ML20212C856
| ML20212C856 | |
| Person / Time | |
|---|---|
| Issue date: | 10/28/1997 |
| From: | NRC (Affiliation Not Assigned) |
| To: | |
| Shared Package | |
| ML20212C791 | List: |
| References | |
| REF-QA-99901052 99901052-97-01, 99901052-97-1, NUDOCS 9710300198 | |
| Download: ML20212C856 (17) | |
Text
- - -.
e U.S. NUCLEAR REGULA10RY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION Report no:
99901052/97-01 Organization:
Nova Machine Products Corporation
Contact:
James Fitzwilliam, Quality Assurance Manager (216)267 3200 Nuclear Industry Activity:
Manufacturer of threaded fasteners and supplier of ferrous and nonferrous bars, threaded fasteners, fittings, flanges, and other products used primarily in nuclear applications.
Dates:
June 2 through 5,1997, and October 2,1997 Inspectors:
Uldis Potapovs, Senior Reactor Engineer Larry L. Campbell, Reactor Engineer Gregory C. Cwalina, Senior Operations Engineer Donald Naujock, Materials Engineer Approved by:
Robert A. Gramm, Chief Quality Assurance Section Quality Assurance, Vendor inspection and Ms...tenance Branch Division of Reactor Controls and Human Factors 9710300198 971028 PDR GA999 EMVNOVA 99901052 PDR
4 lo **
1 INSPECTION
SUMMARY
During the June 2 5,1997, part of the inspection, the NRC inspectors reviewed the implementation of selected portions of Nova Machine Products Corporation's (Nova) quality assurance (QA) program for supplying material under the American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code (Code) requirements and for providing dedicated commercia' g ade material in compliance with the requirements of 10 CFR 50, Appendix B. The inspectors also reviewed the implementation of Nova's program for reporting of defects under the requirements of 10 CFR Part 21. This inspection was a part of an NRC pilot program to develop inspection procedures for assessing licensee oversight of supplier quality.
An inspection report was issued to Nova on July 22,1997. However, by letter of August 18,1997, Nova raised concems over potentially misleading statements in the report. Therefore, the NRC withdrew the initial inspection report pending a review of Nova's concerns. That review was completed on October 2,1997. D.; ring the October 1997 inspection, the inspectors further reviewed Nova's commercial grade dedication procedures and their adherence to customer and material specification requirements, particularly as related to small parts. In addition, the inspectors reviewed Nova's implementation of corrective actions provided in their August 18,1997, letter to the NRC.
The inspection bases were:
- Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants.'
10 CFR Part 21, " Reporting of Defects and Noncompliance."
+
ASME Code, Section lil, Subarticle NCA 3800.
During this inspection, one nonconformance was identified and is discussed in Sections 3.3.b and 3.3.c of this report.
2 STATUS OF PREVIOUS INSPECTION FINDINGS The last NRC inspection of Nova was performed on July 14 through 18,1986. The status of findings from that inspection 'Nas not reviewed.
2
___._.__..-.m 1
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Descriotion of Facilities and Activities l
Nova has been accredited by the ASME as a Material Organization (MO),
manufacturing threaded fasteners and supplying ferrous and nonferrous bars, threaded j
fasteners, fittings, flanges, and other products.. According to Nova management, approximately 60-70% of their sales volume consists of safety related material supplied to nuclear utilities. Of that amount, approximately 50% is sold under their ASME Quality Systems Certificate (QSC) as complying with the requirements of Subarticle NCA 3800 l
of Section ill of the ASME Code Manufacturing capability includes cold and hot forming and mschining. Inspection and test capability includes complete dimensional and visual inse don, tensile and hardness testing, and chemical analysis capability for carbon and loiv alloy steels Nondestructive examination and heat treating are subcontracted to approved suppliers.
i 3.2 Review of Nova's 10 CFR Part 21 Proaram and its lmolementation The inspectors reviewed 10 CFR Part 21 postings at the manufacturing facility and
]
found them to be consistent with the current requirements. The practices for identifying and reporting defects and the time frames for evaluation and notification are described In Nova procedure NP-09, Revision 7, dated March 3,1997. This procedure assigns the responsibility for documenting the results of deviation evaluations to the QA manager and identifies the nonconformance report (NCR) as the record of evaluation.
i The inspectors examined Nova's nonconformance file and did not identify any instances
.where an evaluation resulted in a reportable condition. The inspectors noted that the j
evaluation typically consisted of a check mark in the appropriate Yes/No box on the NCR form with no documented bases for this determination. While it can be argued that, in most cases, the bases are readily apparent from other information cor.iained in the NCR, this did not always appear to be the case. For example, NCR 97 29 l
documented a situation where parts had been rejected and returned to Nova by a L
customer because Nova had failed to perform liquid penetrant testing as required by the customer's purchase order (PO). The NCR identified the cause of the deviation as f#ure to include the liquid penetrant testing requirement in Nova's PO to their vendor.
The *Not Reportable
- Part 21 box was checked with a notation *LP passed" indicating successful testing of the retumed parts. The inspectors noted that the evaluation did not fully address the root cause of this devin, tion to determine why the test requirement was not included in the PO nor the generic potential of such omissions which could require a
l the notification of other customers.
Evaluation of deviations was identified as a weakness in Nova's 10 CFR Part 21 program implementation. Nova management indicated that this issue would be addressed by appropriate changes to the implementing procedure, i
3 h
4
3.3 Review of Recent Sales of ASME Code Material to NRC Licensees a.
Stage The NRC inspectors reviewed the procedural controls and document files for a sample of recently filled orders for safety related matrial supplied under Nova's ASME QSC to evaluate the implementation of the applicable QA program and ASME Code requirements.
b.
Observations and Findinas Nova operates under one quality program. The Quality Assurance Manual (Edition 4 Revision 1) has been written to comply with the applicable requirements of the ASME Code, ISO-9001; Canadian Standard Z-299; 10 CFR Part 50, Appendix B; ANSI N45.2 and other related quality standards. The QA manualis supported by a controlled set of implementing procedures.
Approximately 10 document files supporting the certification of recent material sales under Nova's ASME QSC were reviewed to evaluate the effectiveness of their QA program implementation. Significant observations and findings are summarized below:
Centerior Energy (Davis Besse) Purchase Order (PO) S 078945 D97, dated March 13,1997, for eight 3/4 inch threaded studs,13 inches long, in accordance with ASME SA 193, Grade 87.
This material was certified as supplied under Nova s QSC in full compliance with the applicable PO requirements. Attached to the Nova certification was a material test report issued by Chaparral Steel Co. or MGSn, TX, for 20 tons of hot rolled AISI type 4140 steel bars sold to Cardinal Fasteners, Dedford Heights, OH. Also attached was a heat treatment certification issued to Nova by J.W. Rex Co., Lansdale, PA, (a qualified vendor). The Chaparral certification referenced their QSC number, indicating that, although the material was not provided as ASME Code material, it was processed in accordance with Chaparral's QSC.
The inspector noted that: 1) there was no indication of Cardinal Fasteners role in supplying this material or how Nova acquired the material,2) the heat treatment certification was incomplete, since only a tempering cycle was described, and 3) there was no indication that the materialinternal quality had been verified by I
macroetch as required by ASME SA-193 During the course of the inspection, Nova staff produced records indicating that, although the material had been purcSased by Cardinal, it had been drop-shipped to Nova, and that three samples of the material had been macroetched by an approved laboratory and found acceptable. The records also indicated that the heat treatment certification that was provided to the customer represented only a 4
re temper cycle and did not describe the complete heat treatment that was performed on this material. The re-temper cycle was necessary because the initial temper was conducted at a temperature below the specified minimum.
Records for the original quench and temper cycle were available.
The inspector noted that, since Nova certified this material as provided in accordance with their QSC, supplie :retifications for the complete heat treatment and macroetch examination should have been provided to their customer as required by ASME Code, Section IC, NCA-3862.1 Northeast Utilities (Millstone) PO 00951296, Release 023, dated March 11,1997 for 40 feet of 3/4 inch threaded rod in 10 foot lengths, in accordance with ASME SA-193, Grade B6X This material was certified as supplied under Nova's QSC in full compliance with the applicable PO and specification requirements. Attached to the Nova certification war a Carpenter Technology (qualified vendor) certificate of test for Type 410 annealed ground bar. The certificate also referenced ASTM A 182, Grade F6A (chemist y only). The certificate did not reference or indicate compliance with requirements of SA 193, but contained a statement " Discs Macroetched and Approved." There was no refereve to the macroetch test method or acceptance criteria. Also attached was a certificate from J.W. Rex Co. (approved vendor) stating that 18 pieces and one test piece of this material were hardened and tempered per SA 193, Grade B6X to a hardness level of 24-26 Rc. Also attached was a certificate from Herron Test Lab. (approved vendor) with tensile and hardness test results for this material.
The inspector noted that the mill certification did not contain complete information to verify compliance with the internal quality requirements specified in SA 193 which requires the macroetch to be penbrmed in accordance with ASTM E-381 (or other suitable method as agreed us.on between purchaser and producer), and provides specific acceptance criteria, e accordance with the material certification requirements of NCA-3862, the basis for compliance with SA-193 requirements should nave been included in the original mill certification or supplementary test report and provided to the purchaser of this material.
PECO Energy Co. (Limerick Generating Station) PO LS 608180, line item 03, for Twelve 5/8 inch by 3 % inches long threaded rods, in accordance with ASME SA-193, Grade 87 This material was certified as supplied under Nova's QSC in full compliance with the applicable PO and material specification requirements. The certificate also noted that the material was upgraded per NCA-3855.5. Attached to the Nova certification was a Certification of Inspection and Test, issued to Nova by Walker Wire and Steel Corp. (not on Nova's approved vendor list) which identified the material as.640 inch diameter 4140 cold heading quality hot rolled rod. Also attached to the Nova certification were heat treatment certification from J W.
5
,i I
Rex Company (approved vendor) indicating that the material had been processed per customer requirements to surface hardness of 27 28 Rc, a test report from Herron Testing Laboratories, Inc. (approved vendor) containing the sasults of product analyses performed on four samples of this material, and a test report from Herron Testing Laboratories containing the results of tensile tests on this material.
The inspector noted that the information provided with Nova's certification did not provide an adequate basis for establishing compliance to the requirements of SA-193, since there was no record of verification of the material internal quality (macroetch). Furthermore,it could not be determined if the upgrading was properly performed, since the supplier's certification did not indicate how many separate pieces of material were provided to fill this order.
During the course of the inspection, Nova starf produced additional records indicating that Walker Wire purchased the starting material in four coils from Charter Steel. The mill certification was on file at Nova. Nova also oroduced a test report from Herron Laboratories (approved vendor) indicating tnat one sample of this heat of material had been macroetched in accordance with ASTM E-381 and met the acceptance criteria specified in SA 193.
The inspector noted that, in accordance with NCA 3862.1, the macrootch test results should have been provided to the purchaser of the material. The inspector also noted that, since the coils were purchased as unqualified source material and upgraded in accordance with NCA 3855.5, each of the four coils should have been tested.
c.
Conclusions The document review described above identified three examples where certification packages provided to Nova customers were incomplete in that they did not include the results or certifications of all tests and heat treatments performed by approved suppliers as required by paragraph NCA-3862.1 of Section 111 of the ASME Code. Failure to include certifications from approved suppliers as identified attachments to Nova certifications to Centerior Energy (PO S 078945 D97), Northeast Utilities (PO 00951296), and PECO Energy iPO LS 608180) were identified as Nonconformance 99901052/97-01-01.
During the October 1997 inspection, the inspectors reviewed Nova's corrective actions.
The inspectors determined that Nova provided the applicable certifWions to PECO Energy by letter of August 13,1997, and to Centerior Energy and Northeast Utilitiet by letters of August 28,1997, in addition, Nova revised the certification checklist referenced in Procedure NP-15 to address the Code requirement to provide all heat treat data required by the material specification and to include the macroetch test results, as applicable. The inspectors confirmed the revision had been made and reviewed records indicating that appropriate personnel had been trained in the revision.
6
_.m m
I
.i 3.4 Quahfication of MaterialVendors The Nova QA manual has established several categcries of ve'idors, based on their scope of s'Jpply. Tlie methods of vendor qua'ification or approval for each of these categories are described in Section 6.5 of the manual. Non accredited vendors of ASME Code material are qualified by Nova survey, performed on a three year frequency. The surveys are performed by auditors qualified in accordance with Section 14 of the Nova QA manual and procedure NP-10, " Qualification of Auditors,"
using approved checklists.
A sampling review of audils of ASME Code material suppliers qualified by Nova, including the completed checklists, showed 1 hat these documents did not address the upgrading of unqua5fied source material. According to the Nova QA manager, material purchased from Nova-qualified manufacturers is generally limited to basic mill products which are not upgraded, therefore, upgrading of stock material was not included in the scope of the aud't checklists. The inspectors, however, identified at least one vendor (Flowline Division of Markowitz Enterprises) that Nova had qualified to a scope thet included the upgrading of unqualified source material. A review of the vendor's QA manual, on file at Nova, indicated that the manual contained adequate controls for the r
upgrading process.
Failure to include controls for upgrading of unqualified source materialin the vendor survey checklist was identified as a program weakness.
3.5 Commercial Grade item Dedication Proaram a.
Insoection Scoce The NRC inspectors reviewed the program and procedura! controls applicable.to Nova's Commercial grade item (CGl) dedication activities including the selection of critical characteristics and the use of sampling plans for nondestructive and destructive examinations and tests. The inspectors also observed in-process dedication activities and reviewed document files for a sample of recently filled and in-process materiai orders to evaluato the implementation of the applicable QA program requirements.
b Observations and findinas b.1 Dedication Methodology The requirements, for CGI dedication are prescribed in Nova's Quality Assurance Manual, Edition 4, Revision 1, and Procedure No. NP-12. " Supplemental Testing,"
Rovision 10, dated March 3.1997.
Nova selects critical characteristics for CGls based on the requirements specified by the applicable American Society for Testing and Materials (ASTM) material soecification and any other testing requirements specified by the customer. Critical characteristics, in 7
~
~
.i general, include dimensions, workmanship, and when required by the applicable material specification, chemistr' and mechanical properties and other specified tests.
/
Subsequent to the June 1937 inspection, Nova provided clarification to the NRC on the verification of mNhank'al and chemical critical characteristics as provided for in Section 1 of Attachments 5.0(A),5.0(B), and 5.0(C) of Procedure NP-12. Section 1 states, *Mechanica! and chtmical attributes: V/ hen size allows (as defined by the cpplbabie material specification), test and correlate to the requirements dictated by the customar or applicable mabrial specificatbn.' Nova informed the inspectors that Saction i cf ?N attachments to Procedure NP 12 requires that all testing and analyses requirea by the appNb!w ASTM material specification be performed, however, Nova further explained tnt some material specifications exempt pMucts from certain tests and analyses based on the pr%+>ct size.
b.2 Uste of Sampling Plans for CGI Dedication Sampling plans used by Nova for CGI dedication are based, in part, on a combi,,ation of the & revisions contained in: a) EPRI NP 7218, " Guideline for the Utilbation of Sampling Plans for Commercial-Grade item Acceptance (N.G.-19)," dated June 1992, b) ASTM A-490 (93), " Standard :Decification for Heat Treated Steel Structural Bolts,150 ksi Minimum Tensile Strength;" and c) the degree of confidence that the lot being sampled contains homogenous items.
Nova informed the NRC inspectors that they chose the sample size provided in the shipcing lot sampling plan contained in ASTM A-490 (93) because this plan is one of the most severe sampling plans for chemical and mechanical destructive testing in the ASTM standards. The NRC inspectors exoressed a concern to Nova that the Shipping Lot Size sa 1ple plan provided in ASTM A-490 is for the manufacturer of a commercial product line. Also, the NRC inspectors identified that the quality controls identified in ASTM A-490 do not appear to meet the quality requirements of Appendix B to 10 CFR Part 50 and that such a sample size could include more than one production lot or heat of material. Nova informed the inspectors that it selected this sampling plan based upon the quantity provided in the " Shipping Lot Size" table.
The NRC inspectors also expressed a concern that the qualitative methods for determining whe,1 a homogenous lot exists and that the basis for such decisions, as described in Procedure No. NP 12, were not very prescriptive and placed much reliance on judgement. Other concems expressed by the NRC inspectors are addressed in Section 3.5.c.1 of this inspection report.
Procedure NP 12 lists a number of sampling plan selectien factors and provides four catcgories of san.pling plans based on those factors. Thcse sampling categories are included in Attachment 5.0 to Procedure NP-12 and are discussed below:
I l
8
s.
.c b.2(a) Samolina Plarb ttachment 5.0 (A) of Procedure No. NP-12 A
Nova uses the following sampling plan for its dedication of CGis purchased from a commercial vendor where there is no vendor history, the vendor has not been audited, the part is not identified to a standard, and the part is not marked (Reference is made to the applicable sampling tables contained in EPRI NP 7218.
Inspection Characteristic Sampling Plan 1
Configuration Tightened (Table 2-1 of EPRI NP 7218)
Dimsasions Tightened (Table 2-1 of EPRI NP 7218)
Appearance Tightened (Table 2-1 of EPRI NP-7218)
Hardness
- Tightened (Table 21 of EPRI NP-7218)
Chemical
- See Section 3.5.b.2(e) of this Report Mechanical
- See Section 3.5.b.2(e) of this Report
- When required oy th3 material specification b.2(b) Samolina Plan Attachment 5.0 (B) of Procedure No. NP 12 l
Nova uses the following sampling plan for its dedication of CGls purchased from a j
commercial grade vendor that has been used before (but not qualified) and where parts j
are identified and can be traced to a known standard based on the vendor's nonqualified i
and unaudited markings and documentation:
I Inspection Characteristic Sampling Plan Configuration Normal (Table 21 of EPRI NP-7218)
Dimensions Normal (Table 2-1 of EPRI NP 7218)
Appearance Normal (Table 2-1 of EPRI NP-7218)
Hardness
- Normal (Table 21 of EPRI NP-7218)
Chemical
- See Section 3.5.b.2(e) of this Report j
Mechanicai*
See Section 3.5.b.2(e) of this Report
- When required by the material specificat!on b.2(c) Samolina Plan Attachment 5.0 (C) Procedure No. NP-12 (Condition A)
Nova uses the following sampling plan for its dedication of CGis purchased i
j from a commercial grade vandor when a commercial audit of the vendor has been conducted to verify the items are produced from a single heat / lot of material and 2
traceability has been maintained throughout all processing.
!j 9
J
,a,,,
,-,n--
n,,
w
,,-----,g---
-, ~ - -
,~v-,,,,e m-
-w
l,.
Irispection Characteristic Sampling Plan Configuration Normal (Tabis 21 of EPRI NP 7218)
Dimensions Normal (Table 21 of EPRI NP 7218)
Appearance Normal (Table 2-1 of EPRI NP 7218)
Hardness
- Normal (Table 21 of EPRI NP 7218)
Chemical
- See Section 3.5.b.2(e) of this Report Mechanical
- See Section 3.5.b.2(e) of this Report
- When required by the material specification b.2(d) Samoling Plan Attachment 5.0 (C) Procedure No. NP-12 (Condition B)
Nova uses the following sampling plan for purchases from a commercial grade vendor when a commercial audit of the vendor has been conducted to verify that the items are produced from a single heat / lot of material, traceability has been maintained throughout all processing and other attributes have been reviewed during the audit and found acceptable.
Inspection Characteristic Sampling Plan Configuration Reduced (Table 2-1 of EPRI NP 7218)
Dimensions Reduced (Table 2-1 of EPRI NP-7218)
Appearance Reduced (Table 2-1 of EPRI NP 7218)
Hardness
- Reduced (Table 2-1 of EPRI NP 7218)
Chemical
- See Section 3.5.b.2(e) of this Report Mechanical
- See Section 3.5.b.2(e) of this Report
- When required by the material specification b.2(e) Mechanical and Chemical Testing Per Procedure No. NP-12 Regardless of Nova's activities associated with the vendor supplying CGis or the review of the vendor's quality program implementation, Nova uses the same sampling plan for verifying the CGl's chemical and mechanical properties. This amount of testing is based on the Shipping Lot Size samp..1g plan provided in ASTM A-490 (93).
10
l.
Lot Size Sample Size Acceptance 1 (1 to sell,1 to test) 1 (Note 1) 2 to 25 2
(Note 2) 26 to 50 3
(Note 2) 51 to 150 5
(Note 2) 151 to 1200 E
(Note 2) 1201 to 10,000 13 (Note 2) 10,000 to 35,000 20 (Note 2) 35,001 to 150,000 32 (Note 2)
Note 1: Acceptance Number is 0, test piece must pass.
Note 2: Acceptance Number is 0, all test pieces must pass.
The NRC inspectors and Nova discussed the need for a 95 percent confidence that there are only 5 percent defective items in a lot (95/5) as discussed in documents referenced in NRC Information Notice No. 96-40, " Deficiencies in Mater.at C,edication and Procurement Practices and in Audits of Vendors," dated July 25,1996. The NRC inspectors informed Nova that the 95/5 confidence level is applicable for dedicated CGls used in high-safety-significant applications, as determined by NRC licensees using probabilistic and deterministic methods similar to those described in NRC draft regulatory guides DG-1061, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Current Licensing Basis,"
dated March 28,1997, and DG-1064, "An Approach for Plant-Specific, Risk-Informed Decision Making: Graded Quality Assurance," dated March 24,1997. A 95/5 confidence level may also apply to items dedicated for unrestricted safety application (end use or safety classification not identified) and certified as complying with all requirements of 10 CFR Part 50, Appendix B.
Nova questioned the NRC regarding the need for all customer orders to be processed in accordance with a sampling plan that provides a 95/5 confidence level. The inspectors informed Nova that, as discussed above, the 95/5 confidence level was applicable only to those items dedicated to high safety significant (or unrestricted) applications and that it was the NRC licensee's responsibility to evaluate and accept the sampling plans used 5v Nova for dedicating CGls supplied for its purchase orders, it was further discussed that for dedicated CGls used in low-safety-significant applications, the use of a graded approach, as discussed in DG-1064, to determine the necessary sampling plan would be acceptable. For such applications, a sampling plan having a confidence levelless s
than 95/5 confidence level may be appropriate.
The NRC inspectors informed Nova that, in general, a reduced (less than 95/5) sampling plan should be used only after review and acceptance by the customer. Nova informed the NRC inspectors that its sampling plans were generally reviewed by its customers during their audits and that recently a joint utility audit had been performed and there were no findings in the area of Nova sampling plans used for dedicating CGis.
The NRC inspectors' review of the joint utility audit is summarized in Section 3.6 of this report.
I1
,e b.3 Review of Recent Sales of Dedicated Material l
The NRC inspectors reviewed severalin process and completed CGI dedication packages to determine if the critical characteristics for the iterns had been properly identifiad and veiified. In addition to the identified quality and technical regt'irements, the following POs invoked the requirements of Appendix B to 10 CFR Part 50 and 10 CFR Part 21.
Item 2 on Nova Sales Order No. 49P18, was for the supply of 100, ASTM B 21, Alloy 464, slotted flathead No. 4-40 x 3/16 brass screws, in accordance with Nebraska Pubu Power District PO 97-1872, dated May 6,1997. This sales c
order was being processed during the NRC inspection.
To supply these fasteners, Nova issued a commercial PO and purchased a 1/4 inch diameter X 6 feet long, ASTM B21, Ailoy 464, round bar, from Copper &
Brass Sales in Cleveland, Ohio. Prior to machining each machine screw, Nova cut two pieces of material from the round bar and sent them to an appro,ed supplier, Herron Testing Laboratories, Inc., located in Cleveland, Ohio, for chemical and mechanical testing (tensile, yield, elongation, and reduction in area). Because this sales order was being processed during the inspection, the results of the configuration, dimensional, and final visual inspection were not available to the NRC inspectors.
The NRC inspectors found the dedication activities specified by Nova for dedicating the screws on Sales Order No. 40818 to be satisfactory.
Item 2 on Nova Sales Order No. 39933 was for the supply of 10, SAE J429 (August 83), Grade 8, 5/16-18 UNC 2A X 4 inch, fully threaded hex head capscrews, in accordance with PECO Energy Company Blanket PO No. 159325, Release No. 348002, dated May 13,1997.
This sales order was being processed during the NRC inspection. To supply these fasteners, Nova obtained 5/16-18 UNC 2A X 6 inch fasteners of the specified material from its stock. These stock fasteners were purchased safety-related in 1988, with Appendix B to 10 CFR Part 50 and 10 CFR Part 21 invoked, from Lake Erie Screw Corporation in Cleveland, Ohio. Although Lake Erie Sciew Corporation is not presently an approved supplier, at the time of purchase in 1988, it was on Nova's approved suppher list. Nova informed the NRC that, evt 1 though the fasteners purchased from Lake Erie Screw Corporation were considered basic components, they were being processed as though they were not fully dedicated CGis because they were being altered from 6 inches to 4 inches long and the PECO Energy Company's PO had invoked several unique certification requirements.
To process the sales order, Nova performed hardness testing on two capscrews, proof load tests on two capscrews, and wedge tests on two capsc.rews. Nova did not perform chemical overchecks on the capscrews and jt:stified not 12 J
l e.
l performing chemical tests on the capscrews because they were purchased from an approved supplier. Eight capscrews were subjected to full dimensional and configuration checks and all completed capscrews were suojected to a final visual examination.
The NRC inspectors found the procurement, inspection, and testing activities performed by Nova for the supply c1 toe capscrews on Sales Order No. 39533 to be satisfactory.
Item 5 on Sales Order No. 39089 was for the supply of 100 ANSI B18.21.1(83),
1/4 inch, type 304 stainless steellock washers (regular helical) in accordance with PECO Energy PO 159622, dated April 7,1997.
To supply these fasteners, Nova abtained the lockwashers of the specified size and material from its stock. The lockwashers in stock were purchased commercially from the Mellowes Company in Milwaukee, Wis:onrin in 1987.
Nova dedicated the lockwashers by obtaining 172 from stock and performing the following dedication activities:
A total of 72 pieces were sent to an approved supplier, Herron Testing a.
Laboratories, Inc., located in Cleveland, Ohio. A total of 18 chemical tests and 18 hardness tests were performed.
b.
The remaining 100 lockwashers were subjected to dimensional and visual inspections (inside and outside diameters and thickness).
The NRC inspectors found the procurement, inspection, and testing activities performed by Nova for the supply of the lockwashers on Sales Order No. 39089 to be satisfactory.
On October 2,1997, the NRC inspectors reviewed an additional nine Nova CGI dedication packages, processed during the period March through September 1997, for the supply of small metallic products selected from the following POs:
five for Northeast Utilities (POs 02015122,02018733,02039272, and 00951296, Blanket Releases Nc. 34, and No. 58), two for the Tennessee Valley Authority (PO P95N2G108252, Releases 1183024 and 1184926), one for Nebraska Pu > tic Power District (PO 971872), and one for PECO Energy (PO159325, Release No. 387049).
The inspectors fou..J that all required chemistry analysis, mechanical testing, and hardness testing was performed when required by the applicable PO and material specification.
The inspectors noted that several of the above orders specified ANSI B18.6.3," Slotted and Recessed Head Machine Screws". The inspectors and Nova discussed the intent of the provisions contained in Section 2.7, ' Material," of ANSI B18.6.3 which states the following:
13
e.
2.7.1 MaterialScrews. Unless otherwise specified, machine screws shall be fabricated from carbon steel and shall have a minimum tensile strength of 60,000 psi.
l 2.7.2 Machine Screws, where so specified, may also be made from l
higher strength steels, corrosion resistant steel, brass, monel, aluminum alloys or other materials, as agre d upon between the manufacturer ar;d the purchaser.
Nova informed the inspectors that it interprets the above as meaning that only carbon steel must satisfy the 60,000 psi minimum tensile strength. Nova considers that screws manufactured from stainless steel, brass, or other non-carbon materials are exempt from tensile requirements unless specified by the purchaser. Nova considers stainless steel as corrosion resistant si 'el. Nova further informed the inspectors that customer POs for non-carbon screws manufsetured in accordance with ANSI B18.6.3 typically do not specify tensile test requirements.
c.
Conclusior,a The NRC inspectors concluded that Nova's CGI dedication program addresses the essential elements of the dedication process and that sufficient guidance for performing activities such as inspection and testir.g are provided. However, the NRC inspectors identified the following weaknesses:
The NRC inspectors questioned Nova's practice of stamping all CMTRs, including those received from unqualified suppliers, with a
- Nova OA Approved' stamp, based only on a comparison of the CMTR reported values to the material specificatbn requirements. The inspectors noted that this practice may be misleading to the recipients of such CMTRs, since Nova QA approval may be reasonably assumed to also include other supplier activities such as material melting practice or heat treatment, which may not be directly verifiable for unqualified source material. This practice is of concern also because Nova's CGI dedication program does not specifically require that the results of their verification tests (chemical and mechanical) be compared to the values reported on the original CMTR to validate this document. Significant composition variances, even within une product analysis tolerance limits, between the Nova overchecks and CMTR reported values could be an indicator of mixed heats of material or incorrect CMTR fcr the material being dedicated. Significantly diverging results could, therefore, also invalidate other pertinent information included on the suspect CMTR.
All of the nondestructive sampling plans in the attachments tc NP-12 use the same destructive testing sampling plan regardless of whether the products were purchased as a lot from multiple or a single nonqualified supplier or whether the products were obtained from a supplier that had or had not been subjected to a commercial grade survey.
14
s.
.r The NRC inspectors observed Nova personnel performing the following act;vities: a) receiving, in process, and final inspections, b) chemical analyses, c) tensile, proof load, wedge, and hardness testing, d) order review and e) shop order processing. Personnel performing these activities demonstrated pro'iciency in accomplishing these activities and a high degree of knowledge of the applicable industry codes and standards and Nova procedure requirements for the activities. There were no abnormalities observed.
As a result of their observations and reviews, the NRC inspectors considered the following to be additional strengths:
Nova's attitude to perform work activities correctly was evident during the conduct of the inspection, especially during the discussions between the NRC and Nova concerning the selection of appropriate sampling plans used for CGI dedication.
The intenaces and communications between the various organizations processing a customer's PO were observed to be well established and implemented in a professional manner.
3.6 Nuclear Procurement issues Committee (NUPlC) Joint Licensee Audit The inspectors reviewed information related to a NUPIC joint utility audit, performed at Nova on April 23 through 25,1997. The lead utility '~ this audit was Nebraska Public Power District, Cooper Nuclear Station, with an audit team consisting of a lead auditor and a technical specialist from the Nebraska Public Power District and a quality assurance engir.eer from Niagara Mohawk Power. The technical specialist evaluated Nova's sampling plan including its use in CGI dedication activities. The technical specialist identified no findings, adverse observations, or weaknesses in these areas.
Nebraska Public Power issued the Nova audit report, including completed audit checklists, tc the NUPIC members by cover letter dated May 23,1997. During the NRC inspection, the inspectors identified concerns to Nova regarding the NUPlC audit report and checklists. As a result of a subsequent internal critique and various comments from NUPIC members and by Nova, Nebraska Public Power District issued a revision to
'he audit report, dated June 18,1997, and distributed this revision to the NUPlC members.
Based on a review of the Nova CGI dedication program and its associated sampling plans, the NRC inspectors determined that the origin 01 and revised NUPIC joint utility audit reports and checklists, dated May 23,1997, and June 18,1997, respectively, lacked sufficient detail for a NUPlC member to adeo ately assess NOVA's CGI dedication process.
Specifically, the NUPIC audit report notes that the sampling plans, used in Nova's CGI dedication program, meet the provisions of MIL-STD 105E and/or EPRI NP-7218.
However, the audit reports failed to identify that NOVA's Procedure No. NP-12, Revision 10, permitted the use of sampling plans for CGI dedication that did not meet 15
o
.c either NP-7218 or MIL-STD-10SE. As discussed in Sections 3.5.b.2 and 3.5.b.2(e) of this inspection report, Nova's destructive testing sampling plan used for CGI dedication is based on ASTM A-490(93).
Additionally, although Nova's practice of stamping CMTRs from approved suppliers is addressed, the completed NUPIC audit checklists and audit reports do not address how Nova processes CMTRs received from nongealified suppliers when performing CGI dedication. According to Nova's procedure.s/ practice, CMTRs received from i
nonqualified sources are stamped " Nova QA Approved," meaning that Nova has reviewed and compared the information on the nonqualified CMTR against applicable specification requirements and found it acceptable. Nova then processes the associated CGI through its dedication program, defined in Procedure NP-12. Further, should the chemical and mechanical overchecks performed by Nova vary from the values listed on the CMTR, Procedure NP-12 does not require Nova to take any action l
to investigate such variances. Thus, a customer receiving a CMTR stamped " Nova QA Approved
- would not be aware that; a) the item associated with the CMTR may have been dedicated and the CMTR is from a nonqualified source, b) there is a pos:.:bilRy that test resuits and other the information listed on the CMTR may vary from the actual Nova overcheck test results (Nova does not provide its CGI dedication chemical and mechanical test results unless specifically requested by its customers), and, c) there is a possibility that information listed on the nonqualified suppliers CMTR has not been independently overchecked by Nova.
f in summary, the NRC inspectors noted that the effectiveness and usefulness of joint audits are dependent on capturing sufficient information to enable the potential users of the audit information to reasonably assess the vendor's program controls and implementation for specific activities that are important to the user's procurement needs.
The Nebraska Public Power District joint audit reports and checklists do not appear to contain sufficient detail for NUPlc members to assess Nova's controls for its commercial grade item dedication process. For e.xample, as discussed above, the provisions of Procedure No. NP-12 are not adequately described. The " Test and Inspectivn" section 4
contained in the audit report, could :>e assumed to mean that Nova's destructive testing sampling program meets MIL-STD-105E (neither the audit report nor checklist address 4
the fact that Nova's destructive testing sampling plan is based en ASTM A-490 nor do they address the acceptability of the sample size provided in ASTM A-490 shipping lot sampling plan),
Based on the above discussion, Nova's sampling plan, as described in NP-12 and as evaluated by NUPIC members, may not be adequate for items that are intended to be used in high-safety-significant applications or applications where environmental conditions are of particular concern. For such applications, the licensee may desire the use of a sampling plan that will provide a higher confidence level than that provided by a
. destructive sampling plan based on ASTM A-490. Also, for these applications, the licensee may have the need for a CMTR produced under an Appendix B QA Program or the actual Nova test results rather than using information delineated on a CMTR frorn a nonqualified source.
{
16 4
3.7 Entrance and Exit Meetings I
In the Entrance meetings, on June 2,1997, and October 2,1997, the NRC inspectors
[
discussed the scope of the inspection, outlined the areas to be inspected, and established interfaces with Nova menagement. In the exit meetings, on June 5,1997, and October 2,1997, the inspectors discussed their findings and concerns.
PARTIAL LIST OF PERSONS CONTACTED James A. Mraz, President and CEO Tim Walker, Executive Vice President John Burk, General Manager Jim Fitzwilliam, Quality Assurance Manager Dave Nenstiel, Laboratory Manager
. Sandy Jacoby, QC Documentation Kurt Bartley, i.evel ll Inspector Tony Holloway, Level ill Inspector Tim Corrigan, Materials Director ITEMS OPENED, CLOSED, AND DISCUSSED
' Ooen/ Closed 99901052/97-01-01 NON Incomplete documentation to support CMTR 17
- - _ _ _