audit. Although there was a survey plan that listed the critical characteristics of metals, insula-tion, insulation thickness and metal thickness (per AMP drawings), the survey report contained no evidence that all of these parameters were verified.
In f act, the report criticized AMP's traceat'ility of materials to documentation of particular composition or properties and recommended that IELP not require conformance to certain material specifications because AMP could not certify to them, it was not clear from the receiving documents that anything more than part number, quantity, and receipt of the vendor CoC was verified.
(11) Dedication Package D90-045 dedicated a heater for the standby liquid control (SLC) tank manufactured by Wellmen Thermal Systems.
The safety function of the heater was described as maintaining the temperature of the SLC tank pentaborate solution above precipitation temperature, f ailure inodes or effects were not discussed.
The Q-200 data sheet for the heater (tag number lH3446) gave the safety function simply as
" provide heat." The critical characteristics identified were part number, operating voltage, wattage, dimension "B" and rod diameter from vendor drawing 40A109295, and de resistance. However, the temperature at which the dc resistance was to be taken was not specified, nor was the required resistance.
In addition neither insulation dielectric strength nor insulation resistance normally considered critical characteristics for such applications were specified.
In response to the team's concerns, IELP obtained documentation from the vendor stating the expected " cold" dc resistance was 37.2 ohms 4/-10%.
IELP stated that the heater had met that criterion. With regard to the omission of checking insulation resistance (IR) or dielectric strength, IELP stated "lR prior to installation is a personnel safety characteris-tic," however this statement was considered invalid because the heater would not be expected to be energized prior to installation since no preinstallation operability bench test was specified.
IELP also contended that insulation resistance would be indirectly tested (at nominal service voltage) during post-installation testing, but an operational check of a heater circuit would not necessarily detect the relatively high leakage currents indicative of degraded insulation. 1 l
(12) Dedication Package 090-092, Revision 5, April 10, 1992, dedicated various commercial grade lubricants for use in safety-related equipment, including, for example, MOBIL " Light DTE" oil, EXXON " Nebula EP-0" grease, and MOBILGREASE #28, used in the core spray and residual heat removal pump motors, in safety-related Limitorque motor-operated valves (HOVs), and in Limitorque MOV limit-switch gearboxes, respectively (all environmentally qualified equipment). Visual inspection for such critical characteristics as color, product identification information (e.g., type designations and batch numbers), and container integrity I
were performed on site; whereas, IELP had been using 3rguth Laboratories, an Appendix B, IELP-qualified con) actor, to perform lubricant dedication testing of such critical characteristics as viscosity of oils and dropping point and penetration of greases. The team determined that, in general, IELP's dedication methodology for lubricants was acceptable; although, some concerns were identified.
The sampling plan for lubricants described in the evaluation, in which traceability to the manufacturer was to be established by lot or batch numbers, was based (,n a HIL-STD-105D statistical sample, which assumes batch control and reasonably homogeneous batches (a batch being defined as units of product of a single type, grade, class, size, and composition, manufactured under essentially the same conditions, and at essentially the same time). However, it was.not clear frem the evaluation how IELP had determined what the suppliers' batches or lot controls were, nor was it clear how the suppliers' r.ontrols on product consistency were verified for purposes of determi:itng 'oatch homogeneity.
4 PROCUREMENT AND DE01 CAT 10N TRAINING The inspection team reviewed IELP's training activities in support of the process of dedication of CGis used in safety-related applications performed after January 1, 1990. Until mid-1991, IELP had relied heavily upon contractors to provide engir. sering support for procurement activities.
Since 1991, IELP has developed a Procurement Engineering group of in-house personnel; the group was of comprised four engineers and a group leader. With the exception of the group leader, the engineers had less than one year of experience in the group and had little previous procurement experience. The inspectors noted that the training program for the procurement-angineers was not fully implemented.
IELP stated that it expects full implementation by September 1, 1992.
Past training of the procurement engineers consisted primarily of on-the-job training. Additional training on the EPRI procurement guidelines was given at training seminars held in March 1990 and April 1992.
Since the beginning of 1991, procurement-related topics had been covered during each of the quarterly training sessions provided to the Engineering staff. Among the topics covered were bill of materials, vendor performance-based audits, and the NUMARC comprehensive procurement initiative. An overview of the procurement process is also a part of the general orientation training provided to all new engineers. 3
Training had also been provided to thore most effected by procedure revisions, for example, the procurement engineers received about a half-day of training on the April 1992 revision of the procurement procedures; training consisted of an informal discussion of the changes. Training for the Quality Assurance engineers (OAEs), those engineers who reviewed procurement packages, consisted of a qualification card system that had been in place since the beginning of 1990. Completion of the qualification cards required familiarization with the associated procedcres, performance of specific on-the-job activities, completion of specific training courses, discus; ion of the various topics with the supervisor, and the successful completion of an exam.
For those who had pre"tous experience, completion of the qualification cards was waived on a case-by-case basis.
Only individuals who had completed the qualification cards for a specific function were authorized to perform that function.
As a group, the QAEs had sufficient QA experience but a limited engineering background.
The training given to QC inspectors was also based on a qualification card system. Training had been provided on receipt inspection, CGls, and detection of fraudulent parts.
Interviews with IELP's QC inspectors showed them knowledgeable of the practices to detect fraudulent parts; as a group, they had considerable nuclear and QA experience.
In general, the inspection team considered IELP's training activities to be effective.
5 EXIT MEETING On May 15, 1992, the inspection team conducted an exit meeting with members of tha IELP staff and its management at the DAEC site. During the exit meeting, the team summarized the inspection findings and observations.
The following individuals were present:
lowa_flectric Licht and Power Comony J. Franz, Vice President - Nuclear M. flasch, Manager - Engineering D. Church, QA Supervisor D. Jantosik, Group Leader, Materials and Supplier Quality D. Wilson, Plant Superintendent T. Sims, Nuclear Licensing Specialist J. Howlett, Group Leader, Procurement Engineering H. Johnson, Supervisor, Material Management M. Fairchild, Group Leader, Vendor Programs K. Peveler, Manager, Corporate QA W. Rose, Engineer, Nuclear Safety Committee B. Bernier, Supervisor, Mechanical Engineering D. Engle, Procurement Engineer F. Baines, Procurement Engineer K. Medulan, Procurement Engineer B. Borek, Procurement Engineer D. Baumgartner, Procurement Engineer P. Hansen, Systems Engineer M. Huting, QC Supervisor A. Roderick, Supervisor, Testing and Surveillance b
R.~Aiken,l Senior'QASpecialist
,e
- G. Falta Senior QA Specialist J. Zullo, QA-Specialist D. Homes, QA Specialist!
L. Jenkins, Senior QA-Engineer _ - Procurement
~J. Powers, Senior QA~ Engineer - Procurement z;
D. Podlin, QC Engineer i
r a nef, Senior Civil / Structural. Engineer-
.4
<ng, Manager, Nuclear Licensing-
=
' ~
P.
Esetti, Supervisor, Regulatory Communications.
s'
.iair, Group Latader, Internal _QA Audits L. Mattes, Management Support-R. Salmon, Staff Engineer N. bikka, Supervisor, Electrical Engineering M. Rader, Instrumentation and Control Maintenance G. Whittier, System Engineer J. Gushue, QA Engineer D. Pint, QA Engineer Nuclear Reaulatory Commission L. Norrholm, Chief, VIB F. Jablonski, Section Chief, Rill R. Pettis, Team Leader, VIB S. Alexander, EQ and Test Engineer, VIB L. Campbell, Reactor Engineer, VIB S. Magruder, Reactor Engineer, VIB R. Langstaff, Reactor Inspector, RIII M. Parker, Senior Resident laspector, DAEC Other Oraanizations W. Houston, EPRI
.B.
Bradley, NUMARC W. Ford, Consumers Power Company
- -100-o
/
'c UNITED STATES
/ ;'.
- ?
7 NUCLEAR RESULATORY COMMISSION-g
.)
i tAssmotow, c. c. 20sss -
W$ l October 19, 1992' Docket No. 99901235 b
Mr. Gerhard Liesegang, Chairman Lisega G.bH Industriegebiet Hochkamp l
Postfach 1340 0-2730 Zeven, Germany Dear Mr. Lier 1:
SUBJECT:
NOTICL of NONCONFORMANCE (NRC INSPECTION REPORT NO. 99901235/92-01)
This letter addresses the inspection of your fat.Hity at Zeven, Germany, conducted by Mr. Richard P. McIntyre and Mr..Uldis Potapovs of this office on August 18 through 21, 1992, and the discussions of their findings with you and other members of your staff at the conclusion of the inspection. The purpose
~
of the inspection was to review Lisega's corrective actions for previous inspection findings and unresolved items and to evaluate Lisega's quality assurance (OA) program, including material procurement, control and audit of.
subvendors, material certification, dedication of comercial grade items used in co,ponents sold as nuclear safety-related, and the upgrading of stock material per the requirements of the Americar Society of Hechanical Engineers (ASME) Boiler and Pressure Vessel Code.
l Area m d during the U.S. Nuclear Regulatory Commission (NRC) inspection i
and t - tu ngs are discussed in the enclosed inspection report. The inspes onsisted of an examination of. procedures and representative records,...terviews with personnel, and observations by the inspectors.
The inspection identified that the implementation of your QA program failed to meet certain NRC requirements which are summarized as follows:
1) insuf ficient basis for certificati9n of supplied nuclear componen(ts' as meeting the requirements of ASME Code,Section III, Subsection NF, (2) the failure to include provisions in the Lisega QA Manual and the QA Program implementing Procedural Guidelines (VQSP.'s) for dedicating items purchased by Lisega as-commercial grade and included as part of pipe support components that are sold by Lisega as safety-related 10 CFR Part 50, Appendix B items and not as ASME.
Code items, and (3) the purchasing of items from suppliers who hold a current-ASME Quality Systems Certificate (QSC) or are listed on the German government Register of Approved Material Manufacturers (TOV 1253/1), without performing any implementation audits of these suppliers' quality programs.
-101-
Mr. Gerhard Liesegang '
Pleat.e provide us within 30 days from the date of this letter a written statement in accordance with the instructions in the enclosed Notice of Nonconformance. We will consider extending the response time if you can show good cause for us to do so.
The responses requested by this letter and the enclosed Notice of Nonconformance are not subject to the clearance procedures of the Office of Management and Budget as required by the Paperwork Reduction Act of 1980, P@lic Law No.96-511.
In accordance with 10 CFR Part 2.790 of the NRC's " Rules of Practice," a copy of this letter and its enclosures will be placed in the NRC's Public Document Room Should you have any questions concerning this inspection, we will be pleased to discuss them with you.
Sincerely, v
I li v
Leif J. Nor holn, Chief Vendor Inspection Branch Division of Reactor Inspection and Licensee Performance Office of Nuc'iear Reactor Regulation Entiosures:
1.
Notice of Nonconformance 2.
Inspection Report No. 99901235/92-01
-102-I NOTICE OF NONCONFORMANCE Li5ega GmbH Docket No. 99901235 Zeven, Germany Based on the results of a Nuclear Regulatory Commission (NRC) inspection conducted on August 18-21, 1992, it appears that certain of your activities were not conducted in accordance with NRC-requirements.
A.
10 CFR Part 50, Appendix B, Criterion 111, ' Design Control," requires that measures be established for the selection and review for suitability of materials, parts, equipment, and processes that are essential to the safety-related functions of structures, systems, and components.
The American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code is the basis for demonstrating suitability for application of component supports and hydraulic shock absorbers supplied by Lisega for use in the Grand Gulf, Arkansas, and. Palo Verde nuclear power plants.
Contrary to the above, the material and test documentation for several iter.s which were certified by Lisega as meeting the requirements of ASME i
Code, Section 111, Subsection NF, did not ft.lly support this certification.
Specifically:
1.
Lisega issued Certificate 113 377 for SA 479, TP 410 (1) bar used for piston rods in large hydraulic snubbers ordered by Arkansas Power and Light Company (APL) for steam generator suppoits.
Lisega obtained this material from Gustav Grimm Edelstahl-Werke GmbH (GG) as SA 182 F6A Cl2 forging. GG provided a Certified Material Test Report (CMTR) for this material, including the mill heat analysis, heat treatment description, and NDE certifications on their letterhead. However, GG is not a holder of a Quality Systems Certificate (QSC) nor did their certification to Lisega include the statement that this material had been produced under an NCA 3800 quality program (no evidence that GG had been qualified by Lisega to supply Code material). A CMTR from the melting mill was not included in this documentation and there was no evidence that the mill had been qualified either by GG or by Lisega.
2.
Lisega issued Certificate 111 183 for A 668, Class C (and Lisega Specification 122) material used-for articulated joints in rigid struts supplied to Arizona Public Service Company (APS) under their purchase Order 33801236.
Lisega obtained this material from Lenhauser Hammerwerk GmbH, (LH).
LH provided a CMTR for this material, including the mill heat analysis on their letterhead. -103-
LH is not a QSC holder and the LH CMTR did not demonstrate that this
~
material was produced under an NCA 3800 program that had been approved by Lisega. CMTR from the melting mill was not included in the documentation and there was no evidence that the mill had been qualified either by LH or by Lisega. Additionally, although Lisega specification 122 restricts chromium content of this material to
.30%, analysis for chromium content was marked as not applicable on the Lisega product analysis.
3.
Lisega issued Certificate 115 217 for SA 53 S, Grade A pipe to be used for rigid struts supplied to APS. SA 53 contains restrictions on the maximum amounts of each of the following elements: copper, nickel, chromium, molybdenum, and vanadium as well as the requirement that the maximum combined level of these elements can not exceed 1.00%. The material was supplied by'Benterer as complying with DIN 2448-81/17175-79 without analysis of trace elements.
Lisega Certificate 115 217 (product analysis) reported only the average (combined) value of trace elements and restricted the total to less than 1.00%. This approach does not assure that the individual trace elements do not exceed permitted levels.
Lisega Certifications 115 431, 115 233, 115 284, 115 232, and 115 243 contained similar deficiencies.
4.
Lisega issued Certificate 115 399 for SA 479, TP 410 (1) material to be used for pin / bolt application in rigid struts supplied to APS.
This material was procured from Krupp Stahlag who provided a CMTR.
However, neither the Krupp CMTR nor the Lisega certification described the heat treatment of this material or reported the hardness level of the product as required by the applicable specification. Krupp provided this information by telefax during the progress of this inspection.
E.
10 CFR Part 50, Appendix B, Criterion III, " Design Control," requires that measures be established for the selection and review for suitability of materials, parts, equipment, and processes that are essential to the safety-related functions of structures, systems, and components.
10 CFR Part 50, Appendix B, Criterion V, " Instructions, Procedures, ar'id Drowings." requires that activities affecting quality shall be prescribed by documented instructions, procedures, or drawings, of a type appropriate to the circumstances and shall be accomplished in accordance with these instructions, procedures, or drawings.
Contrary to the above, as of August 21, 1992, neither the Lisega QA Manual nor the QA Program Procedural Guidelines (VQSP's) included any provisions for dedicating items purchased by Lisega as cornercial grade and used as part of pipe support components that are sold by Lisega as safety-related 10 CFR Part 50, Appendix B items. -104-I l
1
.C.
10' CFR Part' 50, Appendix. B, Criterion VII, "Contro11of-Purchased -
lNaterial Equipment,.and Services,' requires that measures shall be-established to assure that purchased material, equipment, and services, whether purchased-directly or through contractors and: subcontractors conform'to the procurement documentsi The effectiveness of the control of quality by contractors-and subcontractors shall be assessed by the applicant or designee:at intervals consistent with-the importance,=
complexity, and quantity of the product or services.
Contrary to the~above, Lisega' purchased items from suppliers who hold a current ASME-QSC or are listed on the German government Register oO Approved Material Manufacturers (TOV 1253/1)lwithout performing any assessments, such'as implementation audits for. verification of the'
_l
-o suppliers' quality programs or testing the supplied material.-
~'
Please provide a written statement or explanation to:the U.S.-Nuclear:
Regalttery Co=.ission, ATIN:. Document Control Desk, Washington,-D.C.
20555 with a copy to the Chief Vendor Inspection Branch, Division of Reactor Inspection and Licensee Performance, Office of Nuclear Reactor Regulation, within 30 days of the date of the letter transmitting this Notice of lionconformance. This reply should'be_ clearly marked as a _" Reply to a Notice of Nonconformance" and should include for each nonconformance: (1) a-description of the steps that_ have or will be taken to correct these items;_
(2) a description of_ the steps that have been or will-be taken to prevent recurrence; (3) the dates your corrective actions and preventive measures:were or will be completed.
l Dated at.Rockville, Maryland day of L. n ;.
, 1992 this t -105-
-..ievauro e i
ORGANIZATION:
=Lisega GmbH
.Zeven, Germany;.
REPORT NO.:
99901235/92-01 CORRESPONDENCE-;
. ADDRESS:
.Mr. Gerhard Liesegang,. Chairman; Lisega GmbH:
Industriegebiet'Hochkamp Postfach 1340
' D-2730 Zeven,:Germanyi o -
ORGAN]IATIONAL Mr. Herbert Bardenhagen,-Quality _ Assurance Manager:
CONTACT:
NUCLEAR INDUSTRY Nuclear pipe support components incidding hydraulici ACTIVITY:
shock absorbers-(snubbers), constant. hangers-and other component standard supports.
INSPECTION CONDUCTEO:
August 18 through 21 -'1992-SIGNED:
7 d
^N b
Richard P. McIntyre,-Teafn Leader #.
-Date Reactive Inspection Section No. 1 Vendor Inspection-Branch,(VIB).
APPROVED:
IONE 2A ?
'I 6
9f-. -
UTcfis Pot'apovs, Chief Datei
~
4
~~
Reactive Inspection Seution No.-1 Vendor Inspection Branch (VIB).
INSPECTION BASES:
10~CFR Part 21 and 10'CFR Part_50,-Appendix B INSPECTION SCOPE:
To review Lisega's Quality.-Assurance' Program relative to the manufacture and supply of pipe support components,. including hydraulic shock-absorbers-(snubbers) to U.S..' utilities for use in nuclear-safety-related applications.
PLANT SITE APPLICABILITY:
Numerous 1
4
-106-
.w-
~*
-1 INSPECTION
SUMMARY
i 1,1 Nonconformances i
1.1.1 Contrary to 10 CFR Dart 50, Appendix B, Criterion !!!, ' Design Contrci," the material and test documentation for several components that were certified by Lisega as meeting the requirements of American Society of Mechanical Engineers-(ASME) Code,Section III, Subsection NF, did not fully -
support this certification.
(92-01-01) 1.1.2 Contrary to 10 CFR Part 50, Appendix B, Criterion 111, " Design Control," and 10 CFR Part 50, Appendix B, Criterion V, " Instructions,-
Procedures, and Drawings," neither the Lisega QA Manual or the QA Program Procedural Guidelines (VQSP's) included any provisions for dedicating items purchased by Lisega as commercial grade and used as part of pipe support components that are sold by Lisega as safety-related 10 CFR Part 50, Appendix B ite :s.
(92-01-02) 1.1.3 Contrary to 10 CFR Part 50, Appendix B, Criterion Vil, " Control of Purchased Material, Equipment, and Services," Lisega purchased items from suppliers who hold a current ASME Quality Systems Certificate (QSC) or are listed on the German government Register of Approved Material Manufacturers (TUV 1253/1) without performing any implementation assessments of the suppliers' quality programs.
(92-01-03) 2 STATUS OF PRFVIOUS INSPECTION FINDINGS 2.2.1 (CLOSED) Nonconformance 99901235/91-01-01 Lisega had supplied components to Georgia Power Company and certified them 'to be in corpliance with ASME Code.Section III, Class 1, Subsection NF.
However, documentation reviewed during the 1991 NRC inspection did not' adequately support this documentation or did not correctly represent significant material parameters.
Lisega's response to this finding was documented in letters dated January 28, 1992 and June 17, 1992. Additionally, Lisega has provided corrected certifications and copies of their corrective actions to Georgia Power Co. and other customers of ASME Code components thus-informing:them of the discrepant conditions. With respect to example 7 of_ Nonconformance 91-01-01, concerning the designation of_ chrome-molybdenum-vanadium alloy steel as SA-696 carbon steel bar, Lisega has discontinued the use of this material for ASME Code applications.
Review of selected documents verified accompiishment of the committed corrective actions. However, selective review of documeritation pertainirig to current orders identified additional deficiencies related to the procurement and certification of ASME Code material.
2.2.2 (CLOSEDI Nonconformance 99901235/91-01-02 Several open bundles of nuclear grade steel material were identified in the -107-
front-parking lot laydown ares that did not contain the' original' Lisega
, nuclear grade material tagging. The individual' material pieces in'the bendles.
had not been stamped by QC as required by procedure.to maintain unobjectionable: identification and material traceability. Also, several
-yellow striped nuclear material tags were loose.in the laydown area.
i Nonconformity Report NCR 03111-6-2291, dated August 23, 1991, was written to document the NRC identified nonconformity and also to ' identify the required corrective actions. The material identified by the NRC inspection team was--
downgraded from ASME Code material to non-Code material. -Also, several training sessions on material storage'and QA requirements'were given-.to-personnel from the various-warehouses, fabrication QA and' control, and material specification and control.
The inspectors toured the outdoor material-storage. areas as part _of-the review of lisega corrective actions. Since the 1991 NRC inspection, Lisega has constructed numerous multi-level storage racks for both ASME Code and non-Code material.
None of the deficiencies identified during the 1991 inspection were noted during this inspection, the overall outside storage was vastly improved.-
and the remaining old laydown areas still utilized are to be transferred to rack storage in the near future.
2.2.3 LQPEN) Unresolved Item 99901235/91-01-03 ASME Code Cases had been used by Lisega for material supplied to the Georgia Power Company.
Lisega had not determined whether Georgia Power Company had approved the use of these code cases for Code material supplied under their.
purchase order, it was verified during this inspection that the code cases in question had been authorized in a Bechtel design specification applicable to this procurement.
It was noted, however, that the authorization to use these code cases was subject to restrictions imposed by NRC Regulatory Guide 1.85.
Lisega did not have a copy of Regulatory Guide 1.85 and had not reviewed it for applicability to this procurement.
Review of procurement requirements for other current contracts identified similar restrictions on the use-of code cases as well as specific restrictions concerning-the use of. article NF 2610 (small parts-exclusion). This issue remains unresolved.
(99901235/91-01-03) 2.2.4 (Cl0 SED 1 Unresolved Item 99901235/91-01 When upgrading stock material, Lisega was not performing product analysis on each piece of the material as required by ASME NCA-3867.4(e). Specifically, only one product analysis per heat of upgraded ASME Class 1 material supplied to Georgia Power Company was previded regardless of the number of pieces involved.
Although this material was originally certified as conforming to ASME Section 111. Subsection NT, Code Class 1, Lisega has obtained authorization to change the code classification to Class.2. By a letter to. Georgia Power Company-dated January 8, 1992, Lisega transmitted revised material certifications and advised that these certifications should be used as Certificates of Compliance -108-
(C0C) as_ permitted by.ASME Code Section III Subsection-NF, paragraph 2130.
-Although the_ Code is not explicit in this: area..Lisegathasimade_an.
'interpretatioithat testing each piece of stock material is not required when C0Cs are acceptable:.Lisega has' revised its procedures-to-require product analysis of_ each piece of upgraded stock material when _this material-is supplied with a Certified. Material Test Report;.This unresolved item is
-closed.
3 INSPECTION FINDINGS AND OTHER COMMENTS-3.1 Backaround a
Lisega GmbH is a manufacturer and supplier of pipe support components'.
including hydraulic shock absorbers (snubbers), constant: hangers,: variable spring hangers,. pipe clamps,-and other component standard supports. The snubbers are manufactured, assembled, and tested in their Bondoufle, France facility. Lisega (Germany) retests the snubbers-upon. receipt'at the Zeven-facility and then certifies them under their American-Society of Mechanical-Engineers (ASME) Quality System certificate (QSC) for ASME! Code orders.
Lisega was issued a QSC as a material-supplier (MS) by ASME on September 9.-
1990. Lisega GmbH also supplies non-ASME Code snubbers and_ other product line coeponents to United States (U.S.) utilities for safety-related nuclear-applications under their 10 CFR Part 50,-Appendix B quality _ assurance (QA) program.
Involvement with U.S. utilities has been ongoing-since 1986-87 and Lisega has supplied various pipe support components to numerous,U.S.. utilities as both-ASME Code and 10 CFR Part 50, Aspendix B items-during the last several years, lisega's Appendix B QA program ans been audited and approved by several-U.S.-
utilities.
Lisega is planning to build a warehouse facility in the Asheville,--Tennessee area in the coming year, with future plans to include manufacturing at the facility.
3.2 Current U.S. Nuclear Activity During the last several years'lisega has received orders from numerous U.S.
utilities for both 10 CFR Part 50, Appendix B components and ASME Code items.
Duke Power (September 1990), Florida Power and Light (September 1991),-and--
Commonwealth Edison (May 1992), have conducted audits of Lisega's QA program and its implementation at Zeven Germany or Zeven and Bondoufle, France.
Several other utilitles,-including-Niagra Mohawk, Northeast-Utilities, and
-Southern California Edison, have used these audits to place Lisega -on their.
approved suppliers list and purchased material. Some of. the more recent orders have come from Commonwealth Edison, Entergy Operations, Duke. Power, Arizona Public Service, Florida Power and Light, Duquesne Light and-Public Service Electric and Gas. The inspection' team chose several of the recent orders for re'tiew of Lisega's-processing of these orders and compliance to the.
utilities' purchase order requirements.
' -109-r
3.3 Procurement and Certification of ASME Code Material Selected material certifications and supporting documentation from recent and in-process orders from U.S. utilities were reviewed to verify specific corrective actions-in response to.the deficiencies identified during the 1991 NRC inspection and to assess.the effectiveness of these actions towards achieving overall improvements in practices related to the stocurement and certification of Code material. While the disposition of tie specific findings f rom the previous -inspection was generally-satisfactory, review of -
current material certifications-identified:the following additional deficiencies which are related primarily.to'the qualification of material suppliers ano verification of conformance with specification requirements:-
1.
Lisega issued Certificate 113 377 for SA 479, TP 410 (1) bar used fo_r piston rods in large hydraulic snubbers ordered by Arkansas Power and Light Company (APL) for steam _ generator supports. Lisega obtained this material from Gustav Grimm-Edelstahl-Werke GmbH (GG) as SA~182-F6A Cl2 --
forging. GG provided a-Certified Material Test Report (CMTR) for this material, including the mill heat analysis, heat treatment description, and NDE certifications on their letterhead.- However, GG is not a holder of a QSC nor did their certification to Lisega include the statement that this material had been produced under an NCA 3800 quality program (no evidence that GG had been qualified by Lisega to supply Code material). A CMTR from the melting mill was not included in this e
documentation and there was no evidence that the mill had been qualified either by GG or by Lisega.
2 -.
Lisega issued Certificate 111 183 for A 668, Class C (and Lisega Specification 122) material used for articulated-joints in rigid struts supplied to Arizona Public Service Company (APS) under purchase order 33801236.
Lisega obtained this material from Lenhauser Hammerwerk GmbH (LH).
LH provided a CMTR for this material, including the mill heat-analysis, on its letterhead. LH is not a QSC holder and the LH CMTR did s
not demonstrate that this material was produced under an NCA-3800 program that had been approved by Lisega. -A CMTR from the melting mill was not included in the documen'tation and there was no evidence that the mill had 'been qualified either by LH or by Lisega, ' Additionally, although-Lisega specification-122 restricts chromium content of this material to 0.30%, analysis for chromium content was marked as not applicable on the Lisega product analysis.
3.
Lisega issued Certificate 115 217 for-SA 53 S, Grade A pipe to be used' for rigid struts-supplied to APS. SA 53'contains restricti_ons on the-maximum amounts of each of the following elements: copper, nickel, chromium,-molybdenum, and vanadium as well as the requirement that the-maximum combined-level of these elements can not. exceed 1.00%; The i
material was supplied by Benterer as complying with DIN 2448-81/17175-79 l
without analysis of-trace elements.
Lisega certificate 115 217 (product.
l analysis) reported only the average (combined) value of trace elements i
and restricted the. total to less than l.00%. This approach'dces not assure that the individual trace elements do not exceed permitted -110-
levels.
Lisega Certificat1cns 115 431, 115 233, 115 284, 115 232, and 115 243 contained similar deficiencies.-
4.
Lisega issued Certificate 115 399 for SA 479, TP 410 (1) material to be used for pin / bolt application in rigid struts supplied to APS. This material was procured from Krupp Stahlag who provided a CMTR.
- However, neither the Krupp CMTR nor the Lisega certification described the heat-treatment of this material or reported the hardness level of the product as required by the applicable s)ecification. Krupp provided this information by telefax during tie progress of this inspection.
These material certification deficiencies are identified as Nonconformance-(92-01-01).
Dif ferences between specific requirements of European and ASME material specifications continues to be a problem in some applications.
for example, Lisega is " upgrading" German standard material to ASME SA 479 TP410 to meet Code requirements. However, their standard product design requires higher tensile and yield values than the SA 479 TP410 minimums. To accommodate their design needs, Lisega is procuring this material to a higher tensile level and adding a footnote to the material designation which reads: " Modified CMTR: To obtain YS and TS for this particular design, we have increased from 275 to 430 and 485 to 600 N/mm." Lisega is of the opinion that this practice is permitted by NCA 3866.6.
While there does not appear to be a technical concern regarding this particular application (the material still meets the SA specification) it does not appear that NCA 3866.6 intended to authorize the supplier to unilaterally modify SA specification requirements to accomodate its design preference, in a dif ferent application, Lisega is procuring hexagonal nuts to SA 563M and Lisega specification 017.
These specifications identify the nuts as plain-carbon steel.
However, the nuts are purchased to a German specification'as chromium / molybdenum alloy steel. Although these nuts meet the mechanical and chemical (for specified elements) composition requirements and, from an application standpoint, are as good or superior, the applied SA specification does not properly characterize the product. This discussion is not intended to identify a significant product deficiency but rather to focus on potential problems with equivalency or interchangeability of international material or design specifications. This area will continue to be assessed during future inspections.
3.4 Review of Liseaa Processina of Utility Purchase Orders 3.4.)
The team reviewed Lisega's processing of Duquesne Light Company's (DLC's) Purchase Order (PO) D111311, dated April 14, 1992, for 20 hydraulic shock absorbers for the Beaver Valley Power Station. This was listed as Lisega job number 1251. All DLC technical requirements were adequately transferred onto the appropriate Lisega fabrication documents, including the provisions of DLC Specification number 10080-DMS-42, " Procurement Specification for Hydraulic Piping Snubbers." The DLC P0 also required Appendix B QA requirements to be invoked on all Lisega subcontractors and for.
Lisega to provide a certificate of conformance to all P0 specifications and -111-
I codes.
The inspectors chose 12 vendoc subcontractors that supplied material certifications for material or niece parts used in the manufacture of the snubbers supplied to DLC.
The inspectors reviewed the contents of the vendor j
material certifications to Lisega as well as the Supplier Certificates provided by Lisega to DLC. Several of the items listed on the Lisega Supplier Certificate were made from material purchased from suppliers that had not been i.udited by tisega and the rAerit) was not tested upon receipt.
Four of the vendors that provided these material certifications were not included on the list of Authorized Suppliers and had not been audited by Lisega. This deficiency is further discussed in Section 3.5 of this report. The material received from these suppliers was used for the paddle ends and locknuts for certain snubbers. The snubbers were functionally tested prior to shipment to Beaver Valley.
3.4.2 The team reviewed Lisega's processing of Southern California Edison Company's (SCE's) PO 6N022003, dated February 4, 1992, for a specially designed adjustable rigid spring assembly for the San Onofre Nuclear Generating Station (SONGS). This was listed as Lisega job number 1232.
Several of the items listed on the Lisega Supplier Certificate were made from material purchased from ;uppliers that had not been audited by Lisega or the material tested upon recaipt. When reviewing the P0 from SCE, the inspectors noted the requirement for Lisega to c'edicate any items purchased as commercial grade as defined in 1" CFR Part 21 using methods similar to those in Electric Power Research Institute (EPRI) NP-5652, " Guidelines for the Utilization of Comercial Gra;e items in Nuclear Safety Related Application (NCIG-07)," as erdorsed by NRC Generic Letter 89-02.
The P0 also included c statement that subtier suppliers shall not be qualified solely on the possession of an ASME Certificate of Authorization or a QSC, but on the basis that the subtier supplier: sre effectively implementing the quality program surveyed by ASME.
Lisega was in noncompliance with these P0 requirements, in a clarification letter to SCE concerning the dedication of comercial grade items, Lisega stated that a separate dedication process is not required or dcfined within the quality program because no comercial grade items are proctred or furnished by Lisega.
This is not considered an accurate reflection of the procurement process being used by Lisega for the manufacture of safety-related Appendix B components. Since Lisega des not impose Appendix B o~ 'n CFR Part 21 on their subcontractors; does not audit all suppliers inc.uded on the List of Authorized Suppliers; and does not include on this list all TUV 1253/1 suppliers being used, Lisega appears to be procuring commercial grade items and should therefore include provisions in the quality program for dedication of these items. Currently, there are no provisions in the Lisega QA Manual and the QA Program implementing Procedural Guidelines (VQSP's) for dedicating items purchased by Lisega as commercial grade that are included as part of the pipe support components that are sold as safety-related 10 CFR Part 50, Appendix B items. These QA program and procedure deficiencies are identified as Nonconformance (92-01-02). 1
-112-l
3.5 Control and Anoroval of Subcontractors During the review of Lisega's processing of utility purchase orders, the inspectors identified numerous vendor subcontractors who supplied material certifications for material or the piece parts used in the manufacture of the pipe support components and snubbers supplied to U.S. utilities. While reviewing Revision R of the Lisega List of Authorized Suppliers, the insrectors determined that all these suppliers were not on this list >,nd all of these suppliers' QA programs and their implementation had not b% n audited by Lisega.
The inspectors reviewed QA Manual Section 5, Revision E. 'r.aterial Procurement Procedures," and VQSP Procedural Guideline No. 37, Revisi s 0, *Haterial Procurement, Goods Receiving Control, and A) proval of Sul.ntractors," to determine the QA program requirements for the approval e subcontractors and the basis for placing vendors on the list of Authorized 'sppl i e r s. The QA manual states that suppliers not holding ASME QSCs shall veyed and audited by Lisega using the checklist QSF 16, "Examinatit
'ubcontractors.'
However, the QA Manual allows suppliers who hold ASME QSCs
- 6. w placed on the List of Authorized Suppliers, P0s processed, and material certifications and certificates of compliance accepted, without any implementation audit being perf ormed of these suppliers. During discussions with Lisega personnel, the inspectors learned that the same process is followed for suppliers who are listed on the German government Register of Approved Material Manufacturers (TUV 1253/1). The inspectors also identified that the majority of suppliers listed on the TOV 1253/1 register who had supplied material certifications to Lisega, were not included on the Lisega List of Authorized Suppliers. The deficiencies identified during this review are identified as Nonconformance (92-01-03),
4 4 PERSONNEL CONTACTED Liseaa GebH G. Liesegang, Chairman H, Hardtke, President H. Bardenhagen, Quality Assurance Manager W. Wagner, Purchasing Manager
-113-
[
- c, UNITED STATES
[' %/?i.
NUCLE AR REGULATORY COMMISSION g/
p WASHING ton. D. C. 20555 N - [el
~-
November 12, 1992 Docket Nos. 50-352 and 50-353 Mr. D. M. Smith Senior Vice President-Nuclear Philadelphia Electric Company Nuclear Group Headquarters Correspondence Control task P.O. Box 195 Wayne, Pennsylvania 19087-0195
Dear Mr. Smith:
SUBJECT:
INSPECTION OF THE PROCUREMENT AND COMMERCIAL GRADE DEDICATION PROGRAMS AT THE LIMERICK GENERATING STATION (INSPECTION REPORT N05. 50-352/92-201 AND 50-353/92-201)
This letter transmits the report of the inspection conducted June 22 through 26, 1992, at the Limerick Generating Station, Units 1 and 2, by R. P. McIntyre, R. C. Wilson, L. L. Campbell, and B. H. Rogers of the U.S.
Nuclear Regulatory Commission's (NRC's) Vendor _ Inspection Branch and by 9 L.
Caphton and E. J. Benner of NRC Region I.
The inspection was related to activities at the plant site authorized by NRC licenses NPF-39 and NPF-85. At the conclusion of the inspection, we discussed our findings with J. F.
O'Rourke, Project Manager, Limerick Generating Station (LGS), and the members of your staff identified in Section 6 of the enclosed inspection report.
The inspection was conducted to review the implementation of the Philadelphia-Electric Company (PEco) program for the procurement and dedication of commercial grade items (CGIs) used in safety-related applications at LGS. The inspection team identified a major program strength in that LGS made use of the PEco corporate laboratories in Valley-Forge, Pennsylvania, which have extensive capabilities for dedication and acceptance testing activities.
Several additional strengths were identified in the PECo procurement and commercial grade dedication program, including LGS's engineering involvement in such areas as the identification of part safety classification, safety functions, failure modes and effects, critical characteristic identification and verification, and performance of replacement part equivalency evaluations.
The PEco policy of purchasing items in accordance with Appendix B to 10 CFR Part 50, when available, is another strength. This practice reduces the number of items that need to be purchased as commercial grade and dedicated for use in safety-related applications.
Finally, the December 1991 assessment by PEco Nuclear Quality Assurance of LGS's commercial grade dedication program against the weaknesses identified by the NRC during the assessments at other utilities led directly to improvements in the dedication program and its implementing procedures.
The results of the inspection also indicate that some weaknesses existed within PEco's program for the dedication of CGIs procured for use in safety-
-114-
Mr. D.M. Smith.
related applications. These weaknesses contributed to the specific findings <
that were identified during the review of recently_ completed dedication' files for CGis installed or available for installation in safety-related plant sys-tems.
Examples of these findings include (1) PEco did not fully specify sampling requirements in order to obtain assurance of batch or lot homogeneity -
from the supplier, (2) did not have provisions to fully specify the ' sample size to be tested during receipt and acceptance testing, and (3) did not
~
perform and document.the environmental and dynamic (seismic) evaluations in certain instances.
This was the last of a series of five pilot inspections being conducted to evaluate the implementation of licensee CGI procurement and dedication programs and to finalize an inspection procedure for future ins?ections of this type. The first three inspections identified numerous wea(nesses in-licensee CGI dedication programs and their implementation. These weaknesses were classified as deficiencies with the potential for subsequent enforcement actions by the appropriate NRC regional office. During this same period, the NRC also received significant feedback from several utilities and the Nuclear Management and Resources Council (NUMARC) questioriing the regulatory basis for the CGI dedication guidance contained in Generic l.etters 89-02 and 91-05, as well as specific interpretations of that guidance by the NRC inspection teams.
To properly respond to these concerns and to_ resolve major. points of contention concerning the CGI dedication process, NRC will be holding a number of public meetings with utility representatives and conducting a public workshop which will address issues concerning the existing dedication guidance and specific interpretations of that guidance, in view of these pending actions, the specific observations concerning your CGI dedication program and its implementation are classified as " findings."
Because of the large number of such findings identified during these pilot inspections and the apparent need for additional clarification of HRC's -
expectations in this area, no enforcement action will be taken against these-findings.
It should be_ recognized, however, that some of these findings may require corrective action based on: (1) their potential impact on the performance of specific equipment, or (2) because of noncompliance with your.
commitments to specific industry standards or regulatory guides endorsing such standards. We expect that you will evaluate the ir.:pection findings in view of these considerations and take appropriate corrective action as necessary.
In accordance with 10 CFR 2.790 of the NRC's " Rules of Practice," a copy of this letter and its enclosure will be placed in the NRC's Public Document Room.
-115-
Mr. D.H. Smith.Should-you have any questions concerning this inspection, we will be' pleased -
to discuss them with you..Thank you for your cooperation in this inspection.
Sincerely, b
.b i,, Dire r
Division of Reactor Pr ects.1/II Office of Nuclear Reactor Regulation-
Enclosure:
Inspection Report 50-352/92-201 and 50-353/92-201 cc w/ enclosure:
See next page
-116-
m t
4 w
Mr. D. M. Smith Limerick Generating Stationt Philadelphia Electric Company:
Units:1 1 2-CC:
J. W.-Durham, Sr., Esquire Mr. William P. Dornsife, Director-Sr. V.P. & General Counsel -
Bureau ~of Radiation Protection Philadelphia Electric Company PA Dept.'of Environmental Resources t
2301 Market Street P.0i Box 2063 Philadelphia, Pennsylvania '19101 Harrisburg,1 Pennsylvania 117120 Mr. Rod Krich 52A-5 Mr. James A.' Muntz Philadelphia Electric-Company Superintendent-Technical 955 Chesterbrook Boulevard Limerick Generating Station Wayne,-Pennsylvania 19087-5691 P.O. Box A-Mr. David R. Helwig,' Vice President Limerick Generating _ Station Mr. Gil J. Madsen-Post Office Box A Regulatory Engineer Sanatoga,-Pennsylvania 19464 Limerick Generating Station P.O. Box A-Regional Administrator
_Sanatoga, Pennsylvania 19464 U.S. Nuclear Regulatory Comission Region I Library-475 Allendale Road U.S. Nuclear Regulatory Comission King of Prussia, PA 19406
. Region 1-475 Allendalc Road L
Hr. Thomas Kenny King of Prussia, PA - -
19406 Senior Resident inspector U.S. Nuclear Regulatory Commission
'Mr. George A. Hunger P.O. Box 596 Project Manager-Pottstown, Pennsylvania 19464
. Limerick Generating Station-P.O. Box-A Mr. Richard W. Dubiel Sanatoga,~ Pennsylvania 19464 Superintendant - Services Limerick Generating Station Mr.. Larry Hopkins P.O. Box A Superintendent-Operations-Sanatoga, Pennsylvania 19464 Limerick. Generating-Station-P.O. Box AL Mr. John Doering Sanatoga,-Pennsylvania 19464.
Plant Manager -
Limerick Generating Station P.O. Box A Sanatoga, Pennsylvania 19464
-117-
=
U.S.-NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION DIVISION OF REACTOR INSPECTION AND LICENSEE PERFORMANCE Report Nos.:
50-352/92-201 and 50-353/92-201 Docket Nos.:
50-352 and 50-353 License Nos.:
Philadelphia Electric Company Correspondence Control Desk P.O. Box 195 Wayne, Pa 19087-0195 Facility Name:
Limerick Generating Station, Units 1 and 2 Inspection at:
Limerick, Pennsylvania inspection Conducted:
June 22 through 26, 1992 I N N Prepared by:
8 Date Richard P. McIntyre, Team Leadei Vendor Inspection Branch (VIB)
Inspection Team:
R. C. Wilson, Senior Reactor Engineer, VIB L. L. Campbell, Reactor Engineer, VIB B. H. Rogers, Reactor Engineer, VIB D. L. Caphton, Senior Reactor Engineer, Region I E. J. Benner, Reactor Engineer, Region I Reviewed by:
- c4 M
//
Leif J(/ipirrholm, Chief tath Vendor Inspector Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation b b3 ll/r/92.
Approved by:
j e Char WI E.
si, Director Dath I Div Wien eactor Inspection
)
and Licensee Performance Office of Nuclear React r Regulation
-118-
e TABLE OF CONTENTS-t Paae EXECUTIVE
SUMMARY
..................................................... i' 1
INTRODUCTION.....................................................!
2 COMMERCIAL GRADE DEDICATION PROGRAM REVIEW....................., 2 2.1 Procedures Review.......................................... 2 2.1.1 Engineering Evaluation Summary (EES) Process......... 3 2.1.2 Saf ety Cl as si fi cat ion................................
3 -
2.1.3 Environmental / Dynamic-(Seismic).
- Q u al i f i c at i on.......................................
4 2.1.4 Repl acement Part Equivalency Evaluation..............
5 2.1.5 Procurement Level Determination...................... 6 2.1.6 Dedication of CGIs.............................'..... 6 2.2 Commerci al Grade Supplier Surveys..........................
7 2.2.1 Third-Party Commercial Grade Surveys..................
9
- 2. 2. 2 Source Veri fications.................................. 9 2.3 Sampling..................................................., 9' 2.3.1 B a s i s fo r S ampl i ng.................................. 9 2.3.2 S ampl e Si ze Sel e ct i on...............................
10 2.4 PECo Corporate Laboratories Capabilities....................
11 2.5 Receipt Inspection..........................................
11 l
2.6 Detection of Fraudul ent Materi al s............................
12
- 2. 7 Trend i ng o f Suppl i e rs........................................ 12 -
3 DEDICATION PACKAGE ~ REVIEW FINDINGS..............................
13 4
NESD SYSTEM UPGRADE DEDICATION ACTIVITIES.......................
21 5
PROCUREMENT AND DEDICATION TRAINING.............................
22 6
EXIT MEETING....................................................
24
-119-
EXECUTIVE SUMMA'"
from June 22 through 26, 1992, representativt-the U.S. Nuclear Regulatory Commission's (NRC's) Vendor Inspection Brancl. s vlB) and Region I conducted an inspection of Philadelphia Electric Company's (PECo's) activities related to the procurement and dedication of commercial grade itt, (CGis) used in safety-related applications at the Limerick Generating Station, Units 1 and 2 (LGS).
The inspection team reviewed PECo's procurement and dedication program to assess its compliance with the quality assurance (QA) requirements of Appendix B to Part 50 of Title 10 of the Code of Federal Regulations (10 CFR Part 50).
On August 24, 1990, the NRC staff fors nded to the Commission SECY-90-304, "NUMARC Initiatives on Procurement," in which the staff reported the status of the Nuclear Management and Resources Council's (NUMARC's) initiatives on general procurement practices. Procurement initiatives as described in NUMARC 90-13 " Nuclear Procurement Program Improvemeats," dated October 1990, committed licensees to assess their procurement programs and take specific action to strengthen inadequate programs. The industry initiative on the dedication of CGIs, which was to be accomplished by January 1, 1990, stated that licensee programs should meet the intent of the guidance provided in the Electric Power Research Institute's (EPRI's) Final Report, NP-5652, " Guideline for the Utilization of Commercial Grade items in Nuclear Safety Related Applications (NCIG-07)," dated June 1988. The staff also stated in SECY-90-304 that it would conduct assessments at selected sites to review the licensees' implementation of improved procurement and commercial grade dedication programs, assess improvements made in the areas covered by the NUMARC initiatives, and report the results of those assessments to the Commission. From February to July 1991, the VIB conducted eight assessments of selected licensees to determine the current status of activities to improve the procurement programs re'ated to industry initiatives and NRC requirements.
On September 16, 1991, the NRC staff forwarded to the Commission SECY-91-291,
" Status of NRC's Procurr..iant Assessments and Resumption of Programmatic Inspection Activity," in which the staff reported on the results of its assessments and noted that it was resuming inspection and enforcement activities.
The NRC conducted this inspection, the fifth of the headquarters pilot inspections in this area, to review PECo's procurement and dedication program and its implementation since January 1, 1990, (the effective date of the NUMARC ir.itiative on dedication of CGis). The inspection included a review of procedures and representative records, including approximately 40 procurement and dedication packages for mechanical and electrical CGis; interviews with PECo staff, including senior management and LGS site personnel; and observations by the inspection team members. The inspection team also held meetings with PECo's management to discuss relevant aspects of commercial grade dedication and to discuss areas requiring additional information. The inspection team findings were discussed with PECo's representatives and senior management at the exit meeting held June 26, 1992.
-j_
-120-e
The inspection team identified several strengths in PEco's procurement -program-and its implementation. A major program strength was LGS's use of the PECo corporate laboratories in Valley-Forge, Pennsylvania. The NRC. inspectors toured the facility and found it to have extensive capabilities to test in mechanical, electrical, chemical, and metallurgical areas, and the' ability to perform numerous calibrations and special-tests. These capabilities could provide a substantial basis and support for the dedication and acceptance testing activities.
Several additional strengths were identified in the PECo procurement and commercial grade dedication program, theluding LGS's engineering involvement in such areas as the identification of part safety classification, safety functions, failure modes and effects, critical characteristic identification and verification, and performance of replacement part equivalency evaluations.
The PECo policy of purchasing items in accordance with Appendix B to 10 CFR Part 50, when available, is another strength. This practice reduces the number of items that need to be purchased as commercial grade and dedicated for use in safety-related applications, Finally, the December 1991 assessment by PEco Nuclear Quality Atsurance of LGS's commercial grade dedication program against the weaknesses identified by the NRC during the 1991 assessments at other utilities led directly to improvements in the dedication program and its implementing procedures.
The inspection team also identified some weaknesses in the generic procurement program and in PEco's actions to implement it. These weaknesses contributed to the specific findings described in Sections 2 and 3 of this report.
Examples of these findings included the lack of specific procedural requirements for determining when it is appropriate to use sampling. As a result, the procurement engineering group (PEG) did not fully specify sampling requirements or the sample size on the engineering evaluation summary (EES) document, in order to obtain assurance of batch or lot homogeneity from the supplier prior to testing. These findings often contributed to the absence of a clearly defined basis to support sampling and the sample size selected f_or the CGIs being dedicated. This is a concern because LGS-predominately uses receipt inspection and acceptance testing at the PEco corporate laboratories to verify the critical characteristics identified on the EES. Weaknesses in this area had recently been identified in a PECo internal audit.
Another finding in the LGS dedication process concerned the lack of specific procedural requirements for performing and documenting how the environmental qualification (EQ) and dynamic (seismic) requirements were being met. This resulted in numerous EES's that did not document the basis for statements that indicated EQ and seismic requirements for the replacement CGI had been met, and, therefore seismically qualified.
The team also determined that a lack of procedural guidance for documenting the acceptance. criteria for the verification methods listed on the EES for the specific critical characteristics identified, resulted in the use of certain discrepant test results from the PECo corporate laboratories. These test results were ultimately accepted by the corporate laboratories-and LGS QC on
-ii-
-121-i m
the quality contrdiLinspection report (QCIR) and were used as part of-the dedicaticn_ basis-for the sp:cific CGI.
LGS Administration Procedure'A-129;2, " Classification and Engineering.
Evaluation of Items _and Off-Site Services," addresses requirements for the
- different procurement levels and, as written,. procurement-level QV permits procurement'of nuclear-unique basic components from other NRC licensees and =
vendors without contractually invoking the requirements-of 10 CFR-Part 21 in the purchase documents. The team considered this a weakness in-the PEco procurement program. PEG informed the team that this practice would be corrected and that PECo Nuclear Group Procedure P-C-2, " Evaluation and Control For Procurement of Plant Items Including Design Equivalent ~ Changes,"
(not yet issued as _of. the time of the inspection), would also reflect proper 10 CFR Part 21 requirements.
o 1
I l
5
-lii-
-122-1
1 INTRODUCTION During this headquarters pilot inspection, the team reviewed the Philadelphia Electric Company (PECo) program and its implementation for the procurement of commercial grade items (CGIs) used in safety-related applications at the Limerick Generating Station (LGS).- The team also reviewed the PECo program -
and its implenentation for determination or verification of suitability of those CGIs for their intended or approved safety-related applications, a process referred to as " dedication."
Part 21 of Title 10 of the Code of Federal Regulations (10 CFR Part 21) defines dedication as the point at which an item or service becomes a " basic component," that is, essentially an item (or service) with safety-related functions. However, the 10 CFR Part 21 definition of a CGI (Section 21.3(a)(4)(a-1)), distinguishes CGls from items procured as basic components. The regulation, then, allows the procurement of items that are to become basic components that meet the definition of CGis without invoking 10 CFR Part 21 in the procurement documents.
When CGis are procured for safety-related service, their procurement and dedication constitute activities affecting quality; therefore, these activities must be controlled in accordance with the requirements of Appendix B to 10 CFR Part 50.
In particular, Criterion Ill, " Design Control,"
and Criterion Vil, " Control of Purchased Haterial, Equipment, and Services "
of Appendix B are most pertinent to the procurement and dedication of CGIs.
Therefore, the team reviewed the PEco program governing these activities and-the implementation of that progre'n for compliance with these and other applicable Appendix B criteria and with the requirements of 10 CFR Part 21.
The NRC has provided further guidance to the requirements of Appendix B as they pertain to the procurement and dedication of CGls in NRC Generic Letter (GL) 89-02, " Actions to Impr've the Detection of Counterfeit and Fraudulently o
Marketed Products," dated March 21, 1989, and GL 91-05, " Licensee Commercial-Grade Procurement and Dedication Programs," dated April 9, 1991.
Therefore, the PECo CGI procurement and dedication program and its iaplementation also were evaluated for consistency with the guidance and NRC staff positions promulgated in these generic letters.
Finally, with respect to procurement in general, including procurement and dedication of CGIs, PEco has committed to various industry standards and other publications (as endorsed or conditionally endorsed by NRC regulatory guides,
']
NUREGs, and GLs); as stated in the PEco quality assurance (QA) program description; as contained or referenced in the PEco Updated Final Safety Analysis Report for LGS, and as expressed for the industry by the Nuclear Man-agement and Resources Council (NUMARC) in the NUMARC initiative on the dedica-tion of CGis as part of NUMARC 90-13, " Nuclear Procurement Program improvements." -123-i
w 2 COMMERCIALLGRADE.0EDICATION PROGRAM REVIEW 2.1-P_tqcedures-Review.
The PECo program for the procurement and dedicationiof:CGIs for safety-related applications at LGS is implemented through. a hierarchy of site procedures --
beginning with LGS Administrative Procedure, A-129, " Procurement of Items;and, Services," Revision 2, dated July 31,-1991.' Thi_s procedure describes the-procurement process _for items and services,-responsibilities and, authorities for site and corporate organizations, and organizational' interfaces..
- Procedure-A-129lalso contains various tern.s and definitions,- including the-following that are referred.to in several LGS proc'edures and in this inspection report.
Critical Characteristics are identifiable-and' measurable attributes /
variables of an item which, once verified,: provide assurance that'the item will satisfactorily perform its intended safety function and that the item has been received as specified..
Procurement levels are assigned to procurement documents'_after review /
evaluation to establish an easily recognizable way_ of identifying general procurement requirements. Alpha _ symbols identify each -level as follows:
0-Soecification (QS) - refers to Q-listed items and services.
procured from a-vendor listed on the PECo Evaluated-Vendors List (EVL). The vendor's quality program meets the requirements of_ ANSI N45.2 or -10 CFR Part 50, Appendix B, as applicable for the items or service and the requirements of_10 CFR Part 213re invoked on the vendor.
0-Verification (QV) - refers to Q-listed items and services that may; be procured from a non-EVL approved vendor' via implementation of PEco's approved QA program.
10 CFR Part'21 is not invoked on-the-vendor but is the responsibility of PECo.
Commercial Grade-(QD) - refers-to-items which meet the commercial _
grade item definition of 10 CFR Part 21 and are procured and designated _for Q-List applications upon completion of_ an appropriate:
dedication plan.
10 CFR Part 21 is not invoked on the vendor.but:
is the responsibility of PECo.
Non-Safety-Related. With Reauirements- (NX) - refers to-non-Q-listed items.and services procured with specific: quality' assurance and/or-code requirements.
10 CFR Part 21 does not apply.
Non-Safety-Related (NN) - refers to items and services procured withi no quality assurance requirements.
Administrative Procedure A-129.1, " Preparation and Processing Preliminary Purchase Requisitions and Prel %inary Requisition-Work Sheets," Revision 2, dated August 28, 1990, provides requirements for the initiation and processing J i
-124-
of preliminary purchase requisitions (PPRs)-and preliminary. requisition-work sheets-(PR-WSs) to obtain-safety-related and non-safety related items-and ser-vices, respectively, for usezat-LGS. The-LGS procurement _ engineering-group
-(PEG) processes PPRs and-PR-WSs_-(for of_fsite services) to ensure that-requirements and criteria such as the-safety classification,-procurement level, procurement requirements, dedication, equivalency evaluation, receipt--
inspection and test, shelf-life, and storage arel properly assigned in=
accordance with Administrative Procedure A-129.2, " Classification and Engineerirg Evaluation of_ Items and Off-Site. Services," Revision 3, dated July 31, 1991.
After PEG has processed a PPR, it forwards the PPR to LGS-Stores-to have PPR requirements transcribed onto a purchase requisition (PR); similarly, PEG forwards a PR-WS to the LGS Business Unit to have PR-WS requirements transcribed onto a PR. The completed prs along with any supporting documents are forwarded to the Limerick Quality Division, Quality Support Section (QSS)-
for review and approval of the quality aspects for-procurement levels QS, QV, QD, and NX. QSS then forwards the approved PR to the LGS Purchasing Department for the preparation of the purchase order (P0).
2.1.1 Engineering Evaluation Summary (EES) Process A PEG evaluator performs an EES to dedicate CGis and offsite services for safety-related applications at LGS in accordance with Administrative Procedure-A-129.2.
The EES includes the determination of safety classification, procurement level, procurement requirements, envir:enmental qualification (EQ) and dynamic (seismic) requirements, replacement part equivalency, alternate replacement items, safety functions, failure modes and effects, critical.
characteristics, verificatica methods and acceptance criteria, receipt inspection, and testing and storage requirements. The principal elements of -
the EES process are described below.
2.1.2 Safety Classification The PEG evaluator determines if su'ocomponents and parts of safety-related components, systems, and structures, (CSSs) are required for the CSSs to perform their safety-related functions. The PEG evaluator records information on the parent CSS (such as manufacturer, specification, applicable drawings, technical manuals, and industry standards, applicable UFSAR sections, and other applicable references used in determining the item's safety classification) on the EES. The evaluator then determines and records the item's function and whether the function is active or passive with regard to-the parent CSS's function..The evaluator then' identifies and documents the critical characteristics of the item that are essential for it to perform any functions that are identified as safety-functions. Using this information, the evaluator identifies and records the potential failure modes of the item that could result in failure to nerform the required safety functions.
Proce-dure A-129.2 requires that each failure mode be evaluated for adverse effects on the safety functions of the parent CSS.
If the' evaluator finds that any failure mode could cause the loss of a safety function, the item is classified as nuclear safety-related (NSR). -125-
The teaa discuss:d t:lth PE'G the fact that Pr:c: dure A-129.2, Revision 3, requires the identification of e:sential characteristics to determine.the item's failure modes from which the critical characteristics would be identified. The team asked PEG to explain the difference between these two types of characteristics. PEG indicated that the essential and critical characteristics were the same an( that the classification process sequence
- would be reviewed and Procedure A-129.2 would be revised, if appropriate.
Additional discussions with PEG revealed that items such as gaskets, 0-rings, lubricants, and valve packing are not generically classified for use at LGS, but are evaluated and are subject to technical and quality requirements applicable for their intended end use.
The team performed a limited review of the recent EES packages listed below and found the classification of these parts as non-safety-related acceptable.
114-71570, fluid for the diesel generator lube oil vacuum pressure manometer 114-76468, shutter assembly for the 4-kV switchgear circuit breaker cubicles 114-76469, screw for the 4-kV switchgear circuit breaker cubicle shutter retainer 114-77992, pin for the refuel platform safety brake The team also performed a limited review of the classification of an annular key in valve HV-57-Il4 (a 24-inch butterfly valve used to exhaust drywell gases to standby gas treatment and to isolate the drywell) and a dowel pin in valve HV-57-109 (a 6-inch butterfly valve used to supply nitrogen to the drywell for inerting and to isolate the drywell). However, there was insufficient information available at LGS for the team to determine if these valve parts had been properly classified as non-safety-related. The classification of the annular key and dowel pin was performed-in 1987 as part of PEco's construction and startup contractor's turn over of spare parts.
2.1.3 Environmental / Dynamic (Seismic) Qualification The PEG evaluator reviews the appitcation of an item to determine if EQ or seismic qualification is required for the replacement item. A PEG EQ engineer then reviews the item being procured to determine if the item meets the requirements specified in the LGS EQ report. This review is coordinated with the Nuclear Engineering and Services Department (NESD) and is documented on the EES form. The assigned evaluator performs a similar activity for the item's seismic qualification.
If an item differs from the EQ report or seismic analysis, PEG is required to complete a replacement part equivalency evaluation (RPE) to evaluate the difference.
The team reviewed several EES packages in order to determine if EQ and seismic requirements had been properly addressed. This review identified that the majority of the EES packages did not document the basis for statements that -126-
-indicat:d EQ and _ seismic. requirements had been met. This itplementation-weakness was the result of Procedure A-129.2 not providing specifict Trenuirements for documenting how EQ and seismic requirements were being met.
^
The following are examples of:EES packages that did-not fully document the conclusion that the' original seismic qualification had been maintained.
EES 114-24964, Wilmar' relays, EES 114-94556, Gould Shawmut fuses,
=
EES 114-76265, Sigma relays, EES 114-94764,- Weidmuller terminal blocks,
EES 114-77164,- Lambda power supplies, and EES 114-76555, level indicating gauges.
2.1.4 Replacement Part Equivalency Evaluation PEG _ initiates an RPE to confirm that an alternate replacement item (not identical to the original) will satisfactorily perform the same design function of the part being replaced.
PEG prepares and processes the.RPE and includes an engineering change request (ECR) prepared ~in accordance with' Nuclear Group Administrative-Procedure (NGAP) NA-10P005, " Design Equivalert Change Control," Revision 0, dated April 15 1991. NESD is required to review RPEs which address EQ and seismic differences, and any replacements-in the main control room, alternate shutdown panel, remote shutdown panel, or emergency shutdown panel.
The ECR form is used to describe a proposed change, provide the engineering disposition of the change, and identify all affected engineering-documents (such as drawings, vendor documents, and calculations) needed to support the change.
There are two types of ECRs: (1) design review ECRs which involve a change in the replacement item's -form, fit, or function and -typically involve alternate replacement items that require installation _ details and/or involve
' multiple component replacements, and (2) document change only ECRs that involve changes in the' replacement item's form, fit, or function that PEG evaluates as equivalent. NESD must review design review ECRs before the affected system is returned to service; however, NESD may review document change only ECRs following return to service.
The team discussed an observation with PEG and NESD that Procedure NA-10P005 did not specify timeliness requirements for the formal updating of design documents associated with document change only ECRs. NESD personnel informed the team that there were approximately 70 document change only ECRs having one
.or more design documents requiring updating to reflect the installed as-built configuration. Most of these ECRs were for changes to systems that had-been declared operable 11/2 to 3 months ago. None of the changes affected the_
operational schematics used by the plant operators.
< -127-1
As part of the RPE process, PEG also performs a 10 CFR 50.59 safety determination on the replacement it:a.
If this deteraination indicates that a 50.59 safety evaluation is required, PEG forwards the RPE to the system engineer for processing as a modification.
2.1.5 Procurement level Determination As discussed earlier, items that are basic components and procured from vendors listed on the PECo EVL. with the reporting requirements of 10 CfR Part 21 applied to the vendor, are designated to be purchased as procurement level 05.
Procurement level QV is generally applied when the vendor is the original equiament manufacturer (OEM) and/or the item is nuclear unique and the OEM or olier vendors may not have a QA program that meets Appendix B to 10 CFR Part 50 or the supplier will not accept the reporting requirements of 10 CFR Part 21.
Procedure A-129.2 addresses requirements for procurement level QV and, as written, permits procurement of nuclear-unique basic components from other NRC licensees and vendors without contractually invoking the requirements of 10 CFR Part 21 in the purchase docur.is.
This is contrary to the requirements of 10 CFR Part 21 which specifies the notificatior, responsibilities of suppliers of basic compoacris. The team consioered th a weakness in the PEco procurement program.
PEG informed the team that this practice would be corrected and that PECo Nuclear Group Procedure, P-C-2 (not issued yet), would also reflect proper 10 CFR Part 21 requirements.
Procurement level QD is assinned to items that meet the 10 CFR 21.3 definition of a CGi and that will be dailcated for use ih safety-related applications at LGS.
2.1.6 Dedication of CGis items purchased as procurement level QD and dedicated for use in safety-related applis ons are done so by PEG through documentation of the following activities as srt of the EES process:
Critical characteristics and acceptance criteris are identified, which, when verified, will provide reasonable a.;surance that the item will per-form its safety-related function and that the item received is the item specified.
Jeceipt inspection, acceptance testing / examination, vendor survey, and source inspection verification methods are identified to verify the listed critical characteristics.
Acceptance criteria are identified.
Commercial grade manufacturing controls and processes are selected to be surveyed to validrie certificates of compliance (CoC) used to verify critical characteristics and requests for surveys are initiated. The team noted that Procedure A-129.2 provided no requirements for determining which critical characteristics are being sufficiently controlled by suppliers on the EVL and considered this to be a program -126-
seakness. PEG informed the teaa that tha procedure would be revised to require a review of the commercial grade survey of the supplier to determine if a supplier's CoC can be used as the basis to verify a critical characteristic.
The EVL is reviewed and updated, as appropriate, to acctpt vendor documentation as part of the dedication process.
Procurement documentation requirements used as part of the dedication process are identified.
Any post-installation testing necessary to complete dedication of the item is identified.
Shelf-life storage and preventive maintenance requirements are identified.
An independent reviewer reviews the completed EES, verifies tne adequacy
+
of the evaluation as documented by the originator, and ensures that system interfaces have been addressed. This independent review also scrves as the independent design verification of the ECR design change.
2.2 Commercial Grade Supolier Surveys Procedure A-129.2 provided for procurement of CGis intended for dedication using procurement level QD.
Section 7.2.4.5 indicated that a vendor survey was an allowable method for verifying critical characteristics. The Vendor
(
Evaluation Section of the Corporate Nuclear Quality Division conducts those audits and commercial grade surveys for both PECo nuclear sites, Limerick and Pea.h Bottom.
PECo Nuclear Quality Assurance (NQA) consists of the Limerick Quality Division, the Peach Bottom Quality Divisicn and the Corporate Quality Division, which is located in the Chesterbrook, Pennsylvania corporate office.
Vendor Evaluation Section Guideline No. 3, " Vendor Commercial Grade Survey Gu.delines," Revision 2, dated May 8, 1992, was reviewed. The team found no direct procedural connection from the HQA guideline to any higher level procedure, whether unique to NQA or LGS, or common. This inconsistency contributed to the absence of a procedural requirem:nt for FEG engineers to support Chesterbrook NQA personnel in planning, performing, or evaluating commercial grade vendor surveys. Guideline No. 3 specified that NQA personnel should consult with PEG engineers and invite their participation, but the guidelir.e was not applicable to PEG. The team noted that a PEG engineer had participated in a commercial grade survey and that PEG engineers had provided input for surveys.
The team also reviewed NQA Vendor Evaluation Section Guideline No. 8,
" Evaluated Vendors List Guideline," Revision 2, May 19,1992. The purpose of th!s guideline was to implement the NQA Plan and Common Nuclear Procedure P-C-9 " Evaluated Vendors ist."
PECo personnel stated that P-C-9 superseded NQA-19, same title, in January 1992. Guideline No. 8 referenced Guideline No. 3, but not vice-versa.
Guideline No. 8 defined the " quality considerations for evaluating a vendor for the supply of commercial grade material," based on a commercial grade survey.
~7-
-129-d
l Section 7.2.3 of Guideline No. 8 required surveys of both a distributor and a parts manufacturer if a distributor was to be itsted. This procedural requirement, and PEco's limited use of distributors for commercial grade items, was considered a strength even though the requirement was not incorporated into procedures until this revision. Section 7.6.9 of Guideline No. 8 initiated a requirement to perform surveys at least every 18 months, although the inspector noted that a resurvey was just being planned for lietal Bellows Co., which was last surveyed in August 1989.
A february 1992 survey report of two Bussmann fuse locations did not fully document verification of critical characteristics. However, the team considered a 1990 survey report for another Bussmann location to be an acceptable example of EPRI Method 2 dedication by commercial grade vendor The team noted that the 1990 survey was performed by a single PECo survey.
NQA representative, whereas the newer 1992 survey was performed by a fot.r-person Nuclear Procurement Issues Committee (NUPIC) joint team led by another utility, which included representatives from both the PEG and the HQA groups of PEco.
PECo personnel stated that the 1992 report of the NUPlc survey did not reflect all the notes taken by the PEto participants.
The NRC team also reviewed a commercial grade vendor survey of Gould, Inc.,
performed in June 1990, by PEco for 42 listed fuse types.
Like the 1992 Bussmann survey, the Gould survey did not fully document the verification of the identified critical characteristics. As an example, the " Critical Characteristics Verification Plan" portion of the PEG EES required verifying the current time limit by meeting unspecified clearance time requirements-(which would vary with fuse ty)e and rating) at 135, 150, 200, and 300 percent of rated current. Page 5 of tie survey report stated that "in conjunction with the survey," clear time testing was witnessed on June 14, 1990; the date of the survey was June 19, 1990. Page 5 stated that "two samples of each type and rating were tested at 200 percent full voltage." The context indicated that three types were tested, but the number of ratings involved was not stated. The report further stated that the TR400 fuses were rejected and an additional three samples were tested without exception.
The team found no further documentation of current time limit testing in the Gould survey report. The basis for representing 42 fuse types by testing 3 types was not given. The basis for testing two fuses of u,r.h type, or for retesting three of one type after test results were rejected, was not given.
Acceptable time limits were not addressed, nor was the basis for testing at only one of the four currents specified for verifying the critical characteristic. The test method was not addressed. The team also noted several other instances where the basis for verifying critical characteristics was not documented. Despite these omissions, page 5 of the survey report stated that "the inspections and tests are adequate to control the critical characteristics."
The NRC inspectors briefly reviewed two other commercial grade surveys:
Parker Hannifan Corp., January 1991 (0 rings); and Leeds & Northrup Co.,
September 1991 (recorders and parts). Both surveys were performed by NUPlc, and the reports appeared to be relatively free from the concerns noted for the Gould and 1992 Bussmann surveys. -130-
2.2.1 Third-Party Commercial Grade Surveys Guideline No. 8, Revision 2, restricts the use of third-party commercial grade surveys to NUPlc surveys. Upon notification of a NUPlc survey, NQA solicits engineering input such as critical characteristics for forwarding to NUPic.
NQA will keep the NUPlc survey report in a suspense file until PEG notifies them that the survey is acceptable. NQA will then add the vendor to the EVL or update an existing listing.
PLCo used third-party (NUPIC) surveys for four of the seven vendors evaluated as part of the EPRI Method 2 dedication process.
in each case PEco provided NUPlc information such as critical characteristics before the survey was performed.
in one case PEC provided two team members as well.
2.2.2 Source Verifications PEco issued Guideline No. 10 " Vendor Source Verification Guideline,"
Revision 0, June 25, 1992, during the inspection. The NRC inspectors did not review the new guideline. PECo personnel stated that they had previously performed only five commercial grade source verifications.
In each case they based the verification on Procedure HQA-20. " Vendor Audits," Revision 2 dated April 22, 1991. This is another example of the 3rocedural hierarchy concern noted above. Although Procedure HQA-20 is a higier ranking document than Guideline No.10, it is not clear that it only applies to safety-related.
Appendix B verifications and that the guideline covers commercial grade verifications.
The inspectors reviewed one of the five source verifications (Lechler, Inc.)
in conjunction with a spray nozzle dedication package discussed in Section 3.
The source verification was acceptable and the team noted that it was clearly arranged according to specific critical characteristics.
2.3 Samnlino The team noted that there were no procedural requirements or guidance in the A-129 series procedures for (1) determining when it is appropriate to use sampling (i.e., selecting other than 100 percent lot size) in the verification of critical characteristics and (2) for selecting the sample size. The use of sampling was of particular concern to the team because LGS relies heavily on the use of receipt inspection and acceptance testing to verify critical characteristics. The team considered the lack of sampling guidance a weakness in the procurement and dedication program.
2.3.1 Basis for Sampling in a discussion with PEG, the team and PEG agreed that tests and inspections may be performed using a sample plan, when appropriate, and when lot homogeneity can be demonstrated. The team noted that further clarification concerning one acceptable use of a sample plan was given on page 2 of to an NRC letter to the Nuclear Management and Resources Council, dated September 30, 1991, which stated, in part: "One approach is to (1) establish batch or lot homogeneity, particularly with respect to the control -131-i
of critical characteristics, and (2) verify through audit-er survey the basis for acce) ting certification regarding lot homogeneity." The letter also stated t1at licensees may implement alternative measures to audit or survey which may be acceptable for demonstrating lot homogeneity, provided that the-alternative measures provide confidence that the items sampled will perform their safety-related function and-the basis for the alternative measures are documented.
PEG agreed with the team that there was a lack of procedural control for verification of lot homogeneity when critical characteristics are verified on a sample basis. The team noted that, during the licensee's internal audit of the commercial grade dedication process performed from November 22 through December 6, 1991 PEco identified a need to develop a sample program document and procedure requirements for the verification of lot homogeneity.
2.3.2 Sample Size Selection Although Action item 4,-Attachment A, item B.6, of the above 1991 commercial grade dedication audit indicates that PEco procedures provide for sampling based on HIL-STD-105E, " Sampling Procedures and Tables for Inspection by Attributes," the team could not find any A-129 series procedure requirements l
instructing PEG to select a sample size based on MIL-STD-105E. The team dis ussed the selection of the sample size with several-PEG engineers who had prepared the dedication packages reviewed during the ins >ection and determined that the sample size selected was at the discretion of tie PEG engineer.
For-example, discussions with PEG engineers revealed that some of the sample sizes selected were based on (1) the criteria used for.the LGS fastener retesting program, (2) selecting 10 percent of the items, and (3) using the requirements of HIL-STD-105E.
The following is a partial listing of dedication packages reviewed that exhibited the absence of traceability to a specific lot or batch.
EES 114-75919 - Of the 100 0-rings purchased, 2 were subjected to material testing and dimensional checks.
j EES 114-81697 - Of the 50 adapters purchased, 2 were subjected to material testing.
EES 114-71882 - Of the 500 resistors purchased, 3 wrere subjected to testing.
EES 114-94556 - Of the 50 fuses purchased, 3 were subjected to testing.
EES 114-77308 - Of the 75 uninsulated wire connectors purchased, 2 were subjected to material testing.
EES 114-24694 - Of the 12 undervoltage relays purchased, 3 were subjected to testing.
The lack of measures to obtain confidence in batch uniformity and the lack of-defined sampling plans is considered a program weakness. -132-l
2.4 'PECo Corporate Labortiories Capabilities The team toured the PECo Corporate Laboratories in Valley Forge, Pennsylvania, to determine the capabilities of the laboratories to support the commercial grade dedication process.
PECo had assigned a coordinator to act as a liaison between the PEG, corporate engineers and the corporate laboratories.
The coordinator assisted in determining the validity and applicability of the tests proposed by engineering, obtained information from the laboratory personnel, and coordinated the testing.
The chemical laboratory tested fuel, lubricants, insulating oils, ion-exchange resins, gases, clastomers, polymers, and sealants. The following tests could be performed: penetration time and liquification for greases; flash point, kinematic viscosity, and rust preventing characteristics for lubricants; testing for such material contaminants as halogens and sulfur; and performance of-infrared-spectroscopy on 0-rings, gaskets, and sealants.
~
1he metallurgy laboratory could perform such tests as full metallography;and photomicrographic documentation; accelerated corrosion testing; hardness.
testing; microhardness examination;- fractography, chemical analysis of metals and corrosion products, and micro analysis using a scanning electron microscope and energy dispersive x-ray Q t W py; optical emission spectrum; carbon determination; and magne> b net h:le examination. Mechanical testing included verification of such-iteh..:s !*sile and yield strength, percentage of elongation, reduction of area, Charpy v-notch -impact andi proofload, dimensional thread examination, fatigee testing, stress rupture testing, and fracture toughness.
The calibration and special tests laboratories could establish and monitor such parameters as air velocity, dimensions, flow rates for fluids and gases, force, humidity, pressure, weight, temperature, tensile strength and proof--
loading, torque, high voltage, and high current.
The electrical calibration laboratory could establish and monitor. such parameters as voltage, amperage, resistance, capacitance, inductance, time and frequency, sound level, light intensity, optical density, distortion, and phase angle.
The tour indicated that the PEco corporate laboratories had extensive testing capabilities, which, if properly utilized, could provide a substantial basis for dedication activities.
The team considered this to be a major program strength.
2.5 Receipt Ingoection Receipt inspection is performed by quality control (QC) receipt inspectors in accordance with common nuclear procedure P-C-3, " Receipt of Items,"
Revision 0, January 9, 1992. The activities controlled by this procedure commence upon delivery of items at the warehouse receiving area and continue through final acceptance.
P-C-3 has provisions to ensure that traceability is maintained, vendor documentation is complete and correct, and that the item received is the item ordered.
Interviews with QC receipt inspectors revealed -133-
good torking knowledge of P-C-3.
When faced with a probica scenario such as a purchase order with a missing critical characteristics verification plan (CCVP), the receipt inspectors correctly identified those actions that would correct the problem.
QC receipt inspectors performed limited dimensional testing; the majority of testing was performed at the PEco Valley Forge Laboratories (vfl).
The team noted a program weakness in that it was not clear who made the final acceptability determination of vfl material and dimensional test results. vfl personnel indicated that, while they could make recommendations as to acceptability of the material and dimensional aspects of a commercial grade item, the final acceptability deter.i! nation was the responsibility of the QC receipt inspector. QC personnel indicated that, unless they were specifically directed by the CCVP to review vfl test results, the material and dimensional acceptability of the item was determined by Vfl.
In addition, QC personnel indicated that, without reference to ~a specific code or more definitive acceptance criteria in the CCVP, it would not be possible to make an acceptability determination of vfl results.
The team concluded that more definitive acceptance criteria, including final acceptability determination responsibility, was needed in the CCVP.
2.6 Detection of Fraudulent Materials Efforts to detect fraudulent materials at LGS are made primarily during receipt inspection.
The lead QC receipt inspector has received training in 4
accordance with Lecture 2960, " Fraudulent Materials Awareness," which includes discussions of appropriate NRC bulletins, generic letters, and information notices as well as specific discussions of documented industry cases of fraudulent materials.
Section 7.2.8 of P-C-3 instructs receiving personnel to check for surplus, j
rebuilt, or fraudulent parts and directs personnel to Exhibit P-C-3-4 which contains EPRI document NP-6629, Appendix C.
This appendix provides personnel with information that will help identify items that may be substandard, misrepresented, or supplied with fraudulent documentation.
Interviews with QC receipt inspectors demonstrated familiarity with the methods. The team concluded that personnel responsible for the detection of fraudulent materials were adequately trained and had properly implemented the appropriate methods.
2.7 Trendina of Suppliers Neither corporate NQA nor the Haterials Management Section were trending supplier quality at the time of this inspection.
The NQA Vendor Evaluation Section (VES) personnel perform annual audits and surveillances of suppliers listed on the EVL to decide if each supplier will be retained on the EVL. This decision is based on audit and surveillance experience, including industry experience Nuclear Plant Reliability Data System (NPRDS) experience, regulatory findings (including bulletins and information notices) and other industry alert-type information.
Following a self-assessment conducted in late 1991, VES has taken steps to improve information flow and access to VES in order to upgrade its evaluation
- -134-
and assessment of suppliers. VES plans te develop improved guidelines to-ensure consistency in the method used to-annually requalify EVL sup) liers, lho VES section superintendent stated that a 1992 goal was establisted to
- develop a system to track and trend the performance of EVL suppliers through reviews of such vendor information as vendor deviation reports, receipt inspection nonconformance rept,rts (NCRs), audits and surveillances.
NQA trending of NCRs is currently proceduralized by NQA-17, "NQA Trending,"
Revision 2, and some supplier-type issues, for example, a supplier problem identified during receipt inspection may be captured within the NCR trending.
However, to date, these have not been integrated by VES into a useful supplier -
trending program.
3 DEDICATION PACKAGE REVIEW FINDINGS To assist ihe NRC inspection team in reviewing individual dedications, PEco prepared, at the NRC's request, a number of files of dedication records compiled from diverse records, but each file pertaining to one dedication, as selected by the team from a review of the list of LGS dedication files.
PEco organized the review packages for the following-disciplines: electrical, instrumentation, mechanical,'and materials (including lubricants). PEco provided the associated commercial audit or commercial grade survey reports separately. The team reviewed the records for the selected dedications including purchase requisitions documents, RPEs, quality control (prs), purchase orders (Pos) PEG EES inspection reports (QCIRs), PEco test results, and maintenance work orders (W0s),
lhe following examples are items that PEco purchased as commercial grade and either installed or made available for installation in safety-related plant applications without performing a complete review for suitability of service, or, in some cases, a design verification review (seismic and environmental evaluation). The NRC inspectors did not consider that the findings documented in the examples resulted in the use of CGis that could cause operability problems, however, they were unable to assure that this was the case.
Accordingly, for those identified and similar items, LGS should review the dedications to assure that all parts are suitable for their intended safety-rclated applications.
(1)
The team reviewed the Jedication package containing Engineering Evaluation Summary (EES) 114-24129, dated May 30, 1991.
P0 LS639447, dated April 11, 1991, was issued to Lechler, Incorporated for the purchase of spray nozzles for the LGS spray pond.
The safety function of the spray nozzles was identified as providing the proper spray pattern for the cooling water. The critical characteristics listed in the EES package were material of construction, configuration, surface discontinuities, part number, dimensions, and pressure boundary integrity.
The acceptance basis for these characteristics was to be established by a source inspection at techler, Inc. and part number verification at receipt inspection. The vendor surveillance report for the spray nozzle indicated that source verification included observation of liquid penetrant inspection and hydrostatic testing and reviews of -135-
equipment, Caterial, and personnel qualifications.. Haterial veri-fication was confirmed using marking and heat numbers as well as actual dimensions being recorded in the report. The team considered the dedication of the spray nozzles acceptable except that there was no-objective evidence that the spray-nozzles were provided in the heat.
treated condition as required by the material specification.
PEG said it had requested the heat treat information from the supplier of the-l spray nozzle.
1 (2)
The team reviewed the dedication package containing EES ll4-75919.. dated November 24, 1990.
P0 LS631820, dated November 29, 1990, was issued to.
Statham Transducer Division of Schlumberger Industries for the purchase of 0-rings to be used in the standby liquid control level transmitter junction box. The critical characteristics listed in the EES package 1
were dimensions and material. The acceptance basis for these characteristics were the inside diameter and thickness and verification that the material was viton by comparison to previously received material. Of the 100 0-rings received, 2 were sent to the PECo laboratories for testing, which included measuring the.inside and outside diameters, thickness, and a material analysis-(by infrared spectrometry). The test results revealed that the. dimensions and material met the acceptance criteria. There was no documentation l
included in the EES package to indicate that any inspections or tests were performed on the remaining 98 0-rings. The team considered that the basis for performing sampilng and determination of the sample size selected was not fully specified as a result of the procurement program not providing requirements for sampling.
No other basis was given for assuming batch or lot homogeneity.
(3)
The team reviewed the dedication package containing EES Il4-81697,. dated.
September 18, 1990.
P0 LS628605, dated January 24. 1992, was issued to Hid Atlantic Instrumentation for the purchase.of 50 adapters (tube end male) for use in various pressure transmitters at LGS.
The critical-characteristics listed in the EES package were dimensions, marking, configuration and material. The basis for acceptence of the char-acteristics included receipt inspection, laboratory analysis, and measurements. Although the critical characteristics and verification methods were aapropriate, the team identified the following deficiencies associated witi the dedication of the adapters:
Of the-50 adapters received, 2 were sent to the PEco' laboratories for material analysis. The team considered that the basis for performing sampling and determination'of the sample ! size selected was not fully specified as a result of the procurement program not providing requirements for sampling. No other basis was given for assuming batch or lot homogeneity.
The material analysis performed on one adapter indicated that two sets of tests were performed on the ada)ter for' determining the chemical composition. The results of tie two tr sts were averaged for each element to determine-the chemical comp >sition. The'two.
tests indicated 0.042 and 0.095 percent carbon, an average of 0,069- -136-l l1 a.
percent carbon. Thu mxiru7 permissible carb:n for the adapter material, SA 479, type 316, is 0.08 percent.
The team pointed out that whether this analysis is considered a product or heat analysis, 0.095 percent carbon was outside the maximum upper limit permitted by Table 1, Product Analysis Tolerances, of the SA-484,
" Specification for General Requirements for Stainless and Heat-Resisting Wrought Steel products." According to section 10 of Specification SA-484, if the results of any test lot are not in conformance with the requirements of Specification SA-484 (i.e.,
excessive carbon content and the applicable product specification),
a retest sample of two specimens may be tested to replace each failed specimen of the original sample.
If one of the retest spect-mens f ail, the lot shall be rejected.
PECo laboratory personnel said they had no written procedures to address the team's concern regarding averaging test results when one result is unacceptable.
Laboratory personnel also said they had very little procedural guidance in performing their daily attivities.
The team observed that the EES package indicated that the adapters were part of the ASME Section 111 Class 2, pressure boundary.
PEG later said this was incorrect information and that the adapters were installed downstream of the instrument valve.
(4)
The team reviewed the dedication package containing EES 114-77164, Revision 0, dated February ll, 1991, which covered 5 volt de power supplies from Lambda Electronics Corporation. This package was satisfactory except for the seismic qualificattor..
Even though the critical characteristics identified for electrical performance, electrical connections, and mounting properly reflected the commercial grade nature of the component, the evaluation statrd that dynamic qualification was "not' af fected - same as originally supplied" (the original components had been supplied safety-grade by a system vendor).
The power supply is a " black box" enclosed within a cover, and weight was not specified as a critical characteristic, so PEco had no way af knowing the makeup of the power supply except from its measured electrical performance.
PEco personnel stated that the evaluation and documentation for review of the seismic qualification of replacement CGis to the original seismic qualification had been identified as a program weakness that would be addressed further.
(5)
The team reviewed the dedication package containing EES 114-76265, Revision 0, dated March 7, 1990. LGS P0 LS623477, dated March 16, 1990, was issued to Newark Electronics, a distributor, for 10 Sigma relays, model 41RO-2500G-BSL. The relays were intended for use as current sensing relays in the alarm system for the dc bus. None of the relays had been installed at the time of the NRC review.
The failure mode of the relay was specified as "N/A" and the failure impact was listed as "no alarm will be generated if relay is inoperable " The safety function of the relay was specified as "part of DC circuitry used in the ground detection system alarm circuit" and the performance requirements as " provide signal to alarm circuit for ground -137-
1 detection."
Das:d on this safety function, the EES listed part number / nameplate data, configuration, dimensions, and pickup and dropout power for adjustment G and S respectively, as critical characteristics.
The PEG engineer stated that the safety function had been incorrectly specified on the EES and that the safety function of the relay whs only to protect the integrity of the safety-related de bus.
The relay, which operated a timing relay for the bus ground alarm, was only required to not fa'l and ground the de bus and was not required to change or maintain a state. Also, listing as critical characteristics pickup and dropout power for adjustment G and 5, respectively, was an error that resulted from the engineer's misunderstanding of the data sheet for the applicable relay.
There was no adjustment for pickup and dropout power on the relay.
This error was corrected by Revision 1 to the EES dated October 19, 1990, before testing the relays, tven with consideration of the clarified safety function, the EES did not consider coil resistance, insulation resistance, and coil and contact voltage and current ratings as critical characteristics.
The EES did not document evidence of the relay's seis.nic qualification and indicated the seismic qualification of the relay was "N/A."
Revision 1 of the EES, dated October 19, 1990, stated that seismic qualification was applicable and that the qualification was maintained due to the item being an identical replacement. The EES did not document qualification testing or an engineering evaluation to show 'St the replacement item was indeed identical and therefore the origina seismic qualification had been maintained.
(6)
The team reviewed the dedication package containing EES 114-94556, Revision 0, dated June 18, 1990.
LGS P0 LS627008 dated June 19, 1990, was issued to Gould Shawmut, the manufacturer, for 500 fuses, model A2Y6. These fuses were purchased exclusively for Modifications 6108 and 6109 of the ground detector and transducer panels. Twenty of the fuses had been installed at the time of the NRC review, The EES did not document an evaluation of the fuse's failure mode, safety function, environmental qualification, or seismic qualification.
PEG personnel indicated that the evaluations were performed and documented as part of the modification packages, for which the fuses were intended, although this information was not documented in the EES package.
The EES listed configuration, dimensions, markings, current carrying capacity, and clearing time as critical characteristics.
Configuration and markings were to be verified for all fuses, current carrying capacity for 3 of the 500, and clear time for 2 of the 500. The EES did not document the basis for sampling or determination of lot homogeneity.
The fuses were tested at the PECo laboratories. The results indicated a failure of one of the three fuses during the current capacity test. The test, which required 110 percent of the 6 amps rated current to be applied for 15 minutes without fuse failure, was incorrectly performed by applying 116 percent of the rated current. The fuse failed the test -138-
at 10 cinute: 48 seconds (Test Result 90-033 dated June 20,1990). The test results were inad:quately reviewed and the fuses were subsequently issued and installed.
The fuse failure and, therefore, indeterminate quality of the remaining fuses was discovered by an additional review of the test results performed during the PEG self-assessment in June 1992.
NCR LG 92-00200, dated June 20, 1992, placed the 27 fuses remaining in stock on hold and revised the EES to perform the current carrying capacity and clearing time tests on 4 of the fuses remaining in stock.
The fuses passed the electrical tests and the NCR was dispositioned such that the fuses installed in the plant and remaining in stock.were acceptable for use.
Revision 2 of the EES dated June 18, 1992, added evaluations for failure mode, safety function, environmental qualification, and seismic qualification. Revision 2 of the EES did not document the basis for sampling although it did provide an evaluation for homogeneity of the lot.
(7)
The team reviewed the dedication package containing EES 114-73267 Revision 1, dated February 14, 1991, which covered temperature switches manufactured by Fenwal Incorporated and obtained from Carrier Corporation, the system vendor. This package was acceptable except the basis for acceptance of the dynamic qualification was not provided for the commercial grade replacement parts.
(8)
The inspectors reviewed the dedication package containing EES 114-94083, Revision 2. dated August 31, 1990, for model MDL-1 Bussmann fuses. The following P0s had been issued:
P0 Date Qty Haterial Status LS630453 10/13/90 100 issued to LGS LS638209 04/05/91 20 Issued to LGS PL394517-5 01/23/92 30 Not yet received PL394517-7 03/30/92 140 On QC hold The EES documented the failure modes and effects, safety classification and performance requirements, and environmental qualification, but did not fully-document evidence of seismic qualification of the fuses.
It indicated that the seismic qualification of the fuses was seismic Category I and that the qualification was maintained beca'ise the fuses were identical replacements to the fuses originally supplied. - The EES did not document qualification testing or an engineering evaluation to show that the replacement item was indeed identical and therefore the original seismic qualification had been maintained.
The EES listed configuration, dimensions, markings, current carrying capacity, and clearing time as critical-characteristics. Configuration, dimensions, and markings were to be verified on two fuses per order.
The EES did not document the basis for sampling or determination of homogeneity for verification of these critical characteristics. Current carrying capacity and clearing time were to be verified by one of two -139-h
=
+
=
w r
methods: (1) verification could be by commercial grade survey if the vendor was on the PEco EVL as an acceptable supplier of the fuses and if an adequate certificate of conformance was received with the material or (2) if those conditions were not met, the electrical test could be performed at the PEco corporate laboratories in accordance with instructions on the critical characteristics verification plan, if the testing was to be performed at the PECo corporate laboratories, resistance measurements were to be taken of the fuses; the two with the highest resistance and the two with the lowest resistance were to be tested. The EES did not document the basis for sampling or provide a complete basis for lot homogeneity.
The current carrying capacity and clearing time were verified by ccamercial grade surveys for the fuses from all four of the P0s. The team reviewed the applicable commercial grade surveys of Bussmann and determined that the surveys used to support P0 LS630453 and P0 L5638209 adequately documented verification of the current carrying capacity and clearing time. However, 'he surveys to support P0 PL394517-5 and P0 PL394517-7 did not fully document verification of the current carrying capacity and clearing time.
(9)
The team reviewed the dedication package containing EES 114-77308, Revision 1, May 15, 1991, which covered 75 uninsulated wire connectors manufactured by Thomas and Betts Company and obtained from Franklin Electric Company.
Although PECO increased the quantity ordered from 62 to the supplier's stated minimum quantity of 75, there was no evidence of lot homogeneity, and the basis for selecting two connectors for material testing was not provided. Otherwise, the package was acceptable.
(10) The team reviewed the dedication package containing EES 114-76555, Revision 0, July 13, 1990, which covered three level-indicating gauges from Dresser Industries Co. Dynamic qualification was addressed with the statement "like-for-like replacement as originally supplied."
Dedication was based on a combination of receipt inspection and post-installation calibration by I&C technicians. With the exception of the seismic qualification, which did not include an engineering evaluation to show that the replacement item was indeed identical and the original seismic qualification had been maintained, the package was acceptable.
(11) The team reviewed the dedication package containing EES 114-24694, Revision 0, dated April 23, 1990. LGS P0 LS625010 dated May 22, 1990, was issued to Wilmar Electronics incorporated, the manufacturer, for the purchase of 12 Wilmar Electronics relays, model 4300CX.
The relays were intended for use as undervoltage sensing relays in the diesel generator control board. None of the relays had been installed at the time of the NRC review.
The safety function was specified as "to assure tte integrity of the class IE circuitry in which it is being used" and "to initiate corresponding actions." The EES listed part number / nameplate data, configuration, dimensions, pickup and dropout voltage, and contact -140-
resistance as critical characteristics. A safety function for a typical application of a relay tsould be to change and maintain state as required under all design conditions.
lhe EES did not evaluate critical characteristics applicable to a fully specified safety function such as I
insulation rt9 stance, coil (sensor) and contact voltage and current ratings, and contact timing The EES required the part number / nameplate data, configuration, and j
dimensions to be verified for all 12 of the relays. Pickup and dropout t
voltage was required to be verified by testing at the PEco corporate laboratories for 3 of the 12 relays. The EES did not document the basis for sampling or determination of homogeneity. One failure occurred when the three relays were tested for pickup and dropout voltages; therefore, the EES was revised to require testing of the nine remaining relays.
i The EES did not document evidence of the relay's seismic qualification.
The EES indicated that the relay was a like-for-like replacement _of a commercial grade item and the dynamic qualification was not affected.
The EES did not document qualification testing or an engineering evaluation to show that the replacement iten was indeed identical and therefore the original scismic qualification had been maintained.-
l (12) The team reviewed the dedication package containing EES 114-94764,-
Revision 0, dated October 18, 1990. LGS P0 LS632464 dated October 26, 1990, was issued to Weidmuller Terminations Inc., the manufacturer, for 200 tcrminal blocks, model SAK4. ' Four of the terminal blocks had been installed in electrical panel 00-C692 at the time of the NRC review.
The EES did not address the possibility of the conductive portion of the-terminal block grounding as an anticipated failure mode.
The EES-documented safety classification and performance requirements, and:
environmental qualification but did not fully document _ evidence of the-seismic qualification of the terminal block, it indicated that the r
terminal block was seismic category I and that the qualification was maintained because the item was an identical replacement to the terminal block originally supplied.
The EES did not document qualification testing or an engineering evaluation to show that the replacement item was indeed identical and therefore the original seismic qualification had been maintained.
The EES listed part number, configuration,-voltage and current ratings, materials, and dielectric strength as critical characteristics. The critical characteristic verification. plan required part number, configuration, and voltage and current ratings, to be verified at-i receipt inspection by comparison to the vendor catalog information.
It is not enough to perform only a visual inspection of a terminal block, compare it to the manufacturers literature, and state that it meets its specified vol_tage and current ratings. Voltage and current ratings are representative of the ability to properly perform under, and withstand, the rated voltage _and current.
Voltage and current ratings can be verified by various methods, including a commercial grade survey or source inspection of the manufacturer to. verify controls or tests were i
. -141-
. -.=
in place and properly implemented, or testing of a sample of terminal blocks received from a homogenous lot.
(13) The team reviewed the dedication package containing EES 114-39;60, Revision 0, dated May 13, 1991.
P0 LS642883, dated July 8, 1911, was issued to Rumsey Electric Company for two limit switch lever ar ns manufactured by Namco Controls. Material composition of the it ser arm parts was listed as a critical characteristic and was to be vetified as i
bronze for the body, beryllium copper for the roller, and stainless steel for the bushing during testing at the PEco corporate laboratories.
Test results from the PECo EDAX chemical analyzer for the roller material did not indicate any beryllium content, only 100 percent copper.
LGS did not have any documented justification as to why the material test results did not include any beryllium content and how the material was accepted by PECo based upon the material test results.
(14) The team reviewed the dedication package containing EES 114-75305, dated January 11, 1992.
LGS P0 LS639725, dated April 4, 1991, was issued to Carrier Corporation for the purchase of six Nusco dehydrator float
-valves for control room chillers OAK 112 and OBKll2.
The safety function of the float valve is to maintain the pressure boundary of the chiller to allow it to perform its safety function.. Critical characteristics-listed on the EES for the float valve were dimensions and materials.
The verification of dimensions and material were to be accomplished during receipt inspection and testing. _ Specific tests were required to verify the valve body as brass, the valve plug as monel 500, two brass washers, one neoprene washer, and one hex nut as stainless steel. Of the six purchased valves, three were sent to the PECo corporate laboratory and only one was tested for materials.
No documentation existed to verify lot homogeneity.
The EES dynamic evaluation for seismic integrity consisted only of a claim that the replacement CGI was identical to the original' part.- No evaluation was done to show that the replacement was indeed identical and therefore the original seismic qualification had been maintained.
(15) The team reviewed the dedication package containing EES 114-73297, dated August 15, 1991.
LGS P0 LS639929, dated June 13, 1991, was issued to Carrier Corporation for the purchase of eight V-ring assemblies for the size 460V-3PH-60 control room chiller. No safety. function was listed for the V-ring assembly itself.
Critical characteristics listed on the EES for the V-ring were material and dimensions / configuration.
The verification of dimensions and material were to be accomplished during receipt inspection and testing. Specific tests were performed-to verify the material as teflon. - 0f the eight purchased V-rings, only one was tested for materials at the PECo corporate laboratory. No documentation existed to verify lot homogeneity.
The EES dynamic evaluation for seismic integrity consisted only of a claim that the replacement CGI was identical to the original part. -No l
-142-u-....
evaluation eas done to show that the replac: ment was indeed id:ntical l
and therefore the original seismic qualification had b:en maintained.
4 SYSTEM UPGRADE DEDICATION ACTIVITIES Nuclear Engineering and Services Department (NESD) under NED is responsible for preparing procurement and dedication packages for CGis required in support of LGS-modification activities and dispositioning nonconformance reports.
Nuclear Engineering Department Procedure (NEDP) 4.10. " Procedure for the Dedication of Commercial Grade items for Use in Safety-Related Applications,"
Revision 1, dated July 1, 1989, and NEDP 3.14, " Procedure for Performing Subcomponent Evaluations," Revision 1 (with NEDP Interim Guidance 92-16, dated March 20, 1992), are the two primary procedures used by NESD for the procurement and dedication of CGis along with several ancillary procedures for performing activities such as safety evaluations, design changes, and procurerent.
The inspection team reviewed an example of the upgrading of a portion of a safety-related system initiated by Nonconformance Reports (NCRs) L90159 and L90161.
These NCRs identified that the safety-related inflatable seals that form part of the secondary containment for the reactor enclosure during normal operation and the inflatable seals that form part of the secondary containment boundary for the refueling floor, were not provided with a safety-related air supply system and the system boundary valves were not safety-related.
The NCRs also indicated that the use of the non-safety 4related air supply system would not ensure that the seals would remair, inflated during all necessary design conditions.
The team reviewed several EES packages prepared by NESD in support of upgrading the existing installed air supply system to the inflatable seals.
Even though the inflatable seals were supplied by service air, which was not safety-related, the seals were always inflated when tested to demonstrate secondary containment integrity. The upgrade included performing activities such as walkdowns, generating as-built design documents, performing engineering analysis, and dedicating existing plant equipment for safety-related application.
NESD generally used the dedication methodology described in Section 2 of this inspection report for preparing the EES packages, however, many items that would normally be tested at the PECo laboratories, were not sent to the PECo laboratories for testing since they were installed at the time the upgrade was performed. PEG and NESD indicated that additional laboratory testing would have been specified if the CGis had been newly procured items as opposed to items already installed. fer example, the EES packages for the inflatable seal system boundary valves and associated components, L 90-015 for a relief valve, L 90-013 for a three-way plug valve, L 90-016 for a pressure switch, and L 90-017 for a pressure indicator, did not list materials of construction as a critical characteristic although material was integral to the system safety function. NESD informed the team that in some instances the material of construction could only be verified using a magnet to determine whether a material was magnetic or non-magnetic during the system upgrade walkdown. The -143-
teaa noted that the use of a tagnet does not' identify the caterial type, but' only identifies whether or not the material is in the ferrous-magnetic family.
i In addition, the NESD evaluation for the air supply system u> grade indicated that any future repairs, replacements, or modifications of tie upgraded air supply system would be performed in accordance with the original design requirements and ASME Section XI Code requirements.
It also indicated that'
+
the valves in this upgraded system would be added.to the LGS ASME Section XI pump and valve test program, and that the ASME Section XI inservice inspection (ISI) drawings would be updated to reflect the upgraded air supply piping.
The team requested the updated ASME Section XI pump and valve test program and-the associated surveillance test procedures for the upgraded air supply system, but the LGS inservice testing (IST)' engineer indicated that the LGS ASME Section XI pump and valve test program had not been formally updated to:
include the upgraded air supply system components and that the surveillance test procedure for the relief valves had nnt been written. Discussions with the IST engineer revealed that certain test frequencies required by the ASME-Section XI Code were not being met and, if these components were within the scope of ASME Section XI, relief requests are required for the reduced test frequencies.
Following an internal NESD meeting with various PEco organizations, NESD informed the team that the engineering evaluation for the upgraded air supply system, as written, was unclear and somewhat misleading, and that it was never intended that the upgraded air supply system would be subjected to the IST and 151 requirements of ASME Section XI. The team noted that LGS has committed to meeting certain recuirements such as Regulatory Guide 1.26, " Quality Group Classifications anc Standards for Vater, Steam, and Radioactive-Waste-Containing Components of Nuclear Power Plants," Revision 3, dated F_ebruary 1976, certain portions of the NRC Standard Review Plan, and that the inclusion or non-inclusion of the upgraded air supply system in the LCS ASME Section XI Code IST and 151 programs should be based on these commitments.
The team reviewed the post-upgrade testing of the upgraded air supply system
~
and components and concluded that the tests performed demonstrated that the upgraded system was functional. _ However, the team noted that no surveillance test procedure had been developed for testing the relief valves in the upgraded air supply system.
S PROCUREMENT AND DEDICATION TRAINING Training and qualifications were reviewed for the personnel directly involved-in the CGI dedication process, namely, PEG, Nuclear Engineering Division (NED), and quality verification receipt inspectors. The-PEG is currently divided into an electrical and mechanical group with a lead engineer over each group.
PEG position guides were inspected.
Procedure MSP-10, " Procedure For Training," Revision 2, provides the minimum acceptable level of job-related training for PEG staff. The procedure lists documents to-be read and understood by PEG staff.
The procedu're specifies- -144-
l that each listed document, ofter being road by the trainee, be discussed with the supervisor to ensure that the trainee understands the subject matter. The procedure also provides an alternate method of assessing a trainee's understanding, i.e., discussions ma; ce held in a classroom or group environment after which the trainee signs beside each listed document.
Discussions held with the lead mechanical engineer verified that an HSP-10 training signoff had been completed for each person in the group. MSP-10 had initially been issued in June 1989. lhe lead mechanical group engineer stated that some record backfitting had been done for experienced PEG staff who had been performing in the job before the implementation of HSP-10 signoff in crder to establish a training record base for PEG. A review of the documents listed in HSP-10 noted the listing to be reasonably comprehensive.
The team determined that at the time i he inspection there was no formalized PEG program for retraining nor was a in PEG designated to handle retraining. During the course of the aspection the Haterials rianagement Supervisor designated the lead mechanical engineer to be in charge of training / retraining.
The lack of a retraining policy within PEG was considered a weakness.
A 2-day EPRl/NCIG training seminar / workshop on " Dedication of Commercial Grade
!tems" was attended in late 1989 and early 1990 by 12 of the current PEG members.
In 1988, 6 of the current PEG group attended a 2-day Babcock and Wilcox training course on " Procurement of Replacement items." five PEG members have attended a 1-day seminar on testing of CGis at the PEco corporate laboratories.
Eight PEG members have attended a 4-day course at PECo corporate laboratories on " Metallurgy, Welding, and NDE." After reviewlag the courses' outlines and taking a tour of the PEco corporate laboratories ano interviewing PEG members, the team concluded that the courses provided meaningful training for PEG members doing CGI dedication work.
The only formal requirement for its NED engineers who perform CGI dedicatir n work is that the engineer must have completed initial qualification trainbg.
All NED CGI dedication packages are required to be approved by the Equipment Qualification (EQ) branch, which has some engineers with CGI dedication training. The current EQ Branch supervisor was noted to be lacking in CGI dedication training. The team concluded that NED training and qualification requirements to perform CGI dedication work was not well formalized and training / retraining of NED personnel to conduct CGI dedications was lacking.
The quality verification receipt inspector qualifications were selectively N
verified to meet qualification requirements. An interview with one receipt inspector verified that he had attended the training courses listed in his training records, including a course entitled " Fraudulent Material Awarene.s."
A review of the training course materials found the course to be reasonably comprehensive. The quality verification receipt inspector also was reasonably knowledgeable about the subject. However, quality verification receipt inspector training relative to the CGI dedication program was relatively old and could be enhanced through updating. -145-
6 EXIT HEETING On March 27, 1992, the inspection team conducted an exit meeting with. members of the PECo staff and management at the LGS site. During the exit meeting the team summarized the inspection findings and observations. The individuals listed below were present.
EhRadelchia Electric Como_ toy _
J. O'Rourke, Project Manager, limerick Generating Station (LGS)
L. Hopkins. 0)erations Superintendent, LGS J. Huntz, Tecinical Superintendent, LGS J. Cotton, General Manager, Nuclear Quality Assurance (NQA)
C. Wyler, Superintendent, Materials Management-J. Evans, Superintendent, Vendor Evaluation Section R. Lesnefsky, Superintendent, Quality Verification Section E. Troy, Superintendent, Quality Support Secticn J. McElwain, Superintendent, Quality Assurance Section J. Gyrath, Branchhead, Engineering Assurance A. Skapik, Supervisor, Procurement Engineering Group (PEG)
J. Seigmund Electrical Supervisor, PEG T. Supplee. Engineer, Materials Management R. Krich, Branchhead, Licensing J. Phillabaum, Engineer, Licensing.
R. Gropp, Licensing _
D. Neff, Engineer, Licensing R. Smith, Senior Auditor, NQA J. DeLong, Section Manager, PECo Corporate Laboratories J. Dinardo, Manager, Plant Evaluation V. Cristniewicz, Superintendent Training, LGS Nuclear Renulatory Commbilga W. Hodges, Director, Division of Re:ctor. Safety. Region 1 N. Blumberg, Section Chief, Region 1 L. Norrholm, Chief, Vendor Inspection Branch (VIB), NRR V. Potapovs, Section Chief, VIB R. McIntyre, Team Leader, VIB R. Wilson, Senior Reactor Engineer, VIB L. Campbell, Reactor Engineer, VIB B. Rogers, Reactor Engineer, VIB D. Caphton, Senior Reactor' Engineer, Region 1 E. Benner, Reactor Engineer, Region 1 R. Clark, Project Manager, NRR R. Lobel, Regional Coordinator, ED0 T. Kenny, Senior Resident inspector, LGS Other Personnel t.
L B. Bradley, Senior Project Manager, NUMARC -146-
/sas n%$
UN11ED STAT [s
!4. J ?
NUCLEAR REGULATORY COMMISSION i
wum9ntow.o c 70555
\\YW}'ls November 3, 1992 Docket No. 99901066/92-01 M<
Martin Hunt Quality Assurance Manager Rotork Controls Limited Bath England BAl 3JQ
Dear Mr. Hunt:
SUBJECT:
NOTICE OF NONCONFORMANCE (NRC INSPECTION REPORT NO. 99901066/92-01)
This letter addresses the inspection of your facility at Bath r gland, n
conducted by Mr. Richard P.- McIntyre and Mr. Uldis Potapovs of
.is office on August 24 through 26, 1992, and the discussions of their findings with you and other members of your staff at the conclusion of the inspection. The purpose of the inspection was to review Rotork Control Limited's (Rotork's) corrective actions for previous inspection findings and unresolved items and to evaluate Rotork's quality assurance (QA) program, including material procurement, control and audit of subvendors, and the dedication of commercial grade items used in the production of valve actuators and actuator spare parts sold as nuclear safety-related.
Areas examined during the Nuclear Regulatory Commission (NRC) inspection and our findings are discussed in the enclosed inspection report. This-inspection consisted of an examination of procedures.and representative records, interviews with personnel, and observations by the inspectors.
During -this inspection it was found that the imalementation of your QA program failed to meet certain NRC requirements. Rotor ('s QA program allowed the procurement of items, that are used in nuclear valve actuators and spare parts, from suppliers listed in the latest edition of The United Kingdom Register of Quality Assessed Companies (DTI QA Register), without aerforming any assessments of the suppliers' quality programs, such as througi implementation audits or testing. Also, Rotork failed to complete and document the corrective actions committed to the NRC for a past inspection finding. The specific findings and references to the-pertinent requirements are identified in the enclosures of this letter.
The inspectors considered as a program improvement, since the last NRC inspection, the fact that Rotork has made considerable progress towards implementing a program for the dedication of items purchased as commercial grade and used in safety-related nuclear valve actuators.
-147-
Hr. Hartin Hunt Please provida us within 30 days from the date of this letter a written statenient in accordance with the instructions in the enclosed Notice of Nonconformance. We will consider extend, i the re O nse time if you can show good cause for us to do so.
The response requested by this letter and the enclosed Notice of Nonconformance is not subject to the clearance procedures of the Office of Management and Budget as required by the Paperwork Reduction Act of 1980, Public Law No.96-511.
In accordance with 10 CFR Part 2.790 of the NRC's ' Rules of Practice," a copy of this letter and its enclosure will be placed in the NRC's Public Document Room.
Should you have any questions concerning this inspection, we will be pleased to discuss them with you.
Sincerely, Leif orrholm, Chief Vendor Inspection Branch Division of Reactor inspection and Licensee Performance Office of Nuclear Reactor Regulation
Enclosures:
1.
Notice of Nonconformance 2.
Inspection Report No. 99901066/92-01 c.c : w/ enclosures Mr. Robert Arnold, President Rotork Controls Incorporared 19 Jet View Drive Rochester, New York 14624 s
-148-2
s Enclosure 1
i NOTICE Of NONCONFORMANCE-j Rotork Controls Limited Docket No. 99901066 Bath, England Based on the results of a Nuclear Regulatory Commission (NRC) inspection conducted on August 24-26, 1992, it appears that certain activities were not conducted in accordance with NRC requirements.
A.
10 CFR Part 50 Appendix B, Criterion VII, " Control of Purchased Haterial, Equipment, and Services ' requires that measures shall be established to assure that purchased material, equipment, and services, whether purchased directly or through contractors and subcontractors-conform to the procurement documents. The effectiveness of the control of quality by contractors and subcontractors shall be assessed by the applicant or designee at intervals consistent with the importance, complexity, and quantity of the product or services.
Contrary to the above, as of August 1992, Rotork purchased items from suppliers listed in the latest edition of The United Kingdom Register of Quality Assessed Companies (DTI QA Register), without performing any assessments, such as implementation audits for verification of the suppliers' quality programs or testing of the supplied material.
B.
10 CFR Part 50, Appendix B, Criterion XVI, ' Corrective Action," requires that measures shall be established to assure that conditions adverse. to-quality, such as f ailures, malfunctions, deficiencies, deviations, defective material and equipment, and nonconformances are promptly identified and corrected. The identification of the significant l
condition adverse to quality, the cause of the condition, and the-corrective action taken shall be documented and reported to appropriate management.
Paragraph 14.1 of **ction 14, " Corrective Action," of the Rotork Quality-Assurance Policy. Manual states that prompt and effective corrective action is essential to the quality system.. Identification and segregation of defective components / products alone is not sufficient; the cause of defects must be found and corrected to avoid repetition.
Contrary to the above, Rotork failed to complete and document the corrective actions committed to the NRC in the Rotork March 13, 1990, letter in response to Nonconformance 89-01-08 from NRC Inspection Report 99901066/89-01.
I -149-a
Please provide a written statement or explanation to the U.S. Nuclear Regulatory Commission, ATTN: Document Control Desk, Washington, D.C.
20555 with a copy to the Chief, Vendor Inspection Branch, Division of Reactor Inspection and Lu.ensee Performance, Office of Nuclear Reactor Regulation, within 30 days of the date of the letter transmitting this Notice of Nonconformance. This reply should be clearly marked as a " Reply to a Notice of Nonconformance" and should include for each nonconformance:
(1) a description of the steps that have or will be taken to correct these items; (2) a description of the steps that have been or will be taken to prevent recurrence; (3) the dates your corrective actions and preventive measures were or will be completed.
Dated at Rockville, Mary' land j
this day of / !/ > ' // !/ / ? 992 J l
-150-
ORGANIZATION:
Rotork Controls Limited Bath, England REPORT NO.:
99901066/92-01 CORRESPONDENCE ADDRESS:
Mr. Martin Hunt Quality Assurance Manager Rotork Controls, limited Bath, England BAl 3JQ ORGANIZATIONAL CONTACT:
Mr. Martin Hunt Quality Assurance Manager NUCLEAR INDUSTRY ACTIVITY:
Supplier of nuclear safety-related and environmentally qualified valve actuators and actuator spare parts.
INSPECTION CONDUCTED:
August 24 through 26, 1992 LO /2o fqt SIGNED:
Richard P. McIntyre,' leam Leade Date Reactive inspection Section No. 1 Vendbr 1.
ection Branch (VIB)
M&QP 10 -2.\\- N APPROVED:
Uldis Potapovs, C1h. ef Date Reactive Inspectiori Section No.1 Vendor Inspection Branch (VIB)
INSPECTION BASES:
10 CFR Part 21 and 10 CFR Part 50, Appendix B INSPECTION SCOPE:
To review Rotork's Quality Assurance Program relative to the assembly, test, and supply of valve actuators and actuator spare parts sold to U.S. utilities for use in nuclear safety-related applications.
PLANT SITE APPLICABILITY:
Multiple
-151-
1 INSPECTION
SUMMARY
1.1 Nonconformanen 1.1.1 Contrary to 10 CFR Part 50, Appendix B, Criterion V!l, " Control of Purchased Material, Equipment, and Services," Rotork Controls Limited (Rotork) purchased items from suppliers listed in the latest edition of The United Kingdom Register of Quality Assessed Companies (DTI QA Register), without performing any assessments, such as implementation audits for verification of the suppliers' quality programs or testing of the suppli3d material.
(92-01-01) 1.1.2 Contrary to 10 CFR Part 50, Appendix B, Criterion XVI, " Corrective Action," and Paragraph 14.1 of Section 14, " Corrective Action," of the Rotork Quality Assurance Policy Manual, Rotork failed to complete and document the corrective actions committed t' the NRC in the Rotork March 13, 1990, letter in response to Nonconformance W-01-08 from NRC Inspection Report 99901066/89-01.
(92-01-02) 2 STATUS OF PREVIOUS INSPECTION FINDINGS 2.1 Nonconforman;n J
2.2.1 (CLO A ) Nonconformance 99901956/89-01-01 The standard QA provisions Rotork imposed on their subcontractors did not require the subcontractors to pass down appropriate QA provisions. This resulted in subcontractors purchasing material and accepting the material based on a certificate of conformance (CoC) or certified material test report (CMTR) without requiring the supplier to have control uf his subcontractors or verifying the validity of the certification.
Rotork corrective action included sending a letter to all suppliers which addressed meeting 10 CFR Part 50, Appendix B, requirements and included guidance on how to provide adequate material traceability. This letter also included a sample heat treatment form and sample CMTR. All Rotork purchase orders for nuclear parts (with an N designation) now include provisions for supplier notification to Rotork of any changes in design, material, or manufacturing and also invoke the requirement to meet QC3/T, Revision 3, "Rotork Supplier Quality Control Requirements," for process / material control and Quality Control (QC) Procedure 97, Revision 3, "Rotork Supplier Material Certification and Traceability Requirements." Finally, Rotork has performed audits of most of their suppliers to verify conformance in these areas.
2.2.2 (CLOSED) Nonconformance 99901066/_89-01-02 The Rotork QA Hanager's Volume 2 of the QC procedures contained four obsolete procedures.
The inspectors verified that the four obsolete procedures (deleted) had been removed from the QA Manager's master copy of QC procedures and also verified -152-
that no other deleted procedures were included in'this master copy. The inspectors also verified that the QA Hanager had in fact received the appropriate revisions of the four QC procedures.
2.2.3 (Cl0 SED) Nonconformance 99901066/89-01-03 Paragraph 4,5 of the Purchasing Department Manual allowed Rotork to place orders with companies not approved by quality control (QC).
Paragraph 4.5 was revised to clearly state in Section 5.3.2 that purchase orders for nuclear items are to be placed only with subcontractors included on the list of approved subcontractors (Approved Supplier File).
2.2.4 (CLOSED) Nonconformance 99901066/89-01-04 Rotork had not established measures to' assess the effectiveness of the control of quality by their subcontractors.
Corrective action for this finding included the establishment of a " rolling" audit program for the audit of subcontractors based on QC Procedure 306 requirements. The audit program requires that all subcontractors be audited every three years against QC3/T or the British Standard BS5750 (ISO 9000).-
The results of these audits, along with historical supplier performance based on rejections; return of corrective action slips from reject notes;.and concessions are factored into the quarterly performance rating for each supplier.
The inspectors verified that a 1992 vendor audit schedule was in place and that the audits were being performed and documented for subcontractors on the approve supplier file. The inspectors also reviewed the September 10, 1991, audit report of KPG Engineering Limited, the supplier of wormwheels and wormshafts, and the November 20, 1990, audit of Robertson Tooling Limited, the-supplier of keys and die tools. Both of these audits appeared to be acceptable.
2.2.5 LCLOSED) Nonconformance 99901066/89-01-05 Testing performed under Procedure QC-80 did not assure that suppliers of.
material effectively control quality. All required material properties were not being verified and the results of material testing were not factored into the vendor's rating.
Proedure QC-80, " Material Certification," Revision 7, dated July 26, 1989, describes an overcheck program of material certification'for a sample of components and parts received, whether purchased as nuclear safety-related or non safety-related. QC-80 was revised as part'of Rotork's corrective action to require material chemical analysis, and.a check of physical properties (such as hardness) to the design specification or drawing, be performed on all material overchecks. Failures and the corrective action are now documented on a Non-Conformance Report, and in some cases, further checks may be implemented. Defective material found as a result of an overcheck is now factored into the vendor rating by means of a rejection note. QC-80 also -
-153-
4 requires traceability of material to its paperwork be maintained, with the specific batch or unique order numbers referenced on all vendor certification.
The inspectors reviewed the 1992 overr. heck log and chose four wormwheels and three wormshafts that had been sent out to a laboratory to be tested for chemical content and physical properties. Non-Conformance Report Nos. 163463 through 163469 had been written on these items. All had materials that were out of specification that were accepted by engineering. The inspectors determined that in some cases the engineer referenced the wrong specification cn the purchase document sent to the testing laboratory, but in other cases the acceptance was based solely on engineering judgement.
2.2.6 (CLOSED) Nonconformance 99901066/89-01-06 Heat treatment certifications for wormshafts supplied by Dava11 and TEE, Limited were not traceable to the individual wormshafts or the base material certification.
Rotork's corrective actions included implementing provisions that require traceability of material against its certification to a unique batch number or the material cast number, if it is unique to the batch, _ and that it be included on all P0s. This is accomplished through Rotork invoking QC3/T for_
QC requirements and procedure QC NO. 97 for material certification and i
traceability requirements on P0s to all subcontractors when applicable. As stated earlier, Rotork also requires that all material certificates received cross reference and identify the unique work order number and batch number.
Section 4.a of QC3/T states, " Ensure that parts are traceable via a unique reference number to any associated material / process certification, the Rotork order, and to the material (eg: heat / batch number)."
The inspectors reviewed several material certificates, CoCs, and CMTRs for drive pins
)urchased from Williams and_0akey Engineering Company Ltd. (WOL) and clutch (eys purchased from Robertson Tooling Ltd. (RTL). The inspectors verified that traceability was maintained by these vendors through referencing the order and batch numbers on all certification paperwork su) plied to Rotork, including the Cots from the subcontractors who performed the leat treatment on the items. Two of the original mill certs (CMTRs) recieved from steel manuf acturers were provided by MaCreadys Steel, London. When reviewing implementation audits for these vendors for inclusion on the Rotork Approved Suppliers file, the inspectors were told that Rotork's QA program allowed the procurement of material from vendors listed in the latest edition of The United Kingdom hegister of Quality Assessed Companies (DTI QA Register),
without performing any assessments of these vendors' quality programs such as through implementation audits or testing.
Therefore, since Rotork does not audit these types of material suppliers or test the material upon receipt, it does not have verified traceability back to the steel mill.
Rotork agreed that this was a program deficiency and stated
, -154-
that it _would be addressed during the completion of the implementation of the commercial grade item dedication _ program and the appropriate procedures, j
This program deficieKy is identified as Nonconformance (92-01-01).
2.2.7 (CLOSED) Nonconfarmance 99901066/89-01-07 Rotork had no system in place to track which vendors had not submitted their corrective action or had returned the tear-off-slips that described their corrective' actions taken on nonconforming products.
As discussed earlier in Section-2.2.4 of this report, procedure QC 306 requires Rotork to monitor and track the return of the corrective action slips for the corrective action taken by vendors for reject notes. _Once reject notes are sont-to verdors, they are allowed six weeks to respond to the reject note by return of the corrective ' action sli). The tracking of the reject notes is computerized and is performed by tie Supplier Quality Auditor. The failure of a vendor to respond to the reject note within the allowed time
-frame is automatically factored into the vendor's quarterly performance ratings so appropriate action can be taken.
The Inspectors reviewed the computerized Vendor Rating Reports for several 1
nuclear suppliers to verify that their ratings are being u> dated quarterly r
using the appropriate performance factors and results of t1e returned corrective action slips from reject notes.
2.2.8 (CLOSED) Nonconformance 99901066/89-01-08 Deviations or nonconforming material discovered as a result of material verification testing performed under Procedure QC-80 were not controlled under a program that requires documentation of the cause of the condition and the corrective action taken to 3revent recurrence. Also, Test Report No. 1769 for drive pin 09332 indicated t1at nickel content was out of specification (low) and no test was performed to determine if the mechanical properties were a f feci.ed.
QC-80 was revised as part of Rotork's corrective action to require material chemical analysis and a check of physical properties (such as hardness) be performed or all material overchecks.
Failures and the corrective action are now required to be documented on a Non-Conformance Report. -The Non-Conformance Report includes a section for documentation of likely causes of the defect and another section for corrective actions to prevent repetition.
The March 13, 1990, Rotork letter to the NRC, with corrective actions for the nonconformances identified in NRC Inspection Report 99901066/89-01, stated that the test report for drive pin 09332 was % be investigated to determine the extent of the actions taken at the time, fhis investigation was to be completed by April 1990, with full details sent to the NRC for review.
The details of the investigation results were not sent to the NRC in 1990.
When the inspectors requested to review the results of this investigation during this inspection, Rotork stated that they had completed the ~155-
. investigation,however,ltheyneverinitiatedLaNn-ConformanceReport-of documented the results of this investigation. _ Rotork issued Non-Conformance t
Report No. 163470 dated August 25, 1992, and documented the results of the -
1990_ investigation of. drive pin ~09332 to address this failure to complete their corrective actions. The drive pin was deemed acceptable by Rotork based upon the engineering department reasoning that low nickel content will not effect the performance of the pin due.to the very small cross -section involved.
This failu e m complete and document the committed corrective actions is -
identified o Nonconformance (92-01-02).
4 2.2.9 (CLOSED) Nonconformance 99901066/89-01-09 The Rotork Purchasing Department could not locate purchase order 19889.
Neither a microfilm nor a hard copy of purchase order 19889 was ever located,=
however, the basic order details are available on the mainframe computet system. Rotork conducted a search of a random sample of 22 purchase orders, and in each case a microfilm or hard copy of the purchase order was located.
Therefore, this appears to be an isolated case.
2.2 Unresolved Items 2.2.1 (OPEN) Unresolved Items 99901066/89-01-10. 89-01-11. 89-01-12 Review of the actuator assembly design basis and the control of design modifications identified that sufficient information was not available to demonstrate that certain design / material modifiutions did not compromise the environmental qualification of this equipment. Three separate items were identified as examples.
In its response, dated November 8,.1990, Rotork stated that the unresolved items would be addressed through the' development and implementation of a commercial grade item dedication program. Review of-dedication activities during this inspection indicated that, although-considerable progress had been made towards the development of a dedication program, this progran was not yet fully implemented and justifications of all previously purchased equipment to the qualification test repert were not completed. These itens remain unresolved.
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Backaround Rotork Controls Limited _(Rotork Ltd.), ath, England purchases parts, then assembles, tests, and supplies valve actuators and actuator spare parts as both nuclear safety-related and non safety-related. The safety-related nuclear actuators (type NA) and spares are produced under their 10 CFR-Part 50, Appendix B, QA program. -The nuclear actuators and spares are sold as safety-related to U.S. utilities by Rotork Controls Incorporated,. Rochester, New York, since Rotork Ltd. does not sell. directly to U.S. utilities. Rotork Ltd. stated that its sale of nuclear actuators and spares to U.S. utilities -156-
makes up :less than one half 1 percent of their total sales and that it-has noti supplied any complete actuators to.U.S.- utilities in thulastiseveral_-years.
3.2 [ommercial Grade item Dedication Proaram-
' As a result _ of corrective action commitmeats documented in' Rotork's. letters _
of March 13, 1990, and'NovemberJ 8, 1990, in response to HRC= letter-'and Notice-.
r
- of_Nonconformance dated January 31,.1990,
- Rotork contracted for consulting -
services to develop _a commercial: grade item dedication program for-its product-line. -The consultant recommendations-have been evaluated by Rotork and assembled into Rotork Technical Report ER 187, Designation, Justification.-
and Dedication of Components. Used in the Rotork NAl. Type Actuators," dated August 18, 1992. - This report contains separate sections on. Technical =
1 Rationale, Part Level Specifications for Components, Justification 'of Components, Dedication of Components and includes a Dedication-Data Base and User Application Manual.
ER 187 provides a rationale for and contains parts level classification for-all purchased items used in the NA1 Type actuators.
Part classification is based primarily on the safety function of the parent component, safety function of the item within the parent component, mode (s) and probabilities of item _ failure, and the effect of such failure on the safety function of the parent component. According to this analysis, the NAl actuator typically contains approximately 45 to 50 safety related parts.
Critical characteristics of safety related. items are base.1 on part number or.
other item identification attributes, physical characteristics and performance characteristics. The acceptance criteria for. identified critical characteristics are based primarily on Electric Power Research Institute (EPRI) 5652 methods 2 and 3 (commercial grade survey and source _ verification),
supplemented by method 1 (special test and inspection) as required.
In:
addressing the use of EPRI 5652 method 2, paragraph 2.3.3 of ER 187 stated' that "... the CG survey is based on either a_ review carried out-by'Rotork'or a third party audit carried out~by an accredited certification authority for those suppliers operating an ISO 9001 program ^1n the UK."
Rotork was cautioned that it may be difficult to use a third party audit of an ISO 9001 program for this purpose since EPRI method 2_is specific to the scope.of the particular commercial grade item being purchased and is intended to provide assurs u that specific cri'.ical characteristics of that-item are being adequately controlled. A broad based programmatic audit would normally not be-expected to accomplish these objectives.
ER 187 also contains a section on justification of components intended.to reconcile all safety related items in the NAl act Jators to the 1978/1979 Wyle Laboratories test reports which provide the basis for environmental and -
scismic qualification of Rotork actuators. -The current Rotork quality system requires engineering changes to be documented and appropriate. evaluations to be carried out to assess the impact of such changes on the qualification status of the actuators. However, this system was_not in effect in 1978 when the qualification actuators were assembled. As a part'of the component justification effort, a number of parts manufacturers have been contacted and requested to provide information concerning changes in material formulations. -157-
While significant progress has been made, the justification effort to date has not fully resolved the issues identified during the 1989 NRC inspection. One of the issues identified during the 1989 inspection was related to an apparent material formulation change in a terminal block (DWG 20885101) which functions as an electrical feed-through as well as an environmental barrier within the actuator electrical enclosure. Recent information supplied by BIP Chemical, Ltd. confirmed the formulation change and stated that, while the two materials were similar in some respects, overall, the materials were not considered similar to each other. No radiation resistance data was available for the substitute material. Rotork stated that testing was currently in progress to verify equivalency or establish qualification for the substitute material.
c was also determined during the last inspection that material specifications for several 0-ring seals were not sufficiently definitive with respect to material formulation to assure similarity between the 0-rings qualified by the Wyle Laboratories test and 0-rings purchased subsequently. The previously referenced consultant report confirmed this concern and recommended that replacement 0-rings be purchased to DOWTY grade 4460 or 4490 formulations (or their ASTM equivalents), both of which have been successfully qualified by the Wyle test.
While Rotork Technical Report ER 187 provides the methodology for an acceptable commercial grade item dedication program, at the time of this inspection Rotork had not developed sufficient instructions / procedures to fully implement this program. Because of this and the incomplete parts justifications, these unresolved items are still considered open.
(99901066/89-01-10, 89-01-11, 89-01-12) 4 PERSONNEL CONTACTED Rotork Controls Limited Martin Hunt, Quality Manager Ivan Burnell, Engineering Manager, Applications and Design Support Dennis Barnett, Purchasing Manager Geoff Hines, Contracts Manager Andrew Weeks, Supplier Quality Auditor
- Attended Exit Meeting
_7
-158-
/p* KIGg"'9, UNITED $T ATES 8
NUCLEAR REGULATORY COMMISSION -
/ ~I
,,- y
{
wAsmwatou. o, caosss t
09I 1932 Docket No. 99900404 Mr.
S. R. Tritch, Manager Nuclear Safety Department Nuclear and Advanced Technology Division Energy Systems Business Unit Westinghouse Electric Corporation Post Office Box 355 Pittsburgh, Pennsylvania 15230
Dear Mr. Tritch:
SUBJECT:
NOTICE OF NONCONFORMANCE (NRC INSPECTION REPORT 99900404/92-02)
This letter addresses the inspection of the Westiaghouse Process control Division (W-PCD) at Pittsburgh, Pennsylvania, conducted by Messrs.
R.
C. Wilson, R. N. Moist, R. K. Frahm, Jr., and R.
E. DePriest of this office on November 16-19, 1992,.and the discussion of their findings with the general manager of W-PCD and members of your staffs on November 19, 1992.
The purpose of the inspection was to review-previous inspection concerns regard-ing the dedication of commercial grade components for nuclear safety-related applications and actions taken by W-PCD relating to 10 CFR Part 21 reports.
Areas examined during the NRC inspection and our findings are discussed in the enclosed report.
This inspection consisted of-4 an examination of procedures and records, interviews with per-sonnel, and observations by the inspectors.
The inspectors found that the implementation of W-PCD's quality assurance program failed to meet certain ERC requirements.
Specifically, in six instances W-PCD' failed to document that possible changes-to purchased commercial grade replacement parts being dedicated for safety-related use were controlled, identi-fled, or evaluated for safety implications.
In four instances W-PCD failed to maintain records of dedication test results, and in two instances W-PCD did not perform dedication activities in accordance with approved procedures.
Although these are noncon-formances in your program which should be corrected affecting a totcl of 96 safety-related replacement parts, we note that they are limited to documentation problems,-and that W-PCD's dedica-tion program is now almost fully implemented.
The specific findings and references to the pertinent require-ments are identified in the enclosed Notice of Nonconformance (Notice).
-159-
[
Mr.
S._R.
Tritch The response requested by the enclosed Notice is.not subject to the clearance procedures of the Office of Management and Budget as required by the Paperwork Reduction Act of 1980, Public Law No.96-511.
In accordance with 10 CFR Part 2.790 of the NRC's " Rules of Practice," a copy of this letter and its enclosures will be placed in the NRC's Public Document Room.
Sincerel O
(
s LNif J N rrholm, Chie Vendor Inspection Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation
Enclosures:
1.
Notice of Nonconformance 2.
Inspection Report 99900404/92-02
-160-l 1
ENCLOSURE 1 NOTICE OF NONCONFORMANCE Westinghouse Electric Corporation Docket No.:
99900404/92-02 Process Control Division Pittsburgh, Pennsylvania Based on the results of an inspection conducted en November 16-19, 1992, it appears that certain of your activities were not conduc-ted in accordance with NRC requirements.
A.
Criterion III, " Design Control," of Appendix B to 10 CFR Part 50 requires, in part, that the design basis for safety-related components be correctly translated into specifications, drawings, procedures, and inntructions, and that measures be established for the review for suitability of application of parts and equipment that are essential to the safety-related functions of components.
Criterion V, " Instructions, Procedures, and Drawings," of Appendix B to 10 CFR Part 50 requires, in part, that activ-ities affecting quality shall be prescribed by documented procedures and shall be accomplished in accordance with these procedures.
Criterion VII, " Control of Purchased Material, Equipment, and Services," of Appendix B to 10 CFR Part 50 requires in part that measures for control of purchased equipment shall include provisions, as appropriate, for source evaluation and examination of products.
Page 6 of Procedure DP 07-003, " Commercial Grade Item (CGI)
Dedication - Nuclear," Revision 4, dated February 14, 1992, which is a part cf WCAP-12710/TP199, " Process control Division [W-PCD) Quality Assurance-Program," states that the quality assurance (QA) requirements of Appendix B to 10 CFR Part 50 apply to the dedication process.
Page 12 requires storing auditable documentation supporting the requirements of 10 CFR Part 50.
Contrary to the above, W-PCD failed to document that possible changes to dedicated commercial grade items were controlled, identified, or evaluated for safety impact in six instances.
In the first two instances, W-PCD-also failed to perform commercial grade vandor surveys specifi-cally required by the appropriate commercial dedication instruction.
The specific commercial grade items involved are (99900404/92-02-01):
-161-
l 1.
Forty-one high voltage power supplies shipped to the Westinghouse Nuclear Services Division (W-NSD) under W-PCD order number N2403 (W-NSD General Order
[GO) RPIN00085 dated August 7, 1992) 2.
Four printed circuit boards shipped to W-NSD under W-PCD order number N4890 (W-NSD GO RPIN00087 dated May 8, 1992) 3.
Twenty-five isolation relays shipped to W-NSD under W-PCD order number N4890 (W-NSD GO RPIN00087 dated May 8, 1992) 4.
Two fan assemblies shipped to the Byron Nuclear Power Station of Commonwealth Edison Company under W-PCD order number N2405 (W-NSD GO RPS22273 dated May 27, 1992) 5.
Five timer modules shipped to the Comanche Peak Steam Electric Station of Texas Utilities Electric Company under W-PCD order number N4925 (W-NSD GO PCD00051 dated August 14, 1992) 6.
Four blower assemblies shipped to the Joseph M.
Farley Nuclear Plant of Alabama Power Company under W-PCD order number N2379 (W-NSD GO RPS22170 dated March 31, 1992)
B.
Criterion XVII, " Quality Assurance Records," of Appendix B to 10 CFR Part 50 requires, in part, maintaining retrievable records that identify the results and acceptability of tests affecting quality.
Page 6 of Procedure DP 07-003, " Commercial Grade Item (CGI)
Dedication - Nuclear," Revision 4, dated February 14, 1992, which is a part of WCAP-12710/TP199, " Process Control Division Quality Assurance Program," states that the QA requirements of Appendix B to 10.CFR Part 50 apply to the dedication process.
Contrary to the above, W-PCD failed to maintain documented test results as part of the commercial grade item dedication process in four instances as follows (99900404/92-02-02):
1.
Four printed circuit boards shipped to W-NSD under W-PCD order number N4890 (W-NSD GO RPIN00087 dated May 8, 1992) 2.
Twenty-five isolation relays shipped to W-NSD under W-PCD order number N4890 (W-NSD GO RPIN00087 dated May 8, 1992) 2 l
-162-
3.
Five timer modules shipped to the Comanche Peak Steam Electric Station of Texas Utilities Electric Company under W-PCD order number N4925 (W-NSD GO PCD00051 dated August 14, 1992) 4.
Fourteen termination modules shipped to the Zion Nuclear Station of Commonwealth Edison Company under W-PCD order number N2393 (W-NSD GO IN18680 dated May 11, 1992)
C.
Criterion V,
" Instructions, Procedures, and Drawings," of Appendix B to 10 CFR Part 50 requires, in part, that activ-ities affecting quality shall be prescribed by documented procedures and shall be accomplished in accordance with these procedures.
Page 6 of Procedure DP 07-003, " Commercial Grade Item (CGI)
Dedication - Nuclear," Revision 4, dated February 14, 1992, which is a part of WCAP-12710/TP199, " Process Control Division Quality Assurance Program," states that the QA requfrements of Appendix B to 10 CFR Part 50 apply to the dedisation process.
Page 3 of W-PCD QA Procedere TP1018, " Commercial Dedication Instruction," Revision 1, dated January 31, 1992, for a commercial grade item being dedicated for safety-related use requires revision of the item drawing using a Development Engineering Order (DEO) and a DEO Notification (DEON),
I listing each critical characteristic and evaluation of changes to it on the DEON, and stamping "DEON CONTROLLED ASSEMBLY" or equivalent on the drawing sheets.
Contrary to the above, W-PCD failed to issue DEONs, provide the required evaluation, and stamp the item drawing in the following two instances (99900404/92-02-03):
1.
Five timer modules shipped to the Comanche Peak Steam Electric Station of Texas Utilities Electric Company under W-PCD order number H4925 (W-NSD GO PCD00051 dated August 14, 1992) 2.
Four blower assemblies shipped to the Joseph M.
Farley Nuclear Plant of Alabama Power Company under W-PCD order number N2379 (W-NSD GO RPS22170 dated March 31, 1992)
Please provide a written statement or explanation to the U.S.
Nuclear Regulatory Commission, ATTN:
Document Control Desk, Washington, D.C.
20555, with a copy to the Chief, vendor Inspection Branch, Division of Reactor Inspection and Licensee Performance, Office of Nuclear Reactor Regulation, within 30 days of the date of the letter transmitting this Notice of 3
-163-I 1
t Nonconformance.
This reply should be clearly marked;as a " Reply to a Notice of Nonconformance" and should include for each nonconformance: (1) a description of steps that have been or will ba taken to correct these items; (2) a description of steps that have been or will be taken to prevent recurrence; and (3) the dates your corrective actions and preventive measures were or will be completed.-
Dated at Rockville, 4aryland this ',' ] ** day of
)ct w Jy v 1992.
4
-164-
~.
ae n
ORGANIZATION -
. WESTINGHOUSE ELECTRIC; CORPORATION PROCESS / CONTROL DIVISION.
PITTSBURGH, PENNSYLVANIAz REPORT NO.:
- 99900404/92-02 t
. CORRESPONDENCE S. A.f Tritch',~.: Manager
~
= ADDRESS:
Nuclear S'afety DepartmentJ._
NuclearLand-Advanced.TechnologyJDivisioni Energy. Systems Business' Unit?
Westinghouse Electric Corporatione Post: Office l Box-355-Pittsburgh,LPennsylvania115230 ORGANIZATIONAL Philip,T.-McManuss QualityfAssuranceiManager CONTACT:
Process Control Division ~
F
~
- Westinghouse: Electric Corporation.
j 2000; Beta Drive-U
- O'Hara Township-Pittsburgh,-Pennsylvania 15238-1987z (412): 963-4815 NUCLEAR INDUSTRY
. Safety-related instrumentation and-control-ACTIVITY:
systems for commercial nuclear: power. plants-a INSPECTION November 16-19, 1992
~
. CONDUCTED:
TEAM LEADER:
92/
- Richard C. Wilson, Senior ~ Engineer
. Date-
~
Reactive Inspection.Section.2 (RIS2) s Vendor Inspection Branch (VIB)-
p OTHER INSPECTORS:
- Randolph N. Moist,-RIS2, VIB-Ronald K. Frahm,cJr., RIS2,--VIB
. Robert E. DePriest,tIntern APPROVED:
(Mr
/7 l'
Gregog g. Cwalina, Chief-
'D ta e.
Reactiv6 Inspection Section 2 Vendor Inspection. Branch-
. INSPECTION BASES:
. 10.CFR Part 21 and 10 CFR Partf50,-Appendix B INSPECTION SCOPE:-
To review: the dedication :of commercial; grades components for nuclear safety-related appli-cations, and to review previous-inspection concerns and 10 CFR~Part'21 notifications-PLANT. SITE Numerous APPLICABILITY:
-165-i w'
e
-er w.r Dr' irWG9 &
wr
?7wt 18' li' M
4 1
INSPECTION
SUMMARY
e
- 1.1 Nonconformangga 1.1.1 Nonconformance 99900404/92-02-01 (Onen)
Contrary toLCriteria III, V,.and VII of Appendix B to 10 CFR Part 50, and to Westinghouse Process Control' Division (W-PCD) Procedure DP 07-003, W-PCD failed to document that possible changes to the commercial grade items _being dedicated.
for safety-related use.in commercial nuclear power plants were=
controlled, identified,'or evaluated for safety impact in six instances (see Sections 3.3.1, 3.3.2, 3.3.3, this report).
3.3.4, and 3.3.5.of 1.1.2 Hgnconformance 99900404/92-02-02 (Onen)
Contrary to Criterion XVII of Appendix B to 10 CFR Part 50,1and to W-PCD Procedure DP 07-003, W-PCD failed to maintain records of-dedication test results for components supplied for-safety '
related use in commercial nuclear power' plants (see Sections 3.3.2, 3.3.4, and 3.3.6 of this report).
1.1.3 Nonconformance 99900404/92-02-03-(Onen)
Contrary to Criterion.V of Appendix B to 10 CFR Part 50, and to W-PCD Quality Assurance (QA) Procedure TP1018, W-PCD failed tct accomplish dedication _ activities in accordance with-established-procedures for. components supplied for safety-related use in~
commercial nuclear power plants (see Sections 3.3.4 and 3.3.5 of this report).
2 STATUS OF PREVIOUS INSPECTION FINDINGS 2.1 H2Dronformance 99900404/91-02-01 (Closed)
Nunconformance 99900404/91-02-01 stated that, contrary to-estinghouse procedures implementing Criterion V,
" Instructions, Procedures, and Drawings," of Appendix?B to 10 CFR Part 50,,W-PCD did not have documented instructionn or procedures to control.all-of the QA activities performed-in dedicating commercial grade items for safety-related use, and that the new procedure "TP1018 DRAFT" was not properly controlled.
At the time of this inspection, W-PCD-had completed issue of written procedures and instructions covering their-dedication program,' principally Section DP 07-003,_" Commercial Grade Item Dedication - Nuclear," Revision 4, dated February 14, 1992, of WCAP 12710/TP199, " Process Control Division-Quality Assurance-Program."
In addition, TP1018, " Commercial Dedication Instruc-tion," Revision 1,' dated January 31,_1992, had been formally 2
-166-
issued'as:a part of TP 1000, " Department Quality' Assurance Procedures, Nuclear Projects "'
The NRC inspectors did not identify'any instances where W-PCDl procedures failed to impose the QA requirements of Appendix B tol 10 CFR Part 50-with respect to dedication:of commercial' grade
-items'for safety-related applications, although the identified nonconformances clearly indicate that the' implementation of the requirements is still deficient.
Based on their review the-inspectors closed Nonconformance 99900404/91-02-01.
2.2 Unresolved Item 99900404/90-01-08 (Closed)
Unresolved Item 99900404/90-01-08 stated that the inspectors had-not reviewed W-PCD's program for complying with-the reporting requirements of 10 CFR Part 21.
During the present inspection the NRC inspectors discussed W-PCD's 10 CFP.'Part 21 program with personnel from both W-PCD and the Westinghouse ~ Nuclear.and Advanced Technology Division (W-NATD), and reviewed applicable-prc.cedures.
W-PCD letter to the NRC dated April 24, 1992,-re-ported a new possible Part 21 concern,_'and-it was reviewed as an implementation example for W-PCD's Part 21 policy as it functions-in conjunction with W-NATD.
The responsible Westinghouse official is Mr. N.
D. Woodson, Vice President and General Manager of the Westinghouse Energy Systems Business Unit (ESBU), which includes both-PCD and NATD..The applicable controlling procedure is OPR-19.0, " Identification and Reporting of Conditions Adverse to Safety," Revisjon-4, dated October 1, 1992, include,d in WCAP-9550, the ESBU " Quality Assur-ance Program for Basic Components."
From November 1991 untilf October 1992 an interim letter was in effect.
OPR-19.0 requires-a division implementing procedure, which for W-PCD'is DP 19.0,
" Identification & Reporting of Substantial Safety Hazards; j
Significant Deficiencies and Unreviewed Safety Questions,"
Revision 6, dated October 1,-1992, included in WCAP 12710/TP199.
The NRC inspectors reviewed these procedures.
They. invoked the current issue of Part 21, important-terms'were correctly defined and the instructions for employee and management activities appeared to be adequate.
The inspectors noted that,1where a Part 21 concern originates in W-PCD, the. concern is acted on by the ESBU Safety Review Committee and reported by W-NATD.
Westinghouse NATD letter ET-NRC-92-3692 transmitted an " Interim-Report of. Evaluation of a Deviation or Failure to Comply Pursuant-to 10CFR21.21(a) (2)" dated April 24, 1992, concerning type 7300 circuit boards and piece parts that may not have been properly dedicated.
During the inspection, W-NATD issued another letter, ET-NRC-92-3771 dated November 13, 1992, describir g - the. evaluation performed by Westinghouse, and concluding.that thi issue was not reportable under 10 CFR Part 21.
The NRC inspectors discussed 3
-167-
'the evaluation with Westinghouse personnel, and concurred that-the evaluation conducted by Westinghouse _was adequate, and_that
-c11 of.the parts were in fact qualified. 'The inspectors, stated that Westinghouse was nonetheless correct in reporting-the possible concern.
The November-NATD letter committed to'" continued corrective a
actions to assure that any vendor or part changes are reviewed-for seismic qualification," but Nonconformance 99900404/92-02-01 of this inspection report identifies six instances where cui eval--
_ ith-the exception;of:
uation of such changes was not documented.
W that concern, which will be followed separately, the inspectors ~
concluded that W-PCD is properly handling 10 CFR-Part 21 con-cerns and closed Unresolved Item 99900404/90-01-08.
2.3 Open Item 99900404/90-01-10 (Closed)
Open Item 99900404/90-01-10 stated that Westinghouse had not completed investigating aiconcern involving electrical isolation between protection and control-circuits'on type 7300 NCT circuit boards, reported by Duquesne Light Company as Licensee' Event-Report (LER) 90-022-00 for the Beaver Valley Power Station.
Westinghouse then determined that the problem only affected certain-plants where the non-safety channel test' bypass'annun-clator output signal'for the energize-to-actuate-containment spray function was not routed through a' qualified isolator card.
The problem had been identified in 1980, but apparently was.not=
corrected for equipment that had already been shipped.
In March 1991 Westinghouse notified'all licensee; customers offthe-concern.
Affected licensees were provided a basis for interim operation involving similarity of'the untested isolator to qualified devices,_and suggested possible long term' corrective actions.
Westinghouse ESBU l'etter NS-NRC-91-3590 dated May 15, 1991, advised the NRC that the issue was closed based on a determination that it was not reportable because'the-isolation function-most likely_would be performed, and.the reasonable:
assurance of safe operation provided to the affected: plants.
Since 1980, Westinghouse has changed evaluation-procedures.to prevent recurrence of this type of concern.
The; inspectors-reviewed an evaluation form used by W-PCD in the= current review as required by the Procedure and Guidance Manual for the W-NATD nuclear safety department.
The new 10 CFR Part 21 procedures:
described'in Section 2.2 of this report also improve Westing :
house's ability to track potential safety _ issues.
Based on-this review the' inspectors closed'Open Item 99900404/90-01-10.
2.4 ppen Item 99900404/91-02-03 (Closed)
Open Item 91-02-03 stated that W-PCD had not completed responding to findings from a recent audit by the Westinghouse' Nuclear 4
-168-
S rvicos Divicion, and had not complotcd-implcm:ntation of up-
. grades to its QA program resulting from that audit and~another by a joint utility group.
The inspectors verified that since the last NRC inspection W-PCD had revised their procedures and insti-
.tuted training to address the open findings.
In addition, the-inspectors reviewed a letter from the lead Westinghouse-auditor dated February 6, 1992, that closed all eighteen findings and twelve observations.
The inspectors also reviewed the report of an audit performed in February 1992 by the joint utility group which verified that W-PCD had performed all required corrective actions resulting from the group's previous audit.
Based on this information the inspectors closed Open Item 99900404/91-02-03.
2.5 Open Item 99900404/91-02-04 (Closed)
Open Item 91-02-04 stated that W-PCD was not adequately verifying critical characteristics by means of commercial grade vendor sur-veys because the surveys were not specific to the characteristics being verified and the surveys were infrequently performed.
W-PCD had performed several commercial grade vendor surveys prior to the inspection.
W-PCD personnel stated that surveys are based on critical characteristics, and are not used as the sole method of dedication.
The NRC, inspector selected. and reviewed two commercial grace surveys that W-PCD conducted during 1992.
The first survey was performed August 1992, at Airpax Co., Cambridge, Maryland, a supplier of circuit breakers, and the second January 1992, at I
ASTEC America Incorporated, Oceanside, California, a' supplier of-power supplies.
The inspectors determined that both surveys verified supplier controls affecting critical characteristics.
The inspectors reviewed W-PCD's QA Commercial Grade Supplier List (CSGL) dated November 16, 1992.
It showed the survey frequency for each supplier based on supplier performance, item complexity, standard receipt inspection results, and procurement frequency.
The survey frequencies are mostly one to two years.
W-PCD uses Vendor Quality Specifications (VQSs) for vendors that supply commercial grade components to W-PCD.
The VQS establishes the requirements for 10 CFR Part 21 applicability, supplier requirements, W-PCD receipt inspection requirements, and W-PCD dedication requirements.
After a first-time procurement of com-mercial grade components from a supplier or initial survey of a supplier, W-PCD QA develops a VQS which is incorporated into the CGSL for that supplier.
The VQS identifies the QA requirements and elements of the supplier's QA program applicable to the items or services procured.
The W-PCD QA group generates, distributes, and controls the VQS.
W-PCD is currently developing or revising VQS's for the commercial grade items used in systems, assemblies and subassemblies.
5
-169-t I
i
Baccd on~thic rcvicwitho_31ncptctore closrd Op;n:ItCm:91-02-04.
._However, the inspectors. identified numerous examples:of-Noncon-formance-99900404/92-02-01, indicating =that W-PCD_needs to con s
-tinue their emphasis on-conducting commercial supplier surveys.
3 INSPECTION FINDINGS AND OTHER' COMMENTS 3.1 Entrance and Exit Meetinas In the entrance meeting on November 16, 1992, the NRC' inspectors.
discussed the scope of the inspection, outlined _the areas to be inspected, and established interfaces with W-PCD management'and staff.
In the exit meeting on November 19, 1992, the inspectors discussed their findings and concerns with W-PCD management and staff.
3.2 Inspection Scone W-PCD manufactures instrumentation and control systems for safety-and nonsafety-related applications. 'The safety-related:
scope was transferred from another Westinghouse division in 1989.-
Previous NRC inspections have monitored'the preparation and-implementation of W-PCD's safety-related-QA program.
The major new area remaining for_this inspection was the W-PCD program to j
dedicato commercial grade items for safety-related applications.
The procedures for W-PCD's commercial grade dedication ~ program l
were completed and in effect at the time of the= inspection.
l W-PCD had completed approximately 108 of 182 commercial dedica-tion instructions.
W-PCD had conducted several commercial grade supplier strveys, and scheduled several more for 1993.
The NRC l
Jnspectors reviewed the dedication program procedures, and selec-ted and examined several packages representing specific dedica-i tion activities, as well as closing the concerns.from previous l_
inspections that-are addressed in Section 2 of this report.
3.3 Dedication Packaoe Reviews Topical Report WCAP 8370/7800, " Westinghouse Electric Corporation Energy Systems Business Unit (ESBU]/ Nuclear Fuel Business' Unit Quality Assurance Plan," Revision 11A/7A, dated December'1988, committed ESBU, of which W-PCD is a part, to a QA' program meeting-the criteria of Appendix B to 10 CFR Part 50.
WCAP-9550, " Energy l
Systems Business Unit Quality Assurance Program for Basic Compo--
nents," Revision 24, dated October 1, 1992, and WCAP 12710/TP199,
" Process Control Division Quality Assurance Program,"~ Revision 11, dated November 16, 1992, imposed these criteria on W-PCD.
i-L In addition, for each dedication package reviewed, the General
~
Order (GO) from the Westinghouse Nuclear Services. Division-I (W-NSD) -imposed the requirements of 10 CFR Part 21 and -stated 6
-170-
h that th3 procuremsnt was2 quality codo A1(ccfoty-rclated). TEcch' GO.also-invoked W-NSD Standard QA Requirements SQAR 101,~ Revision 5,-and Appendix Af Revision 3. -
Appendix A referenced two other-Westinghouse quality system requirements documents,1QPS-320-1 Revision 2,.and_QCS-1 Revision 7, which incorporated the requirements of American Nationalub?.andards Institute'(ANSI) standard N45.2-1971 " Quality Assurance Program Requirements for Nuclear Power Plants."
3.3.1 Order N2403, High Voltage-Power Supplies W-PCD order number N2403 covered-41 nuclearLinstrument-(intier-mediate range) high voltage power supplies delivered to W-NSD-under their-GO RPIN00085 dated August 7, 1992.
The GO specified" style number 3A56943G02, designating Group 2.cn1 W-PCD drawing-3A56943, which specified part number PS20005H04 as item 2 and group 2.
The GO also specified part number UPMD-X 54W,= which was the supplier's catalog number for the power-supply.
Note 1.of Drawing 3A56943 also specified that the item 2 power supply should be inspected and tested per procedure 3A56944.
Develop-ment Engineering Order (DEO) No. 97283 dated June 10, 1992, controlled by DEO Notification (DEON) dated June 10, 1992, revised the drawing to-add Commercial Dedication Instruction (CDI) No. 4A07794 for the group 2 power supplies.
Vendor-Quality Specification (VQS) 042 designated part number PS20005H04 as catalog number UPMD-X54W from Power Designs Inc.
As part of the dedication, CDI No. 4A07794 referenced report WCAP-8687, Supplement 2-E47C, Revision 1,
" Equipment Qualifi-cation Test Report, Nuclear Instrumentation System Tour-Bay Cabinet, Boron Dilution Source Range and Intermediate Range Drawer Assemblies (Environmental and Seismic Testing)," datedL April 1985.
This document identified the high voltage power supply in the intermediate range drawer as position-NQ201, and also referenced intermediate range schematic-drawing 6081D151, Revision E, which showed the intermediate range high voltage-power supply NQ201 as part number 2384A23H04.
Drawing 2384A23 sheet 5 identified part number 2384A23H04 as vendor part number UPMD-X54 W.
This number agrees with the vendor part number.
dedicated in the 1992 dedication activity.
The documentation trail just described established that the com-mercial grade part ordered-was the same catalog number as the part that was originally qualified in 1982 and 1985 qualification reports.
However, W-PCD's dedication activities'did not show that commercial grade power supplies provided by Power Designs Inc. under catalog number UPMD-X54W are' adequately _ controlled to-ensure that the recently dedicated items can' perform the same-safety functions as the qualified items in the same environments.
CDI 4A07794 defined three types of critical characteristics:-
(a) The part number was verified by inspection, and the NRC 7-
-171-I Y
l inspector observed cppropriato inspection dato sheets.
(b) Qual-ification for seismic and mild environments was verified by 1
reference to the qualification test reports, and the CDI stated that a " commercial survey of the vendor for design control helps ensure the changes in the power supply do not affect the seismic qualification."
(c) W-PCD testing in accordance with procedure 3A56844 covered burn-in, voltage and current rating, and other tests,~and appropriate test data sheets for each power supply by serial number were included in the file.
W-PCD, in fact, had not performed a commercial grade survey of Power Designs Inc.
A W-PCD internal memo from engineering t: QA dated August 7, 1992, referred to a future survey, and a schedule provided to the inspectors listed Power Designs Inc. for February 1993.
However, the power supplies under W-PCD order number N2403 were shipped in September 1992 and certified as meeting all requirements of customer order XARN90288 on September 21, 1992.
The NRC inspector. reviewed Westinghouse audits of Power Designs Inc. performed in 1990 and 1991.
W-PCD personnel agreed that these audits do not serve the purpose of a commercial grade supplier survey addressed on page 7 of Procedure DP 07-003.
In addition to the absence of a commercial grade supplier survey, the dedication tests performed by W-PCD were not sufficient by themselves to establish similarity between the dedicated items and the qualification test specimens.
Further, there was no documentation that possible changes to the commercial grade power supplies were controlled, identified, or evaluated for safety impact.
This failure to document that possible changes to the commercial grade items being dedicated were controlled, identi-fled, or evaluated for safety impact is the first example of Nonconformance 99900404/92-02-01, 3.3.2 Order N4890, Printed Circuit Boards and Isolation Relays W-PCD order number N4890 covered 12 line items, including four type NTC printed circuit (PC) boards (line item 6), part number 2837A94G08, shipped to a W-NSD warehouse under W-NSD GO RPIN00087 dated May 8, 1992.
Change Notices 1 and 2 added line item 13 for 25 Struthers-Dunn Co. relays, part number PS12805H01. The inspec-tors reviewed both the item 6 boards and the item 13 relays.
Printed Circuit Boards (line item 6)
Assembly drawing 2837A94, supplemented by DEON, references CDI 4A07736 and test procedure 953A80.
The critical characteristics defined in the CDI included seismic integrity and mild environ-ment qualification as verified by reference to the 1986 qualifi-cation test report for the 7300 Series Process Protection System, with the statement that changes are evaluated under DEO/DEON con-trol.
No documentation indicated that a review was performed to verify whether possible part changes made since the original 1986 8
1
-172-l J
c31cmic cnd mild cnvironmnnt qualification waro ovaluntad.
Thrra was no evidence of a commercial grade survey of the supplier, and the dedication testing described in the next paragraph was not-sufficient to detect all such changes.
This failure to document that possible changes to the commercial grade items being dedi-cated were controlled, identified, or evaluated for safety impact is the second example of Nonconformance 99900404/92-02-01.
Other critical characteristics defined in CDI 4A07736 included repeatability and operability, to be verified as part of PC board-level testing in accordance with test procedure 953A80.
No hard copy record or any other form of documentation indicated that these critical characteristics were verified by testing, or-what results were found.
W-PCD's practice for PC boards, as detailed in Procedure TP 5000, Section 5302, " Card Test," Revision 1, dated October 18, 1991,.is to mark the card edge of boards that passed tests with a desig-nated semi-permanent marker.
The quantitative test results are recorded on the QC Card Test Data Collection Product Assurance data log form, which is used for data collection and evaluation by the Product Assurance Department.
The data are entered into the QA data base for the purpose of trending and other analyses, and eventually the data log forms are discarded.
For serialized PC boards, test results are also recorded on test data sheets that are retained by Product Assurance.
However, for less com-plex, non-serialized boards, the test procedures do not require officially recording test data or retaining hard copy records.
l As part of the CDI justification, the final inspector simply ver-ifies that appropriate tests were performed by observing whether the proper test markings are on the card's edge.
Failure to maintain adequate test records is the first example of Nonconformance 99900404/92-02-02.
Isolation Relays (line item 13)
Assembly drawing 3A98884, supplemented by DEON, references CDI 4A07847.
The critical characteristics defined in the CDI include seismic integrity and mild environment qualification as verified by reference to a 1986 qualification test report for Struthers-Dunn relays, with the statement that a commercial survey of the vendor for design control helps ensure that changes to the relay do not affect the seismic qualification.
The schedule provided to the inspectors indicated that W-PCD had neither surveyed Struthers-Dunn nor scheduled a survey.
There was no documenta-tion that any review was performed under the DEON control system of possible part changes made since the original 1986 seismic and mild environment qualification, and the dedication testing performed was not sufficient to detect possible changes.
This failure to document that possible changes to the commercial grade items being dedicated were controlled, identified, or evaluated 9
-173-u
for ecfety impact is the third excmplo of Nonconformanca 99900404/92-02-01.
CDI 4A07847 also identified isolation as a critical characteris-tic, but not in the critical characteristics portion of the CDI (section E) ; however, isolation was listed in the table in-section H of the CDI.
The CDI did not identify operability as a critical characteristic.
Test procedure 953A20, which was not referenced in the CDI, was referenced by the assembly drawing 3A98884; the test procedure verified operability of the relay by confirming minimal leakage current using a hi-pot tester and by verifying proper contact status for the " operate" and " reset" functions.
Although the W-PCD Certificate of Conformance dated September 3, 1992, stated that the relays were tested in accord-ance with test procedure 953A20, Revision 1, W-PCD personnel agreed with the inspector's comment that isolation and opera-bility should be clearly identified in the CDI as critical characteristics.
W-PCD initiated quality performance feedback QPF 94-060 during the inspection to correct the documentation.
No hard copy record or any other form of documentation indicated that the relays were tested and what results were found.
The final inspector simply verified that testing was performed by observing whether the proper test sticker, which indicated the date of test and initials of the test operator, was affixed to the relay.
Since the sticker was shipped with the relay, no documes. cation was retained to support test acceptance.
This lack of adequate test records is the second example of Nonconformance 99900404/92-02-02.
One relay failed during dedication testing due to high contact resistance.
W-PCD initiated discrepant item notice DIN 231697 in accordance with TP121, " Quality Assurance Inspection Handbook,"
Revision 3, dated August 31, 1992.
The rejected relay was scrapped.
A replacement passed testing and was shipped.
3.3.3 Order N2405, Fan Assemblies W-PCD order number N2405 covered two type 2D34586G01 fan assem-blies shipped to the Byron Nuclear Power Station of Commonwealth Edison Co. under W-NSD GO RPS22273 to W-PCD dated May 27, 1992.
W-PCD drawing 2D34586, Revision 2, as modified by DEON 097510, references CDI 4A07860.
The fan assembly includes a fan, part number 9936A43H01, vendor model MB5100 Type 100,. procured com-mercial grade from EG&G Rotron Co. as their part number 020099.
The critical characteristics defined in the CDI include seismic integrity as verified by reference to the original qualification test report for the 7300 Series Process Protection System, with the statement that changes are evaluated under DEO/DEON control.
No documentation indicated that a review was performed to verify whether fan changes possibly made since the original 1986 seismic 10
-174-
qualification waro adequatoly:ovaluated.- Thsro was no ovidznco of a commercial grade survey of the supplier, and.the~ dedication testing was not. sufficient to detect-all such changes.
This failure to document that possible changes to the commercial grade items being dedicated were' controlled, identified, or evaluated for safety impact is the fourth example of Nonconformance 99900404/92-02-01.
3.3.4 Order N4925,-Timer Modules W-PCD order number N4925 covered five timer module printed-circuit (PC) cards, part number 2839A49G04, for use on type NPL PC boards.
The cards were shipped to the Comanche Peak Steam Electric Station of Texas Utilities Electric Co. under W-NSD GO PCD00051 dated August 14, 1992.
Assembly drawing 2839A49 did not contain a notice requiring DEON control of the drawing..
W-PCD QA Procedure TP1018, " Commercial Dedication Instruction,"
Revision 1, dated January 31, 1992, requires stamping "DEON CONTROLLED ASSEMBLY" or equivalent on drawing sheets for dedi-cated commercial grade items.
It also requires that the DEON list each critical charactoristic and evaluation of changes to it.
Drawing 2839A49 was changed by DEO 052341 dated September 11, 1992, without DEON control.
There was no documentation showing that the change was analyzed for impact on seismic qualification or other safety concerns.
This lack of DEON control is the first.
example of Nonconformance 99900404/92-02-03.
Based on the inspectors' concerns, during the inspection W-PCD issued QPF 92-058 and QPF 92-059 to initiate corrective action to properly mark the assembly drawing and to issue a DEON to.accom-
~
pany the previously issued DEO 052341.
The inspectors also noted-that assembly drawing 2839A49 did not cross-reference CDI.4A07784 as required by TP1018 page 3.
W-PCD issued QPF 92-057 as a result of this NRC concern to assure proper corrective action.
The critical characteristics defined in CDI 4A07784 included seismic-integrity and mild environment qualification as verified by reference to the 1986 qualification test report for the 7300 Series Process Protection System, with the statement that changes are evaluated under DEO/DEON control.- However, no documentation indicated that a review was performed-to verify whether design changes.possibly made since the original 1986 qualification were evaluated under the DEON control system.
There was also no evidence of a commercial grade survey of the supplier, and the dedication testing was not sufficient to detect all such changes.
This failure to document that-possible changes to the commercial grade items being dedicated were controlled, identified, or evaluated for safety impact is the fifth example of Nonconformance 99900404/92-02-01.
11
-175-
f Anoth r critical chtrccteriotic dofin;d in CDI'4A07704 wO0 ro-pratability, to be verified using the' module test procedure dstalled on sheets 13 through 15 of assembly drawing 2839A49.
A3 for other PC boards, as discussed in Section~3.3.2 of this report, no hard copy record or any other form of documentation indicated that this critical characteristic was verified by testing or what results were found.
This lack of adequate test-records is the third example of Nonconformance 99900404/92-02-02.
3.3.5 Order N2379, Blower Assemblies W-PCD order number N2379 covered four type 1864E25G01 blower as-semblies shipped to the Joseph M. Farley Nuclear Plant of Alabama Power Company under W-NSD GO RPS22170 dated March 31, 1992.
As-sembly drawing 1864E25 did not have a DEON control notice.
Draw-ing 1864E25 was changed by DEO 097489 dated September 11, 1992, without DEON control, and there was no documentation showing whether the change was analyzed for impact on seismic qualifi-cation or other safety concerns.
This lack of DEON control is the second example of Nonconformance 99900404/92-02-03.
Based on the NRC inspectors' concerns, during the inspection W-PCD issued QPF 92-058 to properly mark the assembly drawing and to issue a DEON to accompany the previously issued DEO 097489.
QPF 92-059 and DEO/DEON #097627 were also issued to add the "DEON Controlled" statement to the assembly drawing 1864E25.
The critical characteristics defined in CDI 4A07782 inc?.ude sels-mic integrity as verified by reference to the 1988 qualification test report for the Eagle 21 Process Control System, w'.th the statement that changes are evaluated under DEO/DEON cr.ntrol. The test data sheets for the blower assemblies included frocess line tags indicating completion of hardware inspection ar.d testing.
However, no documentation indicated that a review '<as performed to verify whether possible design changes made since the 1988 qualification were evaluated under the DEON control system.
There was no evidence of a commercial grade survey of the sup-plier, and the dedication testing was not sufficient to detect all such changes.
This failure to document that possible changes to the commercial grade items being dedicated were controlled, identified, or evaluated for safety impact is the sixth example of Nonconformance 99900404/92-02-01, 3.3.6 Order N2393, Termination Modules W-PCD order number N2393 covered 14 termination modules shipped to the Zion Nuclear Station of Commonwealth Edison Company under W-NSD GO IN18680 dated May 11, 1992.
The termination modules are manufactured at PCD using commercial printed circuit boards and component piece parts, and tested to the requirements of test procedure 1A90652 as required by CDI 4A07783.
Sections 4-4 and 4-5 of Procedure TP121, " Quality Assurance Inspection Handbook,"
12
-176-
R;vicion 3, d;tcd August 31, 1992, rcquiro attcching C, proc 200
-rccord tcg to ccch-cco:mbly or cub-ccccmbly manuf teturcd by W-PCD.
Section 14 of TP122, " Product Assurance Manual," Revision.
3, dated November 30, 1991, requires that all assemblies except printed circuit assemblies and cable assemblies have a " Process Line Tag" attached.
The process line tags are the normal evi-dence of testing but they could not be located, so only the inspector's signature on_the commercial dedication inspection data sheets verified that.the required testing was performed.
This lack of adequate test records is the fourth example of Nonconformance 99900404/92-02-02, 3.3.7 Order N2389, Power Supply W-PCD Order N2389 covered one power supply shipped to the North Anna Station of Virginia Power Company under W-PCD GO RM36205 dated April 30, 1992.
In contrast to the examples of Noncon-formance 99900404/92-02-01, this dedication activity included considerable effort to identify and evaluate possible changes in a commercial grade item.
W-PCD analyzed the piece parts and materials of the new power supply to determine differences from the originally qualified unit, and documented the results in a report.
The coil was found to be mounted difierently on the power supply converter board.
The new design was seismically evaluated, and the analysis was documented.
The seismic evaluation concluded that the coil and its mounting were acceptable for the seismic requirements.
+
13
-177-s
4 PERSONNEL CONTACTED
?-lQQ:
A.
E. Pauley, General Manager
+
R. L. Loving, Acting Manager, Customer Sys. and Projects
+
E. Toires, Manager, Total Quality
+
R. A. Judd, Acting Manager, Operations
+
K. A. Wells, Manager, Purchasing
+
W. L. Miller, Manager, Nuclear Applications
+
A.
P. Sahasrabudhe, Manager, Special Products & Applic.
+
P. T. McManus, Manager, Quality Assurance
+
L. Kamenicky, Acting Manager, Quality Assurance
+
J. C. McCann, Manager, Test Engineering
+
J. Yurechko, Manager, Just In Time Process Engineering
+
S. Breznican, Manager, Product Support Engineering
+
F.
S. Davis, Senior Engineer, Quality Assurance
+
M.
J. Laubham, Senior Engineer, Quality Assurance
+
R. Stein, Senior' Engineer, Quality Assurance
+
R. M. Roth, Senior Engineer, Quality Assurance
+
F.
A. Postava Jr., Engineer, Quality Assurance
+
M. McCrady, Senior Engineer, Quality Assurance T. L. Marts, Total Quality and Strategic Programs W.
Ritter, Senior Engineer R. Jacobovitz, Senior Engineer L.
Gaussa Jr.,
Senior Engineer P.
Federico, Engineer M. Janosik, Manager, Printed Circuit Assembly R. Evans, Manager, Mechanical / Electrical Repair Area A. Lamanski, Test Technician J.
E.
Leyland, Inspector C.
Riley, Test Engineer S. Martin, Senior QC Inspector W.
Deutsch, Senior QC Inspector J. Duffy, Test Technician R. Stoffiere, Senior Procurement Specialist W-NUCLEAR AND ADVANCED TECHNOLOGY DIVISION:
+
R. D. Petrosky, Manager, Quality Assurance
+
P. J. Morris, Secretary, Safety Review Committee
+
R.
B. Miller, Fellow Engineer
+ Attended the entrance neeting on November 16, 19S2 Attended the exit meeting on November 19, 1992 14
-178-
LSolect d Bullstin3, Gsnnric LGttora, and Information Noticco T
.Concerning' Adequacy:of-Vendor Audits and Quality of Vendor Products ISSUED-TITLE 1.
Bulletin 90-01 Loss of Fill Oil in Supplement 1 Transmitters Manufactured-by-
~
Rosemount 2.
Generic Letter 92-08 Thermo-Lag-330-1 Fire Barriers 3.
Information Notice 92-77 Questionable Selection-.And Review to-Determine.
Suitability of Electropneumatic Relays for-'
Certain Applications 4.
Information Notice 92-78 Piston to Cylinder Liner Tin Smearing on Cooper-Bessemer KSV Diesel Engines Il 5.
Information Notice 92-81 Potential' Deficiency of Electrical Cables With Bonded Hypalon Jackets 6.
Information Notice 92-83 Thrust Limits for Limitorque-.
Actuators and Potential overstressing of Motor-Operated Valves o
-179-
l CORRESPONDENCE RELATED TO VENDOR ISSUES
-180-
y
/
'c UNITED $TATEs l'
Y,i NUCLEAR REGULATORY COMMISSION e
{
.,I wAssmotow. o. c rosss
.,I 00T 1 8 1972 Docket No. 99901020 Mr Clayton W. Brown Consultant Brand Utility Services, Inc.
7?l Racquet Club Drive Addison, Illinois 60101
Dear Mr. Brown:
SUBJECT:
RESPONSE TO OCTOBER LETTER By letter dated October 14, 1992, you requested a clarification of the U. S.
Nuclear Regulatory Commission's (NRt1 role in reviewing nuclear contractors and their technology, especially the ature of the 1987 NRC inspection of Brand Utility Services, Inc. (BRAND).
The t RC does not require that vendors obtain NRC approval before supplying corronent,s or services to the U.S. nuclear industry.- Vendor approval is the responsibility of the nuclear utilities licensec and regulated by the NRC.
The NRC fren e tiy kspects Vendors to evaluate safety issues and to ascertain the ef fectiveness of their quality assurance (QA) programs in relation to licensee purchase orders subject to NRC requirements contained in Title 10 of the Code of Federal Reaulations (10CFR), Part 21 and Part 50, Appendix B.
If and v. hen the f1RC determines that a vendor has supplied components that are umafe, the NRC notifies potentially affected nuclear licensees.
The IEC performed
.n inspection at your facility in January 1987 and concluded, in a letter and Notice of Violation (Notice) dated May 22, 1987, that BRAND was in violation of 10 CFR 21.21. Specifically, BRAND had failed to develop procedures for reporting defects or deviations in components supplied to nuclear licensees. Because the violation was considered to be of minor significance, it was assigned a Severity 1.evel V classification. As required, BRAi.D' responded to that Notice in a letter dated June 4, 1987.
In a letter dated July 13, 1987, the NRC notified BRAND that it had reviewed the reply and found the corrective actions responsive to the concerns raised in the fictice.
BRAND satisfied the NRC with respect to the matters examined durir.g the inspection. The NRC did not conclude that any BRAND component was defective or unsafe.
URC vendor inspections are frequently prompted by an event or allegation which raises concerns of potential safety significance. There have been no events or allegations to indicate that BRAND has supplied poor quality fire seals to the U.S. nuclear industry.
-181-
Mr. Clayton W. Brown Should you have any further questions, please contact Mr. Ronald Frahm, Jr. of my staff at (301) 504-3216.
Sincerely, q
y.
-a,i Leif J. No holm, Chief Vendor Inspection Branch Division of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation
-182-
]
l
.i_J.i..i
,..i...
~ - - ' - ~ - - ~ - - - ~ -
- .Cso w 335
- U S kVCLt AR LtGULt.TO;Y COMul55 TON I WE PORf Nvv81R l
Ab 7* 1"E*ll" h'"I' ** -
in.
mwo BIBLIOGRAPHIC DATA SHEET is=,,i.uwe.a.
,w. eni NUREG-0040 t :iitt ANO susitt Vol. 16, No.-4 Licensee Contractor and Vendor Inspection 3
oats atPORievatissio Status Report l
c i-n aa 4
Quarterly Report January 1993 October - December 1992 uiN OR GRANT Nuva:R 6 AU1MORl51 614P107 RtPQR!
Quarterly
- 7. PE RIOD COV E R L O varsu e o i octot4r. December 1992 8 PER$ORMING ANil AllON - Naut AND ADORES $ ti, hac. am me o=
a o,r
., a us nue=, a., a,
, c n
.ne na,.se, se
,.ree,,,
D3'I[fsfon of Reactor Inspection and Licensee Performance Office of Nuclear Reactor Regulation U.S.
Nuclear Regulatory Commission Washington, D.C.
20555
- 9. $PONSORING ORG AN12 Af TON - N AM E 4o r ')OR E 55 fir mac. erre 1.aw m eeen'.,s te.eeeree,. wimee mac o
- a. on se er any.a. u s muce., a,wes,, c.===u.i and me.6me eds,es,I Same as above 10 $UPf'L(M(pet ARY NOT(5
- 11. A85T R ACT uco.4,e e,==>
This periodical covers the results of inspections performed by the NRC's Vendor 17spection Branch that have been distributed to the inspected organizations during the period from October through December 1992 ia xty wCRoseOtscR:Prossa e,,,-,-
u g.p g a,tgi~i Vendor Inspection
- 14. $t Cunit y CLAMaelCAliON
""tffdlassifled "Tn*c'"lassifled
- 16. NUwSER OF PAGES
- 16. PRICE
' O.RC $ ORM 3M (2 491
Printed on recycled paper Federal Recycing Program
-m-i UNITED STATES..
SPECIAL FCUPTM.MS RATE NUCLEAR REGULATORY COMMISSION restact ano rtrs emo "S"*C
'. WASHINGTON; D.C. - 20555-0001
- e, ;4 5, L = Q 3 C SA;."* --t1 2
2ee
" y renwr we c47 i.~
,l a -
77V F O i a p.
nrtgt7pg77pp#
e
'y
. omCIAL BUSINESS IUS*PE2-?!U"'EIs
. PENALTY FOR PRWATE USE, $300 7,
l I
l.
~
- m. _ __ _