ML20205A599
| ML20205A599 | |
| Person / Time | |
|---|---|
| Issue date: | 03/31/1999 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| NUREG-0040, NUREG-0040-V22-N03, NUREG-40, NUREG-40-V22-N3, NUDOCS 9903310025 | |
| Download: ML20205A599 (103) | |
Text
- . _ . - _ _ . . . .
7 I NUREG-0040
} Vol. 22, No. 3 l ) Licensee Contractor and Vendor Inspection Status Report 1
Quartedy Report l July - September 1998 I
9
' $ 'O
,3l
! E l
S 2 U.S. Nuclear Regulatory Commission ,.*'""
Oflice Nuclear Reactor Regulation ; )
- Washington, DC 20555-0001 ' ,,,,, /
h 4
33 990331 0040 R PDR
i
) Vol. 22, No. 3 l
j Licensee Contractor and l Vendor Inspection Status Report l
Quarterly Report July - September 1998 0 '
/
} fl/
G pO
<: T J
U.S. Nuclear Regulatory Commission ,/ ~~*g Office Nuclear Reactor Regulation i,,
Washington, DC 20555-0001 %, ,
I 0: .
[
'JE8990331 t '
AVAILABILITY NOTICE -
Availability of Reference Materials Cited in NRC Publications NRC publications in the NUREG series, NRC regu- NRC Public Document Room lations. PM Tnia 10, Energy, of the Code o/ Federal 2120 L Street, N.W., Lower Level Regulations, may be purchased from one of the fol- Washington, DC 20555-0001 lowing sources: < http://www. nrc. gov /N RC/PD R/pd r1. htm >
1 -800-397-4209 or locally 202-634-3273
- 1. The Su; + itendent of Documents U.S. Govemment Printing Office Microfiche of most NRC documents made publicly RO. Box 37082 available since January 1981 may be found in the Washington, DC 20402-9328 Local Public Document Rooms (LPDRs) located in
< http://www. access.g po. gov /su_ docs > the vicinity of nuclear power plants. The locations 202- 512 -1800 of the LPDRs may be obtained from the PDR (see
- 2. The National Technical Information Service Springfield, VA 22161 -0002 <http://www.nrc. gov /NRC/NUREGS/
<http://www.ntis. gov /ordernow> SR1350/V9/lpdr/html>
703-487-4650 ;
Publicly released documents include, to name a l The NUREG series comprises (1) brochures ,
few, NUREG-series reports; Federal Register no-(NUREG/BR4000(), (2) proceedings of confer- l' tices; applicant, licensee, and vendor documents ences (NUREG/CP-XXXX), (3) reports resulting from international agreements (NUREG/lA XXXX) and correspondence; NRC correspondence and )
internal memoranda; bulletins and information no-(4) technical and administrative reports and books tices; inspection and investigation reports; licens.
[(NUREG-XXXX) or (NUREG/CR-XXXX)], and (5)
- event reports; and Commission papers and
)
compilations of legal decisions and orders of the the.ir attachments.
Commission and Atomic and Safety Licensing Boards and of Office Directors' decisions under Section 2.206 of NRC's regulations (NUREG- Documents available from publ.ic and specialtech .
nicallibran,es include all open literature items, such XXXX).
as books, journal articles, and transactions, Feder-A single copy of each NRC draft report is available al Register notices, Federal and State legislation, free, to the extent of supply, upon written request and congressional reports. Such documents as as follows: theses, dissertations, foreign reports and transla-tions, and non-NRC conference proceedings may Address: Office of the Chief Information Officer be purchased from their sponsoring organization.
Reproduction and Distribution Services Section Copies of industry codes and standards used in a U.S. Nuclear Regulatory Commission substantive manner in the NRC regulator, process Washington, DC 20555-0001 are maintained at the NRC Liti,2ry, Two White Flint E-mail: < DISTRIBUTION @nrc. gov > North, 11545 Rockville Pike, Rockville, MD Facsimile: 301 - 415 -2289 20852-2738. These standards are available in the library for reference use by the public. Codes and A portion of NRC regulatory and technicalinfo'rma. standards are usually copyrighted and may be tion is availab;e at NRC's World Wide Web site: purchased from the originating organization or, if they are American National Standards, from-
<http://www.nrc. gov >
American National Standards institute All NRC documents released to the public are avail- 11 West 42nd Street able for inspection or copying for a fee, in paper, New York, NY 10036-8002 microfiche, or, in some cases, diskette, from the <http://www. ansi.org>
Public Document Room (PDR): 212 -642-4900 A year's subscription of this report consists of four quarterly issues.
NUREG-0040 Vol. 22, No. 3 l Licensee Contractor and Vendor Inspection Status Report Quarterly Report July- September 1998 Manuscript Completed: March 1999 Date Published: March 1999 Divisi:n ofInspection Program Management Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 gf "*%
4%,.....,/
NUREG-0040, Vol. 22, No. 3 has been reproduced from the best available copy.
1 i
v i
l ABSTRACT l
l l This periodical covers the results of inspections performed between July 1998 and
{ Septernber 1998 by the NRC's Quality Assurance, Vendor Inspection and Maintenance Branch
- that have been distributed to the inspected organizations.
l i
I 1
I l
r l
l 1' ...
Ill i.
4
i l
l l
CONTENTS
)
l l PAGE l 1
I Abstract.................................................................................................................... iii ;
l \
I n trod uction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . !.
i n spe ction R e po rts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 t
, Diablo Canyon Power Plant (50-275/98201 and 50-323/98201)........... 2 l San Francisco, CA Limitorque Corporation- (99900100/98-01 ) .................................. 29 Lynchburg, VA l
)
i Morrison Knudsen Corporation (99901329/98-01 and 99901334/98-01). 43 and The Steam Generating Team Cleveland, OH 1
Siemens Power Corporation (99900081/98-01 ) ............ ..................... 63 Richland, WA Westinghouse Electric Company (99900404/98-02) ................ ................. 78 Pittsburgh, PA l
I t
V
INTRODUCTION A fundamental premise of the U. S. Nuclear Regulatory Commission (NRC) licensing and inspection program is that licensees are responsible for the proper construction and safe and efficient operation of their nuclear power plants. The Federal government and nuclear industry have established a system for the inspection of commercial nuclear facilities to provide for multiple levels of inspection and verification. Each licensee, contractor, and vendor participates in a quality verification process in compliance with i requirements prescribed by the NRC's rules and regulations (Title 10 of the Code of Federal Regulations). The NRC does inspections to oversee the commercial nuclear industry 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 quality verification programs. l The licensee is responsible for developing and maintaining a detailed quality aesurance !
(QA) plan with implementing procedures pursuant to 10 CFR Part 50. Through a !
system of planned and periodic audits and inspections, the licensee is responsible for ensuring that suppliers, contractors and vendors also have suitab!e and appropriete quality programs that meet NRC requirements, guides, codes, and standards.
The NRC 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 construction permits and operating licenses i'n vendor-related areas. These inspections are done to ensure that the root causes of reported vendor-related problems are determined and appropriate corrective actions are developed. The inspections also review vendors to verify conformance with applicable NRC and industry quality requirements, to verify oversight of their vendors, and coordination between licensees and vendors.
The NRC does inspections to verify 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 found, NRC ensures that affected licensees are informed through vendor reporting or by NRC generic correspondence such as information notices and bulletins. I vii ,
3
.. . - .. - .-. - . . . _ - . - . .. ..~ .. . .- - - -.
l l
This quarterly report contains copies of all vendor inspection reports issued during the calendar quarter for which it is published. Each vendor inspection report lists the nuclear facilities inspected. This information will also alert affected regional offices to i any significant problem areas that may require special attention. This report lists selected bulletins, generic letters, and information notices, and include copies of other pertinent correspondence involving vendor issues.
t I
l 1
l l
I I
l 1
l l
l Viii
INSPECTION REPORTS l
1
acna y
't" UNITED STATES NUCLEAR REGULATORY COMMISSION f WASHINGTON, D.C. 20666-4001
%...../ July 16,1998 Mr. Gregory M. Rueger ,
Pacific Gas & Electric Company NPG-Mail Code A10D PO Box 770000 San Francisco, CA 94177 ,
Dear Mr. Rueger; i
SUBJECT:
NRC INSPECTION OF THE DIABLO CANYON POWER PLANT (REPORT NOS. 50-275/98201 AND 50-323/98201)
During the period April 6-9,1998, inspectors from the Quality Assurance, Vendor inspection and Maintenance Branch of the U.S. Nuclear Regulatory Commission's (NRC's) Office of ;
Nuclear Reactor Regulation (NRR), Division of Reactor Controls and Human Factors, ,
performed an inspection of Pacific Gas & Electric Company's (PG&E's) activities related to the L procurement and dedication of commercial grade items for safety-related service in the Diablo Canyon Power Plant (DCPP), Units 1 and 2.
The primary purpose of the inspection was to examine your program and its implementation for upgrading and dedicating commercial-grade materials in the DCPP warehouse for use in safety-related structures systems and components in the plant, particularly those that were not '
originally purchased with the intent of dedicating them. The inspection also focused on the technical and quality assurance aspects of the installation of an incorrect design ; gulator I printed circuit board in the exciter-regulator of the No.1-3 emergency diesel generator at DCPP "
l in 1994. The results of this inspection are contained in the enclosed inspection report.
Within the areas examined, the inspectors identified some instances in which the ,
documentation given to the inspectors for review did not provide reasonable assurance that l
certain technical and quality requirements were met for (1) procurement and installation of a replacement voltage regulator circuit board for the exciter-regulators of the emergency diesel
, generators (including verification of proper configuration and proper part or model number I marking on boards installed and in the warehouse) and (2) dedication of certain commercial-grade items.' In addition, the inspectors identified instances in which changes to procedures were processed in accordance_with the minor change provision in the replacement parts -
evaluation procedure when the changes were not consistent with the requirements of that provision. Because of their relatively low safety significance, these deficiencies are considered ,
minor violations of Appendix B to Part 50 of Title 10 of the Code of Federal Regulations; and in accordance with the NRC's enforcement policy as promulgated in NUREG-1600, Revision 1, no notice of violation will be issued. Nevertheless, we are concerned that sufficient resources may not be devoted to staffing cnd training in the materials services and procurement design >
engineering areas to ensure the consistent quality and suitability of purchased material and services associated with safety-related equipment applications and to avoid further problems ;
with proper review and authorization of document changes, material equivalency evaluations ,
and problems such as those cited in your procurement self assessment. Please respond within i 60 days of the date of this letter with the status of these issues and your plans to address them.
. .~ . . . . . . - . . . . .. .- - -.-.--. -.- ..- .--.-
Mr. Gregory M. Rutger July 16,1 &s5 I
in accordance with 10 CFR 2.790 of the Commission's regulations, a copy of this letter and inspection report will be placed in the NRC Public Documen' Room. Should you have any i questions concerning the attached inspect:er, report, please contact the inspection team leader, I Mr. Stephen Alexander, at (301) 415-2995.
- Sincerely,
- / /
l Suzanne C. Black, Chief i
., Quality Assurance, Vendor Inspection i and Maintenance Branch l Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation
- cc: See next page l
t Docket Nos. 50-275 and 50-323 i License Nos.: DPR-80 and DPR-82 1
Enclosure:
Inspection Report No. 50-275,323/98201 I
s ed' 3-
- . - . - - - , =.. . _
. . ._~ . . - - - ... . _ - .. .- - - ... - . - - . . . . --- - . -
Mr. Gregory M. Rueger July 16,1998 cc: i NRC Resident inspecter Regional Administrator, Region IV Diablo Canyon Nuclear Power Plant U.S. Nuclear Regulatory Commission ;
c/o U.S. Nuclear Regulatory Commission Harris Tower & Pavilion _ ]
P.O. Box 369 611 Ryan Plaza Drive, Suite 400 i Avila Beach, Califomia 93424 Arlington, Texas 76011-8064 l Dr. Richard Ferguson, Energy Chair Christopher J. Warner, Esq. :
Sierra Club Califomia Pacific Gas & Electric Company j 110011th Street, Suite 311 Post Office Box 7442 ;
Sacramento, Califomia 95814 San Francisco, Califomia 94120 Ms. Nancy Culver - Mr. Robert P. Powers San Luis Obispo Vice President and Plant Manager Mothers for Peace Diablo Canyon Nuclear Power Plant P.O. Box 164 P.O. Box 56 Pismo Beach, Califomia 93448 Avila Beach, Califomia 93424 Chairman Telegram-Tribune i San Luis Obispo County Board of ATTN: Managing Editor
- Supervisors 1321 Johnson Avenue i Room 370 P.O. Box 112 l County Government Center San Luis Obispo, CA 93406 l San Luis Obispo, Califomia 93408 l 1
Mr. Truman Bums Mr. Robert Kinosian Califomia Public Utilities Commission 505 Van Ness, Room 4102 San Francisco, California 94102 l Mr. Steve Hsu Radiologic Health Branch State Department of Health Services 1
Post Office Box 942732 Sacramento, California 94232 Diablo Canyon Independent Safety' i Committee !
I ATTN: Robert R. Wellington, Esq.
Legal Counsel 857. Cass Street, Suite D Monterey, Califomia 93940 I
i i
U.S. NUCLEAR REGULATORY COMMISSION l
OFFICE OF NUCLEAR REACTOR REGULATION l l
Docket Nos.: 50-275,50-323 License Nos.: DPR-80, DPR-82 Report No.: 50-275/98201, 50-323/98201 Licensee: Pacific Gas & Electric Company Facility: Diablo Canyon Power Plant, Units 1 and 2 Location: Avila Beach, Califomia Dates: April 6-9,1998 1
Inspectors: Stephen D. Alexander, Reactor Engineer Team Leader, HQMB Gregory C. Cwalina, Senior Operations Engineer, HOMB Billy H. Rogers, Reactor Engineer, HQMB Julio Crespo, Engineer, CSN Observers: Josef Zlathansk , Vice Chairman Nuclear Regulatory Authority of the Slovak Republic Jan Bednsi, Quality Assurance Inspector, NRASR Approved by: Richard P. Correia, Section Chief Reliability and Maintenance Section Quality Assurance, Vendor Inspection and Maintenance Branch Division of Reactor Controls and Human Factors ,
Office of Nuclear Reactor Regulation i Enclosure
. -.- ~.. - - .- -. - - - - - - - - - . - . . _ - . - - _ . _ - - -
EXECUTIVE
SUMMARY
Diablo Canyon Power Plant, Units 1 & 2 NRC Inspection Report 50-275/98201,50-323/98201 During the period of April 6-9,1998, representatives of the Quality Assurance, Vendor inspection and Maintenance Branch, Division of Reactor Controls and Human Factors, NRR, conducted an )
inspection of Pacific Gas & Electric Company's (PG&E's) activities related to the procurement and i dedication of commercial grade items and the technical and quality assurance aspects of the installation of an incorrect design voltage regulator circuit board in the 1-3 emergency diesel '
generator static exciter-voltage regulator in 1994 at the Diablo Canyon Power Plant (DCPP).
Enaineerina
. Emeraenev Diesel Generator Static Exciter-Voltaae Reaulator Circuit Board Reofacement j Review of the technical evaluation and quality assurance aspects of this 1994 circuit board ,
replacement (for which Region IV had cited the licensee for failure to perform a prompt l operability determination after the licensee's discovery of the problem in 1996) revealed )
deficiencies in the original safety classification, procurement process, review for suitability of j application and verification that all the board components met the design requirements. The ]
inspectors determined that the licensee had resolved the technical issues satisfactorily and '
j independently verified that the correct configurations of the circuit boards in question were installed in all six emergency diesel generator exciter regulator cabinets.
. Procurement Enoineerina The program for dedication and upgrading of commercist-grade, warehouse material not originally purchased with intent of dedication for safety-related service was generally technically sound and incorporated adequate quality assurance controls. ,
However, the inspectors identified deficiencies in implementation including .
)
(1) instances in which the documentation of the technical evaluations or acceptance process j did not provide sufficient assurance that all critical characteristics were adequately verified, j (2) instances in which it was unclear that replacement part evaluators recognized the difference between like-for-like (identical) replacements and equivalent replacements because they had erroneously deemed certain commercial-grade items as like-for-like replacements, :
when they should have performed equivalency evaluations, and l (3) instances in which changes to replacement part evaluations were processed using the minor change provision in the replacement part evaluation procedure when the changes did not meet the requirements of that provision. ,
i I
. Quality Assurance in Enaineerina A self assessment of replacement part evaluation activities by the Diablo Canyon independent Safety Engineering Group identified some of the same l weaknesses identified by the inspectors. The self assessment was thorough, insightful and candid. The inspeciors determined that planned corrective actions discussed with the self l essessment team and the cognizant licensee staff should adequately a' ddress the concerns identified.
1 t
~ -- .--.- -.- - - .- - - . - - . - - _ _ - _ - _ . - . - -
t i
Report Details Ill. Enaineerina E1 Conduct of Engineering E1.1 Emeroenev Diesel Generator Exciter-Reaulator Circuit Boards
- a. Insoection Scone in April of 1996, the licensee identified an incorrect component configuration on the voltage regulator circuit board in the static exciter-voltage regulator cabinet for what )
PG&E calls No.1-3 Diesel Emergency Generator (DEG 1-3) and took corrective action, but did not promptly perform an operability determination. Subsequently, NRC Region 1 IV cited PG&E for failure to perform a prompt operability determination for DEG 1-3
{
(Inspection Report 50-275,323/96-16, Violation 96-16-06). Based on RIV's review of the quality assurance (QA) aspects of the procurement of the replacement voltage regulator circuit boards in 1977 and the use of one of them in DEG 1-3 in 1994, it appeared that !
- the licensee had otherwise followed its procedures in effect at the time of the original I procurement of the spare boards and of the DEG 1-3 board replacement. However, the NRC determined that further examination of the circumstances surrounding the procurement of the spares and subsequent activities that might have detected or prevented further problems, including a more detailed review of the technical and QA aspects of this issue was warranted.
Accordingly, the inspectors reviewed current procurement procedures, those in effect at the time of the original procurement of the replacement circuit boards, and those in effect at the time of the replacement in question. The inspectors reviewed technical information and procurement records for the original procurement of DEG voltage regulators and components and also the procurement records for the spare circuit ;
boards, one of which was used in the 1996 replacement in the DEG 1-3 voltage regulator. The inspectors reviewed documents associated with the 1994 installation of the incorrectly configured replacement circuit board for DEG 1-3, the 1996 installation of a correctly configured replacement board for DEG 1-3 and three others. Finally, the inspectors performed a visual examination of all six DEG voltage regulator cabinets to verify that the present configurations of the components of concem was correct on the ;
installed circuit boards. Subsequent to the onsite portion of the inspection, on several j occasions, the licensee provided additional information that enabled certain remaining ,
issues to be resolved. !
- b. Observations and Findinas in reviewing the technical documentation associated with the DEG voltage regulator issue, the inspectors noted that the DEG static exciter-voltage regulator schematic diagram was on PG&E Drawing DC 663082, Sheet 166, Revision 2, which was a redesignation of Basler Electric Drawing D90 67100 010, Revision "l" (as in " India"), '
titled " Schematic and Interconnection Diagram for Series Boost Exciter Regulator and 3
Associated Control Functions." On this drawing, the exciter-voltage regulator was !
designated "Model No. 90 67100100." Within the exciter-regulator low-voltage chassis, the block representing the voltage regulator itself was designated " Voltage Regulator Assembly,3 PH Sensing, Model No. 90 78600101," and referenced " Schematic Dwg No. D90 78600 910."
The detailed regule. tor schematic, Basler Drawing D90 78600 910, original date: I November 12,1970 (revision block blank), titled " Schematic -Voltage Regulator - SVR 1 Modified" showed the regulator components that are mounted on the circuit board in question within a dashed-line block annotated " Etched Circuit Board." This drawing did not give the model number for the regulator or for the printed circuit board itself. The l board-mounted components comprise most smaller electronic control components of l the regulator; the larger, heavier ones, such as the transformers, being separately
]
mounted on the metal voltage regulator chassis in the exciter-regulator cabinet upon which the circuit board is also mounted. The circuit board-mounted components in )
question were resistor R26 and capacitor C5. The regulator schematic showed R26 as two resistors in parallel and originally showed CS as a separately designated single i capacitor in parallel with capacitor C4. The Basler detailed regulator schematic had been redesignated PG&E Drawing DC663082, Sheet 60. Revision 4 of this drawing, dated September 25,1996, showed capacitor C5 removed by a drawing change annotation that cited Field Correction Transmittal 20914, dated July 3,1996. The operation and service manual for the exciter-regulator, stated that capacitor C5 (with .
C4) was part of the feedback loop in the voltage regulator stability circuit and that R26, !
in conjunction with capacitor C8, formed a filter network across the control rectifier bridge. It was the requirement for C5 and the corresponding values of associated .
feedback loop resistors R15 and R16 when C5 is installed that were in question initially in 1996. Subsequently, the configuration of resistor R26 came into question when the licensee discovered that the R26 for the DEG 1-3 and then the DEG 2-3 voltage regulator circuit boards were different from the drawings and from the other R26s.
The inspectors' review of microfilm prints of archived procurement and maintenance records revealed that the voltage regulator printed circuit board has had a complicated history of changes, errors and confusion involving its part number, model number, and ,
proper configuration for the Diablo Canyon DEGs. The original PG&E purchase order *
(PO) for the four spare boards, PO No. 558018, dated November 10,1977, ordered them as part number 90 32101-105. The associated purchase requisition, No. 7745715, dated July 29,1977, originally specified Basler Part No. BE-32101-105. Review of the licensee's history and evaluation cf the circuit board problems indicated that this part number was chosen because the circuit boards were silk screened with that number and no part number for the circuit board itself appeared on the Basler drawings. On the basis of telephone conversations and correspondence with Basler in 1996, the licensee determined that the first error was Basler's marking the origina! circuit boards with the !
wrong part or model number. In addition, Basler had supplied a generic voltage regulator schematic to PG&E. Although the 1970 drawing did show R26 as a double resistor, it also showed C5 which is not used on all models and without any notes to indicate which circuit board types were supposed to have which components, and more importantly, which one Diablo Canyon was supposed to have.
4
According to a Basler letter, dated April 25,1996, in response to licensee inquiries, Diablo Canyon's voltage regulators were Model 90 786100101 in accordance with the i Model 90 6700100 exciter-regulator drawing. However, that type of voltage regulator i can have any one of five different configurations of " etched" circuit board. The letter stated that Diablo Canyon's regulators were all built according to configuration codes specified in Basler's " Build Book " (one of the mar,ufacturer's master design control documents). Various codes from the Build Book form a so-called " style number" which determines how the voltage regulator is to be built, including, for example, major 4
component specifications and interface features. The style number of Diablo Canyon's regulators was "SVRO1 A05B2B1B." The "05" code in the Build Book" specified circuit board assembly Part No. 90 32101104. However, even though the boards were reportedly assembled as 104s, the style number code 05 was apparently misinterpreted by the personnel marking the boards and the model number 90 32101 105 was si!k-screened onto the fronts of the boards instead of 90 32101 104 as it should have been.
When the spare boards were supplied in 1977 under PO 558018, Basler shipped boards that were marked 90 32101 104 (consistent with the 05 Build Book code) and not with
, the 105 suffix as ordered. Basler first explained this substitution in a February 26,1984, letter as discussed below. Consistent with the 104 design, capacitor C5 was not installed and R15 and R16 were of the correct values for when C5 is not installed.
However, in May of 1996, the licensee learned that the second error by Basler was that one of the several modifications (component upgrades) and model redesignations 4 prescribed on Basler Drawing 90 78600, Revision 8, dated November 12,1970, to be l performed on all circuit board types (90 32101-100 through 104 became 101 through 109 respectively) used in the 90 78600-series regulators, had not been done on the circuit boards shipped under PO 558018. Specifically, the boards had not been ,
modified to change R26 from a single 100-ohm,20-watt resistor to two 200-ohm,25- '
watt resistors on a bracket and had not been remarked with modified Model No. 90 32101 109. Further, Basler had not changed the part number of the modified boards i
< used to build Diablo Canyon's original regulators from 90 32101 104 to 90 32101 109 as l required by the instructions on the drawing. Therefore, when determining that the Diablo Canyon regulators were built with boards that were originally Model No. 90 32101 104, Basler should have verified that they were configured in accordance with the 1970 modification of the double R26, and remarked them 90 32101 109 before shipping.
.Thus, the so-called "new" number described in a Basler memorandum faxed to PG&E on May 7,1996, i.e., 90 32101 109, "to avoid confusion" was not new at all, but had been established in 1PO, and not put on either the original regulator boards or the spare boards supplied in 1977. The spare boards had not been remarked with the 109 suffix possibly because they had not been modified with the double R26 as they were supposed to be; although the original boards supplied with the regulators had double R26s and the other 1970 upgrades.
During further review, the inspectors noted additional discrepancies in the earlier documents. The 1977 spare board purchase requisition indicated that the voltage regulator was Model No. "SBSR-HV" The inspector noted that this model designation ,
was in the title of the Basler operation and service manual for the exciter-regulator, but not on the equipment name plates. The inspector noted during the walkdown that the 5
9-
model number on the name plates on all of the original five exciter-regulator cabinets :
was SBHV 67100-100. The model number on the exciter-regulator cabinet of the sixth
! DEG, DEG 2-3, was SBHV 9244900-100. The inspector determined that the exciter-regulator name plate numbers, also referred to as system part numbers, except for DEG 2-3, corresponded to an exciter-regulator design consistent with the exciter-regulator drawing, but Model No. 90 67100100 was actually shown on the drawing. The inspectors later determined that the number on the DEG 2-3 exciter-regulator name l plate (92 44900100) was a Basler generic system part number for an exciter-regulator cabinet because this unit had been built about 20 years later than the original five cabinets, had been purchased separately from the other equipment as a commercial-grade item, and had been wired on site. The equipment was also identified on the l purchase 1977 requisition as "Ser.126," which, the inspectors noted, was the serial number for the DEG 1-2 exciter-regulator; although it was not clear why the serial ;
number, if relevant, of only one of the five original DEG exciter-regulators was included i
on the purchase requisition for spare boards for all the regulators.
)
The parts list sheet attached to the original 1977 purchase requisition and purchase !
order for the spare boards also indicated that the stock code for the board was 42-1841. ;
Note that this is the stock code that was later (1996) assigned to the so-called "new" l Basler part number for the boards configured for Diablo Canyon's DEGs,90 32101 109.
The record then contained what appeared to be a photocopy of the original requisition l
parts list sheet on which the part number was changed (pen and ink) to BE-32101-104 1
(with the annotation "Model 32101-104") and the stock code was now 37-3770, but the original stock code did not appear. It was not clear when these changes had been '
made. The QA requirements sheets filed with the purchase order and requisition indicated that Basler was not on the qualified suppliers list (QSL), but was exempt because of being the original equipment manufacturer (OEM). Although it was clear that the purchase was for material to be used in safety-related applications, no quality requirements were specified except receipt inspection. The typed "PG&E Code" on the purchase order was the material class (87) and the material code (4200). However, on the receiving copy of the purchase order, the 87-4200 was lined out and changed to 37-0770. On the copy that indicated it would go to the vendor, the 87-4200 was changed to 42-1841, but again, it was not clear when or why these changes were made.
On January 6,1978, PG&E issued Minor Change 1 (MC01) to its PO 558018 to incorporate a statement of the applicability of 10 CFR Part 21 to the purchase. Although Basler was not a 10 CFR Part 50, Appendix B, supplier, and was not on the Diablo Canyon Qualified Suppliers List (as mentioned above, it had been exempted because of being the OEM), the inspectors did not identify any Basler correspondence in which the vendor took exception to that added purchase condition.
The receipt inspection report, indicating initial shipping damage inspection on February 8,1978, and item receiving inspection on February 9,1978, indicated that all as-received conditions considered applicable by the QC inspector were acceptable (including " Gas Blanket" which was inexplicably checked as "OK" and not marked "N/A" as were some other blocks that could be applicable, e.g., " desiccant"). Also checked as "OK" was the "l.D. & Markings" block, yet the circuit boards received were marked with 6
1
( the model number 90 32101-104 instead of BE 32101-105 as ordered. Although the l receipt inspection report did not document the discrepancies, it indicated under l
" Document Verification" that the "PN Clarification," was acceptable.
The part number clarification referred to was a Basler letter, dated February 6,1978, in l
response to PG&E questions and received shortly after the shipment. The Basler letter
- referenced PO 558018 and a telephone conversation with a Mr. J.W. Weldy [of PG&E]
and stated Basler's understanding of the intent of the PO, i.e., that it was for "the correct Printed Circuit Board Assembly used in Basler Senior Voltage Regulator, Model SVR01 A0582B1B"[ sic). However the letter did not explain at the time what this number was (later determined to be a style number, not the model number), nor did PG&E I documents. The fact that it was completely different from the model numbers on the drawings and on the equipment caused considerable confusion during later reviews, but was not questioned at the time. The letter then explained that " Circuit Board Assembly, P/N 90 32101104 is the correct unit, and therefore was supplied on the referenced i purchase order." The inspectors concurred that the vendor had supplied incorrectly j configured and marked spare circuit boards, but it was at this point that the licensee I
missed its first opportunity to detect the problem and prevent the eventual installation of the nonconforming material. There was no documented evidence that the apparent discrepancies in circuit board part numbers and in the regulator model numbers prompted the licensee to verify the information against name plates and drawings and also perform a detailed physical comparison between the spare boards received and l those already installed in the regulators. However, the procedural requirements in effect I at the time also did not prompt the licensee to perform such verifications in view of the assurance by the original equipment designer and manufacturer that the parts supplied were correct.
On June 21,1983, the PG&E QA Department issued Audit Report No. 83074P for the audit completed April 29,1983, on Diablo Canyon's procurement and acceptance l process. As a result of this audit, the PG&E QA department initiated Nonconformance Reports (NCRs) DCO-83-PG-N020, N021 and N022. N020 reported that there was insufficient documented evidence of the attributes checked (and acceptance criteria) in receiving inspections for "off-the-shelf" items used in Class 1 applications. N021 documented that nine 1982 procurements had been identified in which material that was required to be purchased from qualified suppliers, had been purchased from suppliers that had not been qualified by PG&E QA. But this NCR did not list PO 558018 among the nine cited as deficient. NO22 dealt with procurement involving material substitutions without prior approvals. The prescribed resolution for the NCRs, and in particular, for the one most applicable to PO 558018, NO20, was to review receipt inspection records for this type of material and take corrective action on an individual basis.
l One of the Purchase Document Change / Variance forms generated as a result of this review, dated January 6,1984, addressed the items purchased under PO 558018, although a block on the form was checked indicating that it was associated with the resolution of NCR DCO-83-PG-N021 instead of N020. Nevertheless the evaluation addressed receipt inspection of the "off-the-shelf" items purchased on the referenced PO (the NCR N020 issue) and not whether the supplier was on the QSL (the NCR N021 l
7 f
issue). This evaluation was inadequate in that it did not directly respond to the NCR issues and simply asserted that the purchased items were " standard, commercial grade, off-the-shelf items," that "There are no special requirements beyond routine assembly line testing," that " qualification for use in safety related applications is through receipt inspection," and that " Proper receipt inspection was performed and documented." Thus it completely ignored the NCR N020 issue of inadequate inspection attributes and acceptance criteria for "off-the-shelf" material. This evaluation reaccepted the 1977 procurement of the spare DEG voltage regulator circuit boards on the basis of what was deemed a satisfactory receipt inspection despite the NCR N021 issue. Thus, another
- opportunity was missed to correct the original error, i.e, that Basler had supplied incorrectly configured boards.
The next opportunity identified by the inspectors to correct the circuit board problem came in 1994 when one of the spare boards was used in the DEG 1-3 voltage regulator. ;
Work Order C0133243, Work Activity 02, indicated that Diablo Canyon technicians replaced the original voltage regulator circuit board in the exciter-regulator of DEG 1-3 on December 15,1994, because the regulator had been found to be faulty after erratic voltage control had been experienced during a surveillance run. According to the work '
record parts documents, including the "PIMS lssue Ticket," (PIMS is the " Plant information Management System" datab&se) for the spare regulator board, the faulty regulator board was replaced with one of the four spare replacement boards from the onsite warehouse, Stock Code 42-1841, that had been originally procured from Basler under PO 558018 in 1977. Review of the Integrated Parts Catalog (IPC) data in the PIMS database revealed that Materials Services had changed the stock code from 42-1841 to 37-3770 in 1984. Then in 1994, at the time of the replacement, Stock Code 37-3770 had been changed back to 42-1841 (and the purchase class changed from 7 to 5),
apparently on the basis that the original procurement under PO 558018 was viewed as a safety-related procurement.
The 1984 evaluation (resolution to NCR DCO-83-PG-N020) which changed the stock code to 37-3370, as discussed above, indicated that the boards were acceptable on the basis of satisfactory receipt inspection as stated previously On this basis, Materials Services did not perform any additional replacement or equivalency evaluation of the circuit board prior to changing the stock code back to 42-1841 when it was installed in 1994. Although the inspectors determined that the NCR evaluation was inadequate, the use of the spare board from warehouse stock without fuither evaluation was allowed on this basis by procedures in effect at the time of the DEG 1-3 voltage regulator circuit board replacement on December 15,1994. In particular, Revision 2 of Administrative Procedure AD9.lD4, " Establishing Procurement Technical and Quality Requirements,"
Section 5.4.2, " Acceptance," allowed acceptance on the basis of receipt inspection (5.4.2.a.2), alone if certain requirements were met. In addition, Section 5.4.2.a.5 ;
provided for acceptance on the basis of post-installation testing (with a prerequisite of receipt inspection). The inspectors confirmed that satisfactory post-installation tests and results were documented in the WO C0133243 file for the DEG 1-3 voltage regulator (and governor) troubleshooting and circuit board replacement. Nevertheless, this was the third missed opportunity to detect the incorrectly configured circuit boards. The last missed opportunity to detect and correct the circuit board component configuration 8
l l
problems was the installation itself. The work order did not require the technicians to document performing a detailed comparison between the installed board and the replacement, apparently because the part from the warehouse was presumed to be correct. The notes section of the work record did not centain any reference to the fact i
that the replacement circuit did not have capacitor C5 and that R26 was a single resistor instead of double.
The inspectors' review of documents associated with the licensee's 1996 discovery of l the problems with the circuit boards revealed the following: In April 1996, Procurement l Design Engineering (PDE) personnel were developing Revision 3 to Replacement Part Evaluation (RPE) E-6802 which was originally developed to accept the voltage regulator supplied as a separate commercial-grade item by Basler in 1990 for DEG 2-3. Revision 3 of RPE E-6802 was to be used to accept regulator circuit board that had been removed from the DEG 1-3 exciter-regulator cabinet in 1994 and repaired by Basler in 1995 and any future replacement boards.
Under PG&E PO D554605 issued February 1,1995, Basler performed a failure analysis and then repair of the board removed from the DEG 1-3 voltage regulator. PO 554605 compounded previous errors by requesting that Basler perform a failure analysis on a voltage regulator board "Model 32101-10, Part# 90-32101-00113"[ sic). The incorrect I model number had been written on the associated purchase requisition (Material Requisition 4073677, dated January 26,1995) and the incorrect part number was the result of misreading the part numt,er written on the requisition 32101-001B. It was not clear where these numbers came from, or which is correct, if either of them is, in addition, the standard clauses attached to PO 554605 stated (on the same page) that 10 CFR Part 21 was applicable (Standard Clause 5012), but that the purchase was not safety related (Standard Clause 5021). The licensee could offer no satisfactory explanation for these contradictory statements except that on the procurement package cover sheet in PO 554605 record, PSG File.No.15523, which referred to Material Requisition 4073677, the "Nonsafety-related" block under " Classification" was erroneously marked, in the course of examining the records and the equipment in connection with the replacement part evaluation, PDE discovered discrepancies between installed circuit boards and the design drawing referred to above. These were first documented in Action Request (AR) A0399437 on April 17,1996. AR A0399437 stated that capacitor C5, which according to Drawing DC663082, Sheet 60, was supposed to be in parallel with capacitor C4 in the voltage regulator feedback loop of the stability circuit, was missing on the regulator boards in the exciter-regulator cabinet for DEG 1-3 the cabinet is designated SED 13). The AR stated that a walkdown of Unit 2 revealed that the regulator board in SED 21 (for DEG 2-1) also did not have C5 installed. This AR was written before PDE had determined that C5 was supposed to be absent, and that the boards with C5 (in SED 12,22 and 23) were incorrect. The AR then listed severalitems it characterized as documentation and procedure problems: (1) That Basler had repaired a circuit board Model No. 90 32101 100 which was placed on hold in the Diablo Canyon warehouse under Stock Code 42-1841 and listed as Part No. 90 32101 104, (2) that the board now had C5 installed, which was not like other 90 32101 104s, (3) that 9
PO 554605 incorrectly listed Part No. 90 3210100113 instead of 0018 (as discussed above), (4) that the PO listed Model No. 90 32101 10 (also discussed above) which the AR incorrectly stated should have been 90 32101 100, presum6oly based on Basler's erroneous failure report, (5) that there were three more boards in stock under 42-1841 that do not match the design drawings, (6) that they were transferred from stock code 37-3770 to "O class" stock and accepted for safety-related use, and (7) that this was contrary to Section 5.3.5 of Procedure AD9.lD4 which the AR stated required an "NES l evaluation" to dedicate comma cial grade items for safety-related use. Although the !
inspectors determined (as discussed below) that dedication was not required at the time l of the original purchase (1977), nor at the time of the installation of a spare board (1994) by the procedures in effect at the time, the AR further stated, quite correctly, that the boards were upgraded from Purchase Class 7 (nonsafety-related) to Class 5 (for safety-related use) and that they were accepted : Based on PO 558018 without any dedication !
or evaluation of the differences between the original and the replacement. The AR then went on to relate the history of the issue, generally as also determined by the !
i inspectors, with a few factual errors such as misquoting the style number from the 1978 Basler letter explaining its substitution of 90 32101 104 boards for 90 32101 105s. l i
1 It was later determined (in part by the Basler failure analysis repwt for the board) that because PO 554605 had incorrectly specified Model No. 90 32101 10, and also because the board was marked with Model No. 90 32101 105 (which was originally a 90 32101 100 before the 1970 R26 modification) Basler assumed that the board was a l i
Model No. 90 32101 100 (which was supposed to have C5) and treated it as such, in fact, even though_ the board was really a 90 32101 109 (s modified 104 with double R26 ,
and no C5), Basler, without verifying the number against its original manufacturing l records, believed C5 to be missing and installed one. It was not clear whether Basler checked the values of R15 and R16 to confirm whether C5 should be installed. ;
Subsequent related ARs A0399596 and A0401207 and Quality Evaluation QOO11867 l further addressed the issues first identified in AR A0399437. One of the corrective actions was to walkdown all six exciter-regulators as the inspectors did during this inspection. The licensee found that C5 was not present on the boards in exciter-regulators SED 11,13 and 21. C5 was found in SED 12,22, and 23. The licensee determined that SED 12 had C5 because the entire voltage regulator had been replaced in 1995 wit a spare regulator purchased from Basler that had the incorrect stability circuit feedback loop. However, no discrepancies were identified during this installation either and no evaluations or detailed inspections were performed or required by ,
procedures. The faulty voltage regulator replaced in 1995 had been examined and 3
repaired by Basler under PG&E PO D554794, issued September 26,1995. SED 23 had l l
C5 in part because of the way it was procured and wired separately in 1990 as part of !
the addition of the sixth DEG. The licensee was not able to determine why SED 22 was l originally supplied with C5 on its regulator board.
i During one of the walkdowns in response to the ARs, the licensee found that in exciter-regulator cabinets SED 13 and 23, R26 was a single,100-ohm,20-watt resistor instead !
of two,200-ohm,25-watt resistors in parallel and mounted on a bracket as in the other j
, four units. AR A0401207 was written to address the R26 issue and revised to address
- 10 I
l
- . - - _ . - . - - - - - . . . . - . - ~ - . - - - . . - . - - - . - - . - - -...
l the fact, discovered later, that the R26 in SED 13 had overheated, become discolored and had scorched the printed circuit board. The licensee also found that the replace-men boards from the warehouse had a single R26 of 100 ohms (equivalent to the two 200-chm resistors in parallel). However, the single resistor, while having the same l
resistance as the two in parallel, was only rated for 20 watts and therefore could not dissipate as much heat as the double R26 configuration without becoming too hot. In addition, not being mounted on a metal bracket which acts as a heat sink (as were the double R26 resistors), and therefore also being closer to the printed circuit board surface, was the reason the single R26 in SED 13 had overheated.
t in consultation with Basler, the licensee confirmed that the correct configuration of the
- comoonents in question on the circuit board should be a double R26 on a bracket, no C5, and R15 and R16 of the values required with no C5 (100k ohms and 350 ohms respectively). As stated previously, Basler had assigned part number 90 32101 109 for this configuration, to which Diablo Canyon Materials Services reassigned Stock Code 42-1841. Accordingly, Revisiori 04 to RPE E-6802, Minor Change 01, dated July 11,1996, specified Standard Clauce 9792 (Revision 2) which, among other inspection acceptance criteria, specified the correct configuration for C5 (and R15 and R16) and R26, along with voltage regulator performance criteria.
4 During the visual examination of the intemals of all six exciter-regulator cabinetc, the inspectors noted that the model number for the whole voltage regulator (90 78600101, shown on the exciter-regulator schematic, drawing DC663082, Sheet 166, Revision 2) was not visible inside the cabinets. However, stamped or stenciled on the front of the voltage regulator printed circuit boards in question were several different circuit board moc'e! numbers. The circuit boards for DEGs 1-1,1-3, and 2-1 bore the model number 90 32101-105, the same number used to order the spare boards in 1977. However, the number nn the board for DEG 1-2 was 90 32101-109 (with the "109" suffix apparently handwritten with a permanent marker). The number on the DEG 2-2 circuit board was 32101-104 (the number that appeared on the spare boards supplied by Basler in 1977).
The third digit of the suffix of the number on the DEG 2-2 board appeared to have been altered, but was illegible. Nevertheless, the inspectors' visual examination of the circuit
- boards in all six DEG voltage regulators also confirmed that all the R26 resistors were of the correct double 200-ohm,25-watt resistor, bracket mounted configuration, and that all C5 capacitors had been removed.
- c. Conclusions On the basis of records associated with the 1904 DEG 1-3 voltage regulator circuit board replacement, the spare board procurement, and subsequent evaluation, the inspectors concluded that there were deficiencies in the original safety classification, procurement process, review for suitability of application and verification that all the board components met the design requirements. Many of these deficiencies were identified by the licensee in Action Requests (ARs) written by Procurement Design Engineering in 1996. However, although those procurements and acceptance processes would not meet the licensee's present standards, the inspectors concluded that the 1977 procurement was done in accordance with the procedures in effect at the 11
4 time. The inspectors further concluded that the procedures applicable to the
- replacement part installation in effect in 1994 were followed, but that they were weak in
~
that they did not require a reevaluation of the replacement part when it had been originally accepted solely on the basis of (1) having been purchased from the OEM and j exempted from having to be on the QSL when the spare boards were purchased for the
- exciter-voltage regulator, (2) had been reaccepted on the basis of receipt inspection by 1984 NCR evaluation which the inspectors concluded was inadequate. The fact the
] board had pas ,ed post-installation tests, even though the supplier qualification and l . technical and quality requirements imposed on Basler originally would not have met the i licensee's procurement and acceptance requirements in 1994, partially compensated for .
the lack of reevaluation, but did not preclude installation of the incorrectly configured I l
{ board. The inspectors further concluded that the work authorization and control '
l documents used in 1394 when the circuit board was replaced with one of the incorrect i part number Ond with an incorrectly configured R26, were weak in that they did not i effectively ensure a close visual comparison to verify that the circuit board component
! configuration and markings matched that of the original circuit board being replaced and l that all discrepancies were resolved. i l
The inspectors further concluded during the onsite portion of the inspection that the
- licensee had resolved the technical issues satisfactorily with the exception of the
!' necessity for capacitor C5 and the correct marking of the spare and installed boards.
Subsequent to the inspection, the licensee provided additional documentation from
- Basler that es isfactorily addressed the inspectors' concerns with respect to C5. On the l basis of that information as well as on satisfactory voltage regulator operation during
- tests, the inspectors concluded that the DEG voltage regulators were acceptable as-configured without C5. Confirmation of the licensee's strengthening of procedures as required, verification of the configuration of the boards in the warehouse and of placing l
the correct markings on all the boards on site is identMied as inspector Followup Nm j (IFI) 50-275,323/98201-01.
E1.2 Reolacement Parts Evaluation
- a. Insoection Scogg The inspectors reviewed Diablo Canyon procedure CF3.lD13, " Replacement or New Part Evaluation (RPE)," Revision 4, dated December 10,1996, and several RPE packages to determine how the licensee evaluates replacement items for use in safety-related structures, systems and components. Although an RPE can cover a wide variety of items, the inspectors limited their review to commercial grade items, with emphasis on those that were not originally purchased with the intent to dedicate them for safety-related service. To evaluate implementation of the program, the inspectors l
reviewed selected RPE packages and also the data in Diablo Canyon's parts database called the Inventory Parts Catalog (IPC). ,
i
)
i 12
- b. Observations and Findinas b.1 RPE Procedure RPE procedure CF3.lD13 established the process by which the Diablo i Canyon Procurement Design Engineering (PDE) and Materials Services (MS) staff ;
evaluates, specifies and verifies both technical and quality requirements for replacement !
or new items. According to CF3.lD13, an RPE is used to perform equivalency evaluations (confirms that a replacement part, which is not line-for-like, will perform its l intended function), specify verification activities and implement minor modifications. !
CF3.lD13 specified the responsibilities for those involved in the process, including the j
requestor, preparer, independent technical reviewer and the supervisor. Attachment 8.1 to CF3.lD13 provided the forms to be used in the RPE, although not all forms are i needed for every RPE. CF3.lD13 required, for replacement or new parts, that the preparer identify the safety function of '}e parent component (component in which the specific item is installed), function of the item, credible failure modes of the item and the effects of the failure modes. Lastly, the preparer must determine if the failure of the item would prevent the parent component from performir;q its safety function.
The inspectors found that CF3.lD13 was comprehensive and adequately addressed commercial grade dedication requirements with some minor exceptions. Appendix 7.2, RPE Form Instructions," included a description of Verification Activities as the four acceptance methods presented in Electric Power Research Institute (EPRI) Report NP-5652, " Guideline for the Utilization of Commercial Grade items in Nuclear Safety Related Applications (NCIG-07)." However, the description of Method 2, Commercial Grade Survey, did 'not include the NRC restrictions described in NRC Generic Letter 89-02, " Actions to improve the Detection of Counterfeit and Fraudulently Marketed Products." In addition, Page 12 of Attachment 8.1 required the preparer to determine if the replacement item met "all three CGI criteria," referring to the previous definition of a commercial grade item in 10 CFR Part 21. The NRC changed that definition for 10 CFR Part 50 licensees in 1995. Materials Services-PDF. personnel committed to change the 1 procedure appropriately. No further concems were identified.
b.2 RPE E-1011- Fasteners for Electrical Connectors RPE E-1011, used to dedicate commercial grade fasteners (bolts, cap screws, nuts, washers and lock washers) for Class 1E electrical connections, had been prepared in accordance with CF3.lD13. The RPE identified the critical characteristics of dimensions and material, specifying dimensions to be verified by measurement (with samples sizes and acceptance criteria given) and material to be verified, also on a sample basis, by an " analysis method consistent with material acceptance requirements," The RPE specified the acceptance criterion for material simply as: " Material exhibits the characteristic of silicon bronze (copper-silicon alloy)," and did not specify the verification method or a more specific set of acceptance criteria. The inspectors questioned the material analysis method and i acceptance criteria. The licensee explained that test methods and acceptance criteria i were, at one time, specified in RPEs. However, the licensee stated that experience had )
shown that test personnel were more familiar with test equipment (including new or updated equipment) capabilities and limitations and are, therefore, better qualified to determine tre best test method and equipment. Further discussion identified that test personnel utilize the standard specified in Diablo Canyon procurement documents.
13 1
- The licensee's inventory Parts Catalog (IPC) data sheet for Part No. 94-5914, a silver-bronze hex head screw, indicated that the part was dedicated in accordance with RPE E-1011. The inspectors noted that the IPC included a limitation (Standard Clause 6177) regarding the use of the screw stating, "This item has been evaluated by RPE E-1011 as safety related when used in bolted electrical connections only. Any other safety j related use must be approved by engineering prior to installation." That clause i effectively restricted the use of RPE E-1011 and parts dedicated with it for specified applications. The inspectors considered the use of such clauses in the IPC, noting a limited dedication, as a good practice which should help prevent part misapplication.
Based upon the inspectors' interviews of materials analysis personnel and a detailed guided tour of the materials testing laboratory, the inspectors were satisfied with the test personnel's ability to properly verify material critical waracteristics. The inspectors ;
considered that consulting with the materials testing laboratory to establish the ,
applicable standards and the best material verification methods and acceptance criteria )
to be an appropriate, even necessary practice. However, the inspectors pointed out that the licensee had established the RPE to be the design output document in which the entical characteristics, verification methods and acceptance criteria are specified. ;
Although the inspectors did not identify any inadequate dedications attributable to the l use by the materials testing laboratory of inapproprMe material verification methods or ,
acceptance criteria, even though they were not specified in detail in the RPE, the ;
inspectors considered the practice of not specifying any verification methods, standards or acceptance criteria, once recommended by the materials testing laboratory, to be a weakness.
b.3 RPE E-8272 "EDG Voltaae Reaulator Linear Reactor TS1" RPE E-8272 is used for l dedicating what the RPE refers to as " current boost transformers"(CBTs) made by. f Basler Electric Company (Basler Part number BE 10771-002) used in the emergency diesel generator static exciter built by Baer (Note that PG&E uses the initialism "DEG" for " diesel emergency generator" elsewhere). According to the DEG static exciter-voltage regulator schematic diagram, PG&E Drawing DC 663082, Sheet 166, Revision I 2, a redesignation of Basler Drawing D90 67100 010, Revision "l"), the three l components with Part No. BE 10771-002 are designated T51, T52, and T53. According !
to Basler Operation and Service Manual 90 67100-990, the components designated '
TS1, T52 and T53 are called saturable transformers. Apparently used as CBTs, the three saturable transformers are a type of magnetic amplifier that help control the DEG field excitation current in response to voltage regulator output sigrwis on their control windings. The inspectors noted that the title of the RPE was incorrect in that T51 is not a the only CBT and it is not a linear reactor. The Basler manualindicated that the linear ]
reactors are designated L51, L52 and L53. Nevertheless, the RPE identified ,
I appropriate dedication activities to be performed, including verifying insulation resistance and no-load voltage ratio of the CBT. RPE E-8272 stated that upon l
successful dedication of a CBT, its Diablo Canyon part number or stock code should be changed from 27-6697 (CBT commercial grade part number) to 95-4535 (CBT safety- ;
related part number). I
\
14
. _ . - ~.- - . - - - . - - - . - - - - - - . - - . - . - - -
I in reviewing the Diablo Canyon Inventory Parts Catalog (IPC) data sheet for the CBT, commercial-grade part number 27-6697, the inspectors noted that Materials Services had changed the part numbers of two CBTr M 27-6697 from 95-0357 which was the safety-related part number for a current trst:ormer (CT), Basler Part No. 10772-001.
l Basler BE 10772-001 cts are used as CT1, CT2 and CT3, as shown on the DEG static l exciter-voltage-regulator schematic, to obtain power from the DEG generator output for field excitation. The inspectors questioned why a safety-related part number had been i changed to the commercial grade part number of a different part.
i The licensee reported that it had determined that the Basler BE 10771-002 CBTs' had I been misidentified upon receipt at Diat:lo Canyon as Basler BE 10772-001 cts, most probably due to the very similar Basler part numbers, and therefore had been mistakenly assigned the CT commercial part number 27-6694. According to the licensee, Materials Services then attempted to upgrade the commercial grade CBTs to 4
safety-related status by use of RPE E-6795, the RPE for dedicating the cts, because l the CBTs had been erroneously assigned the CT commercial part number. During the dedication process, Diablo Canyon technicians identified the incorrect part number and l
4 effected a change from the safety-related part number to the correct commercial grade part number. At the completien of the inspection, the dedication of the CBTs using the correct RPE, E-8272, was in progress. The inspectors were satisfied with the explanation for the changed part numbers. The inspectors also considered Diablo Canyon's ability to track each component through the system and identify problems associated with each part as a system strength.
b.4 RPE E-8346 - SF6 Breaker Charaina Motor Mountino Fasteners Revisions 00 through 02 of RPE E-8346 had been developed to upgrade commercial grade fasteners for use in SF6 safety-related breakers. Revision 02 provided acceptable specifications for bolts and nuts, but the inspectors noted that Revision 02 identified both a like-for-like and an equivalency evaluation. The inspectors questioned why a like-for-like and an equivalency were both identified and therefore, reviewed all earlier revisions to the RPE.
The inspectors found that the previous revisions had identified other acceptable fastener specifications as follows:
Rev.No. Soecification ggy 0 SAE J429 Grade 8.1 or higher (screw / bolt) medium carbon alloy SAE J429 Grade 8 or higher (nut) medium carbon alloy 1 DIN 933 Class 8.8 (screw / bolt) medium carbon alloy DIN 934 Class 8 (nut) low carbon MC01' ISO 898-1(screw / bolt) medium carbon or medium carbon alloy ISO 898-2 (nut) low carbon
' Minor Change 15
-~ -- . .- - -. _ - . . - . - .
l 2 ISO 898-1 or ASTM F568 or medium or low SAE J1199 Class 8.8 (screw / bolt) carbon q ISO 898-2 or ASTM A563M or medium or low SAE J 1199 Class 8 (nut) carbon Based upon their review, the inspectors were concemod regarding the licensee's i determination of like-for-like in Revision 2 of the RPE. The inspectors interviewed a procurement engineer involved in the process. The engineer stated that he believed 1 that the like-for-like block was checked on the RPE form for clevision 2 because MC01 ,
had already identified the ISO standards as acceptable r devision 2 added the ASTM and SAE standards.' Therefore, like-for-like was appropi. .e for the ISO standards while l an equivalency evaluation was needed for the others. Although the explanation may have seemed reasonable from the engineer's standpoint, it raised a concem regarding the RPE process.
l Procedure CF3.lD13 prescribed the process for implementing a so-called minor change (MC), i.e., a document used to approve and transmit minor corrections to an issued RPE, but which does not require the same levels of review and approval as a revision to an RPE. The procedure required that for a change to an RPE to be treated as a minor change, it must be within the scope and intent of the RPE and not conflict with any statement, basis, logic or requirement of the effective revision of the RPE. The inspectors determined that the use of the MC process here to incorporate a specification change, as well as to expand the scope of the acceptance criteria (including medium carbon steel) was contrary to the minor change requirements in CF3.lD13 as well as to the intent of its minor change provision. Further, MC01 did not address the basis for changing the specification from the DIN to the ISO specifications. The inspectors determined that the lack of a documented basis for changing the material specification <
constitutes a failure to review material for suitability of application as required by :
Criterion lli of 10 CFR Part 50, Appendix B. However, the changes made to the material ;
specification were not safety significant and Revision 2 provided an acceptable equivalency evaluation. Another example of what the inspectors considered to be an inappropriate use of the minor change provision involved RPE E-8001 (Revision 0) which was developed to accept a replacement component cooling water pump motor (early 1995 or late 1994). Subsequent to issuing E-8001, Revision 0, the vendor proposed to substitute a different frame number for the motor ordered. MC01 to Revision 0 was used to approved the replacement frame number which tumed out to be the wrong size. While the wrong size motor was never shipped, and would not have been able to be installed, the inspectors determined that use of a minor change to approve a change in an attribute as fundamental as the size of the replacement component was inappropriate in that it did not get the same level of review as a full revision presumably would have received. More complete evaluation and review might have prevented approval of an improperly sized replacement component.
Further discussions with the procurement engineers indicated that Diablo Canyon will, on occasion, use the MC process to expedite needed changes to an RPE. The inspectors were concerned that this misuse of the MC system may result in materials l being utilized in the plant that have not had a proper RPE performed. The inspectors 16 1
did not identify any other instances of improper use of the MC provision, however, I based upon the discussions with Diablo Canyon employees, the improper use of the MC process is considered a weakness. However, after discussions with the leader of the licensee's procurement and dedication self assessment team, the inspectors were satisfied that the licensee's planned corrective actions in response to the self assessment (discussed in Section E2 below) should adequately address the concerns identified.
b.5 RPE M-8572 - Motor Driven Auxiliarv Feed Pumo Parts The inspectors reviewed RPE M-8572, used to evaluate the acceptance of chrome plating an auxiliary feedwater pump internal rotating element (impellers and balance sleeve). The chromium buildup was determined necessary because the existing rotating element was found to be undersized. The manufacturer of the element recommended the chromium buildup as an acceptable repair method and performed the actual repair.
The inspectcrs found that Revision 00 of the RPE (April 17,1997) stated that the item was a like-for-like replacement. The rationale for this statement was that there was no change to the rotating element materials, despite the fact that the RPE identified the addition of chromium. During an internal review of the RPE process by what the licensee called the SPlGOT (not an acronym) Action Request (AR) Review Team, Diablo Canyon personnel reviewed Revision 00 of the RPE and determined that the like-for-like determination was in error. The reviewer determined that an equivalency l evaluation should be performed and initiated AR A0452626 (Request date: February 6, 1998). Further, the SPlGOT team determined that the like-for-like determination was a violation of CF3.lD13 and was, therefore, also a quality problem.
Revision 01 of RPE M-8572 was issued on April 4,1998, correctly identifying the RPE ;
as an equivalency evaluation and including the necessary evaluations. The inspectors '
had no further technical concems regarding this RPE.
However, the circumstances surrounding the RPE raised an additional QA concern.
The initial determination that the chromium repair was a like-for-like replacement raised a question as to whether Diablo Canyon staff consistently recognize what constitutes a like-for-like replacement. The definition contained in CF3.lD13 stated that a like-for-like replacement is an " item with the same part, make and model number, that exhibits the same technical and physical characteristics as the original." Although this definition seemed clear and restrictive, RPE M-8572 indicated a lack of attention to detail on the part of the RPE preparer, reviewer and supervisor, if wide spread, this problem would be considered a program implementation weakness. However, as mentioned above, the licensee's corrective actions in response to the self assessment should adequately address the concems identified, b.6 RPE-E00025 Fuses. Fuse Blocks. and Fuse Holders The inspectors reviewed RPE-E00025 for upgrading commercial grade fuses, fuse blocks, and fuse holders, from various manufacturers, for safety-related service. The inspectors noted that the RPE described the dedication activities to be performed. The 17 RPE identified the components to be dedicated (fuses, fuse blocks, and fuse holders),
the parent components for the fuses, fuse blocks, and fuse holders (safety-related electrical panels and equipment), and the item model type and part number. The RPE identified the safety-related function of the parent components; the safety-related function of the fuses, fuse blocks, and fuse holders; the credible modes of failure such i as premature opening of a fuse, failure of the fuse block base material or poor contact f between the fuse holder and an installed fuse; and the effects of failure of a fuse, fuse i block, or fuse holder on the system where installed (these items were all identifiec' l separately for the fuses, fuse blocks, and fuse holders). l The RPE also identified the critical characteristics of the fuses, fuse blocks, and fuse holders which included dimensions, current carrying capacity, and clearing time (fuses);
and dimensions, insulation resistance, and fuse fit (fuse blocks and fuse holders). Fuse dimensions were verified by measurement, current carrying capacity was verified by testing, and current clearing time was measured by testing. The fuse critical characteristics were verified on a sample of the items received with the sample size ;
specified for each critical characteristic. Fuse block and fuse holder dimensions were verified by measurement, insulation resistance verified by testing, and a " pull-out test" verified that the appropriate sized fuse could not be easily pulled out of the fuse holder.
The fuse block and fuse holder critical characteristics were verified on samples of the items received with the sample size specified for each critical characteristic.
In addition, the inspectors reviewed the IPC for Part No. 94-2045, a 30 amp fuse block, dedicated in accordance with RPE-E00025. The IPC specified Standard Clause 7071 which provided instruction on the dedication activities to performed. The inspectors noted that dedication activities specified by Standard Clause 7071 were in agreement with RPE-E00025. The inspectors also reviewed Work Order No. C015667 which provided instructions on a fuse holder replacement. The inspectors concluded that the RPE, IPC, and Work Order reviewed had been adequately prepared in accordance with the applicable Diablo Canyon procedures and guidelines.
b.7 RPE E-1013 Transistors (Generic)
The inspectors reviewed RPE E-1013 for upgrading commercial grade transistors, from various manufacturers, for safe:y-related service. The inspectors noted that the RPE described the dedication activities to be performed. The RPE identified the component to be dedicated (transistors), the parent components for the transistors (various applications), and the item model type and part number. The RPE identified the safety-related function of the parent component, the safety-related function of the transistor, the credible modes of failure of a transistor such as short circuit or open circuit, and the effects of the failure of a transistor on the system where installed.
The RPE also identified the critical characteristics of the transistors which included configuration, verification to the transistor characteristics cur 3, forward-current transfer ratio (HFE), saturation voitage and resistance (VCE), and breakdown voltage emitter to base (VEB). Configuration was verified by inspection. Verification of performance in accordance with the transistor characteristics curve, HFE, VCE and VEB were verified 18
. ... _ --.- _ - - _ _ . - . - - - - . ~ . . - - - . . - - - . . .
by. testing. The transistors' critical characteristics were verified on a sample of the items received with the sample size specified for each critical characteristic. In addition, the inspectors reviewed the IPC for Part No. 93-3052, an NPN transistor, dedicated in <
accordance with RPE E-1013. The IPC specified Standard Clause 9941 which provided instruction on the dedication activities to performed. The inspectors noted that dedica-3 tion activities specified by Standard Clause 9941 were in agreement with RPE E-1013.
The inspectors also reviewed Work Order No. C0152725 which provided instructions on a transistor replacement in a boron concentration measurement device. The inspectors concluded that the RPE, IPC, and Work Order reviewed had been adequately prepared in accordance with the applicable Diablo Canyon procedures and guidelines.
b.8 RPE-E00039 Thermal Overload Relavs and Overlos.d Heaters The inspectors reviewed RPE-E00039 for upgrading thermal overload relays and overload heaters, from various manufacturers, for safety-related service. The inspectors noted that the RPE described the dedication activities to be performed. The RPE identified the components (overload relays and overload heaters), the parent components for the thermal overload relays and overload heaters (various applications),
and the item model type and part number. The RPE identified the safety-related
- function of the parent component, the safety-related function of the thermal overload relays and overload heaters, the credible modes of failure such as open circuit, high resistance due to material failure, and shori circuit, and the effects of failure of the thermal overload relays and overload heaters on the system where installed.
The RPE also identified the critical characteristics of the thermal overload relays and overload heaters which included configuration, dimensions, insulation resistance, contact current rating, contact configuration, and operability (thermal overload relays) and the applicable overcurrent trip time range and continuity (overload heaters).
Configuration, dimensions, contact current rating, contact configuration, and the applicable overcurrent trip time range were verified by inspection, and insulation resistance, operability and continuity were verified by testing. The thermal overload relays and overload heaters critical characteristics were verified on a sample of the items received with the sample size specified for each critical characteristic. In addition, the inspectors reviewed the IPC for Part No. 93-1335, an overload heater, dedicated in accordance with RPE-E00039. The IPC specified Standard Clause 7057 which provided instruction on the dedication activities to performed. The inspectors noted that dedication activities specified by Standard Clause 7057 were in agreement with RPE-E00039. The inspectors also reviewed Work Order No. C0156917 which provided instructions on an overload heater replacement in an auxiliary feedwater pump motor controller. The inspectors concluded that the RPE, IPC, and Work Order reviewed had been adequately prepared in accordance with the applicable Diablo Canyon procedures and guidelines.
b.9 RPE P-1194: Kunkle Relief Valve RPE P-1194 documented the dedication of a commercial-grade 3/4-inch, Kunkle relief-valve to be installed in the exhaust end of the auxiliary feed water (AFW) pump turbine casing (DCPP 10 Nos. MS-1(2)-RV-57), a Design Class 1 (safety-related) application. The licensee developed this RPE because 19
~ . _ _ _ _ _ _ . _ _ _ _ _
l the valve manufacturer, Kunkle Valve Co., was no longer on the Qualified Supplier List (QSL) and did not have an "N-Stamp" under the Boiler and Pressure Vessel Code of the American Society of Mechanical Engineers (ASME Code). Revision 01 of RPE P-1194, dated January 10,1996, stated that the functions of RV-57 were overpressure protection for the exhaust end of the AFW pump turbine casing and system pressure i retention. The postulated failure modes given were fracture of the spring or valve body. l The RPE indicated that the postulated failure modes could prevent the parent component from performing its safety function. Revision 00 of the RPE had specified that the valve body material (specified in both revisions as " typical stainless steel") was to be verified by X-ray fluorescence test, but the verification method was deleted in Revision 01. See the previous comments on not specifying verification methods.
To verify performance of the overpressure protection function, both revisions of the RPE specified performance of the DCPP maintenance procedure for safety relief valve set l point calibrr.?on (25.9 psig for RV-57), but only Revision 1 specified verification of the function of . trsure retention (by a 1-minute seat leakage test at 25 psig). However - i having' stated that fracture of the spring was a credible failure mode and that it could I adversely affect the safety function of the turbine, the RPE did not identify spring ,
material as a critical characteristic. Although a fractured spring would be detected by the performance tests upon acceptance, not specifying the material of the spring would not provide reasonable assurance of acceptable corrosion and erosion resistance. Also, ,
the RPE did not discuss the effects of the failure modes, e.g., wouH a failed open RV-57 (due to a fractured spring) impact turbine operation or could the esciping steam potentially damage adjacent safety-related equipment?
However, upon further review in response to the inspectors concem9, the licensee determined that contrary to the RPE, the pressure retention functic,n was not, in this '
case, a relevant function for the system since a complete loss of 9V-57 was considered to have a negligible effect on AFW turbine operation. In additior , although this was not clear from the RPE, the inspectors leamed that the particular model of Kunkle relief valve used for RV 57 was a sentinel valve whose main purposa was to serve as an audible alarm in case the exhaust of the turbine should becomo blocked. Therefore the i RPE had erroneously ider'tified pressure retention as a safety function of the valve.
With regard to the effect of escaping steam (from a fractured valve body as opposed to normal valve discharge-same as failed open valve) on adjacent equipment, the licensee had not evaluated that scenario, but committed to doing so and to include a more ,
complete discussion of the various failure effects in the RPE. ;
- c. Conclusions The inspectors found that CF3.lD13 was comprehensive and, if properly implemented, I should assure that dedicated commercial grade items will perform their intended safety i function. However, the inspectors considered not specifying standards, verification !
methods and acceptance criteria in the RPE to be a weakness. The inspectors also considered misuses of the RPE minor changes a weakness. In addition, based upon an l RPE review, the inspectors were concemed as to whether Diablo Cenyon procurement l engineers consistently recognize what constitutes a like-for-like replacement versus i 20 1
l what needs en equival:ncy evaluation. After discussions with the leader of the licensee's procurement and dedication self assessment team, the inspectors were satisfied that the licensee's planned corrective actions in response to the self assessment should adequately address the concems identified. With respect to the Kunkle relief valve dedication, the inspectors concluded that RPE P-1194 had failed to identify the appropriate safety related functions of the relief valve and did not adequately address spring material. Taken in the aggregate, the deficiencies identified above could be considered violations of the Criteria Ill, V, and Vil of 10 CFR Part 50, Appendix B.
However, in view of their relatively low safety significance, the deficiencies are being treated as a minor violations consistent with Section IV of NUREG-1600, Revision 1, May 1998, General Statement of Policy and Procedures for NRC Enforcement Actions, and no notice of violation will be issued.
E2 Review of Diablo Canvon Reolacement Parts Evaluation Self Assessment
- a. Insoection Scooe The inspectors reviewed the recently issued Diablo Canyon independent Safety Engineering Group (ISEG) Report No. 980360053. The ISEG report documented an assessment performed to review the RPEs issued since January 1995 and the process used to upgrade items initially purchased as non-safety related for safety-related use.
The inspectors also reviewed the ARs which had been generated as a result of the it.;EG RPE assessment. The inspectors reviewed the corrective actions which had been documented in the ARs and the weaknesses identified.
- b. Observations and Findinas The assessment team's report indicated that it had reviewed approximately 50 RPEs covering a variety of equipment, components, and material including Velan valves, Crosby relief valves, steam generator primary manway gaskets, Gould pump replacement parts, emergency diesel generator saturable transformer, Sorrento power supplies, and 4-kV circuit breaker auxiliary switches. The results of the assessment identified seven quality problems which had been documented in Action Requests (ARs) and had identified nine weaknesses. The assessment team had concluded that "the RPE process is in a state of decline" and that"the rate of decline, and the reason for it are unknown at this time, but merits follow up in the future."
The ISEG RPE ARs documented problems such as an inadequate technical evaluation for two stock codes, misclassification of identical versus equivalent, inadequate ASME Code reconciliation during procurement, misclassification of an RPE, an RPE evaluated as generic as opposed to specific, an inadequate equivalency evaluation, and lack of a design change evaluation. The inspectors noted that all of the ARs had documented corrective actions, most of which had been completed.
The inspectors discussed the assessment with both Nuclear Quality Services (NQS) and procurement personnel to obtain further insight on the significance of the conclusion drawn by the ISEG assessment team concerning the state of the RPE process. NOS 21 e
indicated that the assessment was a component of a continuous audit cycle and was j therefore most accurately viewed in that broader context. The audits were typically performed on a quarterly basis and any additional information received (including corrective actions) after the assessment was distributed, but prior to the issue of the Quarterly Audit Report, would be considered. The assessment would receive further ;
review when discussed in the Quarterly Audit exit meeting and the re' '. would be i incorporated into the Quarterly Audit Report. NOS indicated that the " state of decline" was relative to the last assessment performed in the RPE area and did not indicate a i breakdown of the RPE process. In addition, NQS stated that the assessment results l indicated the possible need for increased monitoring an 3 oversight of the process and i the need for another assessment of the area after an appropriate period of time, but that :
the results of the assessment were not significant enough to require the generation of a nonconforming condition report (NCR) to address the issue. Procurement Services indicated that they had reviewed the assessment and were developing an approach to
) ,
j address the general concems documented in the ARs and in the identified weaknesses. )
i l
- c. Conclusions I l On the basis of the inspectors' own review of weaknesses identified by the ISEG, the l
inspectors concluded that the ISEG had adequately documented areas where attention
- couH be focused by personnel using the assessment to provide input or develop actions to further improve the RPE process. The inspectors further concluded with respect to ;
the self assessment that the results of the assessment and the general conclusions I
given by the ISEG assessment team should be viewed in the broader context of the
> Quarterly Audit process and that the summary of the current state was relative to earlier assessments and not necessarily indicative of an absolute status.
The inspectors concluded, based on the review of RPEs, IPC, Standard Clauses, and Work Orders, performed during the inspection, that in general the RF'E program was adequate, but suffered from some implementation problems. The inspectors were concerned that sufficient resourcss may not always be devoted to staffing and training in the materials services and procurement design engineering areas to ensure the consistent quality and suitability of purchased material and services associated with safety-related equipment applications and to avoid further problems with proper review and authorization of document changes, material equivalency evaluations and problems such as those cited in the procurement self assessment.
E3 Procurement Enaineerina E3.1 Warehouse
- a. Insoection Scooe As a part of the evaluation of the effectiveness of the Diablo Canyon QA Commercial Grade Dedication Program, the inspectors reviewed the material control processes for handling and storage of items in the main warehouse. The inspectors also performed a walkdown of the warehouse.
22
b.- Observations and Findinos The warehouse was divided into two main areas designated Level B and Level A in accordance with requirements of ANSI N45.2.2. The storage level was specified in the Inventory Parts Catalog (IPC) by the plant procurement group personnel in accordance with plant .ucedure CF5.lD2," Control of Materialin Storage." Plant Material Services personnel store the materials in the designated warehouse locations. The inspectors ,
noted that the Level"A" area was a segregated zone with controlled access and
{
appropriate means for the control of temperature and humidity, although it was mainly occupied by items permitted to be Level"B" storage.
All items stored in the warehouse were uniquely identified in order to permit traceability with procurement documentation. The identification of items consisted of labels directly affixed to the item or on bags holding small spare parts. The identification code consisted of a six-digit number, called a stock code, which correlated to the PG&E purchase order for the item. This stock code was also used in the IPC computer data base allowing warehouse personnel to find the physicallocation of each individualitem.
During the warehouse walkdown, the inspectors randomly selected samples of stocked items to verify that they had been stored in the appropriate location. One of these items (caps 3/8" SS Hex Tubes - valve spare parts), physically located in bin AB-58-07-5 was labeled with stock code No. 93.9919, but with warehouse location identification No.
AB-58-06-H. The inspectors also observed that bin AB-58-06-6 was actually occupied by paramedic medical scissors with stock code number 76.6355.
The licensee explained this apparent error by stating that the designation of all physical locations for items in the warehouse was controlled in the IPC data base via the stock i codes. The inspectors verified that the IPC data base actually listed stoex code No. !
93.9919 as having warehouse location identification No. AB-58-07-5 as its physical I location. Upon further scrutiny, the inspectors leamed that material under stock No.
93.9919 had been relocated to bin AB-58-07-5 but that, at least one of the individual items remained identified with the previous bin designation (i.e., AB-58-06-6).
Apparently, warehouse personnel had relied on the bin location identified for the item in the IPC data base and had not updated the package's original bin location marking. The !
inspectors' review of procedure CFS.lD2 " Control of material storage, and the warehouse notebook instructions, confirmed that this was an acceptable practice.
- c. Conclusions The inspectors concluded that the facilities and associated controls in the main warehouse were adequate. However, absent specific instructions requiring that warehouse personnel verify the physical location / relocation based only on the IPC data, there is the possib!!ity that the printed location (in the label affixed to the items) could become a source of errors in the storage of individualitems. The licensee agreed to evaluate this possibility and take appropriate action if warranted.
23 Management Meetings A formal entrance meeting was held with Diablo Canyon management and cognizant staff on Tuesday, April 7,1998. The inspectors explained the scope and objectives of the inspection, and discussed report issuance and how any enforcement or followup items would be handled. I Logistics and communications arrangements during the during the inspection were agreed upon )
including special arrangements for the concurrent visit of, and inspection observation by, two ,
representativeu of the nuclear regulatory agency of the Slovak Republic. !
I The inspectors conducted the exit meeting with Diablo Canyon management and cognizant l staff on Thursday, April 9,1998. The inspectors briefed the licensee on the major inspection l findings and their possible and probable disposition, subject to NRC management review.
i PARTIAL LIST OF PERSONS CONTACTED !
.PGGF (Ucensee for Diablo Canyon) !
R.P. Pocars, Vice President, Diablo Canyon Power Plant ,
i Stan Ketelsten, Supervisor, NSAL-Regulatory Services Dwight Christensen, Engineer, NSAL-Regulatory Services Charlie Nichols, Director, Procurement Services Michael Jacobson, Sr. Engineer, Nuclear Quality Services Aaron Silva, Intem, NSAL-RS Dave Taggart, Director, Nuclear Quality Services Tom Fetterman, Director, l&C/ Electrical Design Engineering David Oatley, Manager, Maintenance Services Teny Grebel, Director, Regulatory Services Jim Love, PQA, Procurement Services Paul Milne, Engineer, PPE Jeff Nubbe, Procurement Specialist, Materials Services Bill Crocket, Manager, Nuclear Quality Services ;
John J. Griffin, Supervisor, Nuclear Quality Services l Stefan Bednarz, System Engineer, Diesel Emergency Generators ,
t Usama Elbakahish, Engineer, Procurement Design Engineering Mike Williamson, Meg Supervisor, Materials Services Jay Young, Director, Nuclear Quality Services USNRC Ken Perkins, Acting Director, Walnut Creek Field Office David Proulx, Senior Resident inspector, Diablo Canyon Power Plant Brad Olson, Region IV, Walnut Creek Field Office, Project Engineer Don Allen, Region IV, Diablo Canyon Resident inspector 24
. .~ . .-.
pa a849 4
g t UNITED STATES g j NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20066 4001 e
s ****/ July 13,1998 l
Mr. Henry G. Grilk, Vice President Engineering Limitorque Corporation 5114 Woodall Road P.O. Box 11318 '
Lynchburg, VA 24506
SUBJECT:
NRC INSPECTION REPORT 99900100/98-01 AND NOTICE OF NONCONFORMANCE
Dear Mr. Griik:
This letter addresses the inspection of your facility at Lynchburg, Virginia, conducted by Bill Rogers, with assistance from Thomas Scarbrough, of this office on April 21 through 23, 1998, and the discussions of their findings with Patrick McO.uillan and Bill Miluszusky of your staff at the conclusion of the inspection.
Areas examined during the inspection and our findings are discussed in the enclosed report.
This inspection consisted of an examination of procedures and representative records, interviews with personnel, and observations by the team.
During this inspection, it was found that the implementation of your Quality Assurance (OA) program did not meet certain NRC requirements. Limitorque Corporation (Limitorque) had not taken adequate corrective action conceming the adequacy of the relaxation of the sizing equation criteria for ac-powered motor actuators, which could potentially allow licensees to use ac-powered motor actuators which had not been adequately sized. In addition, Limitorque had not performed corrective ' actions, which had been documented as required to be performed, for your 10 CFR Part 21 evaluations.
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.
The response 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, Pub. L. No.96-511.
Mr. Grilk July 13,1998 in accordance with 10 CFR 2.790 of the NRC's " Rules of Practice," a copy of this letter, its enclosures, and your response will be placed in the NRC Public Document Room (PDR). To the extent possible, your response should not include any personal privacy, proprietary, or safeguards information so that it can be placed in the PDR without redaction. However, if you find it necessary to include such information, you should clearly indicate the specific information that you desire not to be placed in the PDR, and provide the legal basis to support your request for withholding the information from the public.
Sincerely, C-x ,
Suzannt !ack, Chief Quality Assurance, Vendor inspection and Maintenance Branch Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation Docket No. 99900100
Enclosures:
- 1. Notice of Nonconformance
- 2. Inspection Report 99900100/98-01 l
1 1
, NOTICE OF NONCONFORMANCE i
1
- Limitorque Corporation Docket No. 99900100 Lynchburg, Virginia j Based on the results of an inspection conducted un April 21 through 23,1998, it appears that j certain of your activities were not conducted in accordance with NRC requirements.
A. Criterion XVI," Corrective Action," of 10 CFR Part 50, Appendix B," Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants,"(Appendix B), requires in part that " measures shall be established to assure that conditions adverse to quality, such
, as failures, malfunctions, deficiencies, deviations, defective material and equipment, and
, nonconformances are promptly identified and corrected," "in the case of significant conditions adverse to quality, the measures shall assure that the cause of the condition is determined and corrective action taken to preclude repetition," and "the identification of the ;
significant condition adverse to quality, the cause of the condition, and the corrective action I i'
taken shall be documented and reported to the appropriate levels of management."
4 Limitorque Corporation's (Limitorque) Quality Assurance Procedure QAP 14.1, " Corrective Action Procedure," Revision 3, dated February 7,1997, states that "upon identification of a
!' condition or trend adverse to quality, Quality Assurance will initiate the appropriate C/A ,
[ corrective action] document and provide recommended corrective actions," and "the recipients of the C/A document shall determine the cause, develop and implement I appropriate corrective action that corrects the deficiency and the steps to prevent i recurrence."
Contrary to the above, the NRC found that Limitorque had obtained information by ;
September of 1997 which indicated tilat the use of relaxed criteria for the sizing equation for ac-powered motor actuators could potentially allow licensees to use ac-powered motor actuators which had not been adequately sized and that Limitorque had not taken adequate corrective action regarding licensee use of the relaxed criteria in the sizing equation guidance by the time of the NRC inspection in April 1998.
(Nonconformance 99900100/98-01-01) i B. Also contrary to the requirements of Cnterion XVI, " Corrective Action," of 10 CFR Part 50, Appendix B, the NRC found that the Limitorque 10 CFR Part 21 Evaluations No. 45, dated September 28,1993, and No. 47, dated December 9,1993, documented certain corrective -
actions to be performed and that Limitorque had not completed these corrective actions.
These evaluations concemed notifying customers of the appropriate hardware to be used with SMB-000 actuators and correcting diserapancies in previously supplied motor .
revolutions per minute (RPM) curves.
(Nonconformance 99900100/98-01-02)
Enclosure 1 3
f w - . . . . . , , - , . . . , , , _ r , , ,- , -
)
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, Quality Assurance, Vendor Inspection and Maintenance Branch, Division of Reactor Controls and Human Factors, 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 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 1
Dated at Rockville, Maryland this 13.tbday of July 1998 2
U.S. NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGUI.ATION Report No: 99900100/98-01 Organization: Limitorque Corporation
Contact:
Bill Miluszusky, Quality Assurance Engineer (804)528-4400 l Nuclear Activity: Motor-operated valve actuators, replacement parts, and services Dates: April 21-23,1998 Inspector: Bill Rogers, Quality Assurance, Vendor inspection and Maintenance Branch Participating Personnel: Thomas Scarbrough, Mechanical Engineering Branch Approved by: Richard Correia, Chief Reliability and Maintenance Section .
Quality Assurance, Vendor inspection and Maintenance Branch Division of Reactor Controls and Human Factors Enclosure 2 i
1 INSPECTION
SUMMARY
On April 2123,1998, the U.S. Nuclear Regulatory Commission (NRC) performed an l inspection at the Limitorque Corporation (Limitorque) facility in Lynchburg, Virginia, i l The inspection was conducted to review selected portions of the Limitorque quality l l assurance (QA) program, and its implementation, which included review of applicable l programs and procedures used in the current supply of motor-operated valve (MOV) actuators, replacement parts, and services to NRC licensees. The team reviewed activities related to motor actuator sizing criteria, including actions taken in response to related industry and NRC information,10 CFR Part 21 activities, and Limitorque's l conformance to customers' procurement requirements and compliance with NRC l regulations.
The inspection bases were:
. 10 CFR Part 50, Appendix B, " Quality Assurance Criteria for Nuclear Power ;
Plants and Fuel Reprocessing Plants." i
. 10 CFR Part 21, " Reporting of Defects and Noncompliance."
The team identified two nonconformances, related to corrective actions, which are discussed in Sections 3.1 and 3.2 of this report.
2 STATUS OF PREVIOUS INSPECTION FINDINGS Unresolved item (99900100/93-01-03) (Closed)
Unresolved item 99900100/93-01-03 concemed Limitorque's review of a situation, which was identified by a licensee, regarding SMB-000 actuators with incorrectly ,
installed SAE Grade 1 and 2 hardware. During this inspection, it was determined that the licensee had adequately addressed the situation and that Limitorque had evaluated the issue. These actions adequately addressed the Unresolved item; however, Limitorque had not completed the corrective actions which had been determined to be required and documented during Limitorque's evaluation process.
The failure to complete the required, documented corrective actions was identified as Nonconformance 99900100/98-01-02 and is further discussed in Section 3.2 of this Inspection Report.
Unresolved item (99900100/93-01-04) (Closed) ]
l Unresolved item 99900100/93-01-04 concerned the sampling frequency used by Limitorque for accepting bar stock and tubing. Limitorque had indicated during the 1993 NRC inspection that the applicable Inspection Plan would be revised to include !
adequate sampling plans. During this inspection, the team reviewed Inspection Plan l' No. 23, "Bar Stock and Tubing," dated May 24,1993, which included appropriate sampling requirements and adequately addressed the Unresolved item.
' 2 i
1
l Unresolved item (99900100/93-01-05) (Closed)
Unresolved item 99900100/93-01-05 concerned informing licensees of errors in the Limitorque Efficiency Table for SMB-3 actuators. During this inspection, the team determined that Limitorque had provided the correct information to the licensees by I
issuing Limitorque Technical Update 93-01, "SMB-3 Gear Efficiencies," dated June 8, 1993, which adequately addressed the Unresolved item.
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 Limitoraue Activities Related to Motor Actuator Sizina. Sorina Pack Curves. Structural Ratinas. and Bolt Toraues
- a. Insoection scom .
The team reviewed procedures and documentation and interviewed personnel to assess Limitorque's activities related to motor actuator sizing, spring pack curves, l structural ratings, and bolt torques, and review of NRC and industry information.
- b. Observations and Findinas b.1 AC-Powered Motor Actuator Sizing Criteria !
As discussed in NRC Inspection Report 99900100/93-01 (IR 93-01), Limitorque had earlier established a standard practice for sizing motor actuators as described in its "SEL" documents (dated 1977 and earlier). According to the SEL documents, Limitorque predicted the torque output of actuators used in motor-operated valves (MOVs) as follows:
Actuator predicted output torque = MT x Eff x AF x OAR x DVF (referred to as the " Sizing Equation") where MT = motor torque Eff = actuator gear efficiency AF = application factor OAR = overall actuator gear ratio DVF = degraded voltage factor in the 1977-era SEL documents and more recently in its Technical Update 92-02, Limitorque specified for sizing ac-powered motor actuators:
(1) Motor torque to be the nominal motor starting torque, (2) Actuator gear efficiency to be the " pullout
- gear efficiency from a table provided by Limitorque for the specific actuator size and gear ratio, (3) Application factor to be 0.9 or lower depending on the particular application, (4) Overall actuator gear ratio to be based on the particular actuator, and (5) Degraded voltage factor to be the square of the ratio of actual to rated motor voltage.
I s 3 l
The NRC team had reviewed Limitorque's basis for the calculations during the 1993 NRC inspection and determined that, although these assumptions were based primarily on engineering judgement and that the individual parameters in the ac-powered motor actuator sizing criteria had not been independently justified, the confidence in the prediction of output torque had been obtained from the success of the Limitorque criteria in sizing motor actuators for many years.
In the early 1990s, dynamic tests of MOVs by NRC licensees revealed that valve 4 vendors had significantly underpredicted the torque and thrust (related to torque by ,
the valve stem factor) required to operate many safety-related valves under their ;
design-basis conditions. This determination led licensees to request that Limitorque )
relax certain criteria used in the sizing equation for ac-powered motor actuators, which would result in higher actuator predicted output torque. For example, in a letter j to Cleveland Electric illuminating Company dated September 17,1992, Limitorque stated that, for a particular MOV and its application, the licensee could assume 110% !
of the rated motor starting torque and a "run" efficiency (up to 20 to 30% greater than ,
the " pullout" efficiency). Similarly, a Limitorque letter dated September 24,1993, to j Boston Edison Company allowed that licensee to assume "run" efficiency for another )
MOV and its specific application. Further, Limitorque Technical Update 93-03 (dated )
September 1993) allowed licensees to assume an application factor of 1.0 (rather than 0.9) where the motor voltage was less than 90% of the motor's rated voltage.
As a result of such relaxations of Limitorque's motor actuator sizing criteria, most licensees revised their MOV calculations to incorporate the relaxed criteria for all ac-powered motor actuators.
In 1994, concerns were raised regarding the adequacy of Limitorque's relaxed criteria, used in determining the actuator's predicted output torque, to reliably predict the actual output torque of ac-powered motor actuators. in a paper presented by Texas Utilities Electric at an NRC/ASME Symposium on Valve and Pump Testing in 1 July 1994 (see NUREG/CP-0137), it generally appeared that the actuator sizing method would only be adequate if the original sizing criteria were used with " pullout" efficiency and a 0.9 application factor assumed. At another NRC/ASME symposium in July 1996 (see NUREG/CP-0152), Commonwealth Edison Company reported that its review of available test data indicated that use of " pullout" efficiency combined with a 0.9 application factor was needed to establish a conservative lower bound for actual i actuator gear efficiencies. Also, at the July 1996 symposium, the Idaho National Engineering and Environmental Laboratory (INEEL) reported that preliminary results of NRC-sponsored tests similarly suggested the need to use " pullout" efficiency combined with a 0.9 application factor to reliably predict the actuator gear efficiency:
On August 21,1996, the N AC !ssued Information Notice (IN) 96-48, " Motor-Operated Valve Performance issues," which alerted licensees to these results of NRC-sponsored and industry tests of motor actuator torque output. In July 1997, the NRC issued NUREG/CR-6478, " Motor Operated Valve (MOV) Actuator Motor and Gearbox Testing," which described in detail the results of the NRC-sponsored tests of motor actuator performance. The testing program revealed that motor output was typically greater than nominal starting torque but that actual actuator efficiency was normally less than *run" efficiency (and in some cases even less than " pullout" efficiency), and 4
motor output under degraded voltage conditions was less than predicted by the
. square of the voltage ratio.
During this inspection, the team determined that Limitorque had received the information conceming ac-powered motor actuator performance, as discussed in NUREG/CP-0137, NUREG/CP-0152, IN 96-48, and NUREG/CR-6478, by approximately September of 1997. At an MOV Users Group meeting held in December of 1997, Limitorque had stated to NRC licensees that it was considering retracting its previous relaxation of the sizing criteria and that a documented notification would be issued in early 1998. However, at the time of this inspection, Limitorque had not yet completed an assessment or finalized a position on the adequacy of the relaxation of the criteria used in the sizing equation for ac-powered motor actuators and had not provided any information to NRC licensees. The team determined that there was significant indication that relaxing the criteria used in determining the actuator predicted output torque could result in a predicted torque which would be greater than the actual torque. The team concluded that the use of 4 the relaxed criteria, which could potentially allow licensees to use motor actuators which had not been adequately sized, was a significant condition adverse to quality and that Limitorque had obtained information on the condition by September of 1997, but had not taken adequate corrective action or dot -"*nted the condition in a corrective action program by the time of the NRC in.pection in April of 1998. The team concluded that Limitorque had not taken corrective action for a significant condition adverse to quality which constituted a nonconformance to the requirements of Criterion XVI, " Corrective Action," of 10 CFR Part 50,' Appendix B.
(Nonconformance 99900100/98-01-01) )
Subsequent to this inspection, Limitorque prepared Technical Update 98-01 (dated May 15,1998) to provide updated guidance to determine the output torque capability l of an ac-powered Limitorque motor actuator. In Technical Update 98-01, Limitorque l specified that, in the sizing equation, licensees should use nominal motor starting torque, " pullout" efficiency, an application factor (typically 0.9), and an exponent of two as applied to the voltage ratio. Where voltage is greater than 90% of the motor rated voltage, Limitorque allowed licensees to eliminate the degraded voltage factor because of standard motor design characteristics. Limitorque also allowed licensees to use more optimistic assumptions than specified in Technical Update 98-01 where actual test data or certain engineering data are available. In light of available test data, Limitorque has alerted licensees to specific motor actuators that require special attention in determining cutput torque capability. In addition, Limitorque has indicated that the guidelines for considering the effects of elevated temperature on motor starting torque are provided in Limitorque Technical Update 93-03.
b.2 DC-Powered Motor Actuator Sizing Criteria The sizing equation developed by Limitorque for de-powered motor actuators in the SEL documents is similar to the equation for ac-powered motor actuators. The assumptions in the sizing equation for de-powered motor actuators include nominal motor starting torque, application factor, " pullout" efficiency, and a degraded voltage 5
l I
factor where an exponent of one is applied to the ratio of actual to rated motor voltages, (Limitorque has not relaxed its guidance for sizing de-powered motor i
actuators from ils SEL documents). However, initial de-powered motor test results l (INEEL tests) described in NUREGICR-6478 suggested that a tested motor did not l produce the expected torque and that the degraded voltage effect is more severe than predicted by the ratio of actual to rated motor voltage.
l At the time of this inspection, Limitorque had not documented any activities related to
- the information regarding the adequacy of its de-powered motor actuator sizing ,
criteria; however, Limitorque indicated that a review of de-powered motor performance test curves had been initiated and that Limitorque had plans to discuss these curves in more detail with the motor manufacturer. Limitorque also had assessed the general applicability of the INEEL tests based on the age of the tested i I
motor and its maintenance history. Limitorque stated that it would consider any additionalinformation developed from follow-up INEEL tests of de-powered motors. l During this inspection, Limitorque indicated that the questions raised regarding its de-powered motor actuator sizing criteria would be resolved through planned {
improvements to its corrective action program, b.3 Spring Pack Curves in IR 93-01, the NRC team noted that Limitorque was developing information to revise its curves that compared spring pack displacement to actuator output torque. During this inspection, Limitorque indicated that the work had not been completed because interest in this information had decreased as licensees typically use more sophisticated valve diagnostic equipment than in the past. Limitorque iridicated that ,
its effort would resume if a need for the information was revealed. The team did not identify any concerns in this area.
b.4 Actuator Structural Ratings Limitorque clarified its position, previously provided in a Limitorque letter to Babcock
& Wilcox, dated July 26,1990, regarding thrust and torque delivered at the MOV torque switch setting in relation to actuator structural ratings. Limitorque indicated that it did not have a concem with thrust and torque output at the torque switch setting in terms of the actuator structural rating provided the allowable structurallimits are not exceeded for the entire valve stroke based on appropriate consideration of ;
torque switch repeatability and diagnostic equipment error. The allowable structural j limits may range from 110% to 162% of the original ratings depending on a licensee's l reliance on recent studies of structural capability, and the conditions and limitations j associated with those studies. The team did not identify any concems in this area.
b.5 Actuator-To-Yoke Bnit Torque l l
In Maintenance Update 89-1, Limitorque referred licensees to valve manufacturers for j guidance on the proper torque value for bolting connections betweets the actuator base and the valve yoke. In Technical Update 92-01, Limitorque provides a specific 1
l 6
-3 8- i l
._ ._m ._ _ _ _ _ _ _ _..-.. _ __._ _ .__ _ _ _ ._ _ _ __.
l I
torque pattem for actuator-to-yoke bolts where higher structural limits are allowed and also refers licensees to the valve manufacturers for additional guidance. During this )
inspection, Limitorque indicated that licensees are responsible for any differences I from the guidance. The team did not identify any concems in this area.
- c. Conclusion 4
The team concluded that Limitorque had obtained information by September of 1997 which indicated that the use of relaxed criteria for the sizing equation for ac-powered motor actuators could result in a predicted torque which would be greater than the actual torque and that the use of the relaxed criteria could potentially allow licensees to use ac-powered motor actuators which had not been adequately sized. Limitorque i had not taken adequate corrective action regarding licensee application of relaxed criteria used in the sizing equation by the time of the NRC inspection in April 1998, which was identified as a nonconformance. In addition, the team determined that
. Limitorque was aware of the INEEL test results which suggested that a tested i de-motor did not produce the expected torque and that the degraded voltage effect is more severe than predicted, but concluded that Limitorque was taking adequate actions to address this issue.
3.2 10 CFR Part 21 Proaram
- a. Scope The team reviewed procedures and documentation and interviewed personnel to determine whether Limitorque had established a program meeting the requirements of 10 CFR Part 21 (Part 21) and was adequately implementing the program.
- b. Observations and Findinas b.1 10 CFR Part 21 Procedure The team reviewed Quality Assurance Procedure (QAP) 13.2, " Reporting of Defects for Safety Related Equipment," Revision 11, dated February 7,1997, which established the requirements for the identification and evaluation of deviations and for the reporting of defects to the NRC.
Paragraphs 4.1,4.2, and 4.3 of QAP 13.2 discussed employee's responsibilities for identifying non-conforming conditions or deviations and further describes these conditions as " conditions of a recurring nature,""significant conditions adverse to quality," and " substantial safety hazards." The team determined that the terminology used in QAP 13.2 did not completely agree with Part 21 which defines the term deviation as "a departure from the technical requirements included in a procurement document" which is then evaluated to determine whether a defect or substantial safety hazard exists. In addition, the persons reviewing the various documents which the QAP indicated were used to document the non-conforming conditions - Variation Reports (Quality Control), Field Service Reports and Customer Reported Problems 7
l
-m .
l (Parts Service), Limitorque Corrective Action Requests and Audit Deficiency Notifications (Quality Assurance)- would not perform the evaluations of the deviations to determine whether a substantial safety hazard exists (this was accomplished by the Part 21 committee). Therefore, QAP 13.2 was inappropriate in describing the conditions to be reported by QC, Parts Service, and QA as substantial safety hazards !
(which would have required the evaluation of a deviation). I 1
I b.2 10 CFR Part 21 Program Implementation The team reviewed the Limitorque Part 21 files which documented ine evaluations performed during the period from 1993 to 1998. Limitorque had performed eleven i evaluations during this period which concemed issues such as cracks in the frame :
seam weld on motors, torque switch roll pin failures, a loose set screw on a bearing l locknut, failed clutch sleeves, motor seals fabricated from incorrect material, failed !
actuator component supplied as a replacement part, missing torque switch spacers, j and potential for motor pinion key failure due to excessive torque. ]
Part 21 Evaluation No. 45 concemed the review of a Washington Public Power i Supply (WPPS) Part 21 report to the NRC which discussed SMB-000 actuators with incorrectly installed SAE Grade 1 and 2 hardware. A Limitorque memorandum dated September 28,1993, which documented the Evaluation No. 45 Part 21 Meeting stated "this irdormation will be re-stated together with reference of the WPPS Part 21 report in a Limitorque Maintenance Update with a mailing no later that 10-29-93."
The Limitorque Evaluation, dated September 28,1993, had an attached Limitorque Maintenance Update, 89-1, which indicated where SAE Grade 5 and 8 hardware was l to be used for the SMB actuator and identified that the corrective action was to issue i an additional Maintenance Update to the nuclear industry. Limitorque had no record of the Maintenance Update being issued to NRC licensees and indicated that it had not been issued.
Part 21 Evaluation No. 47 concemed the discrepancies in fullload revolutions per minute (RPM) on performance curves for certain Reliance motors identified to .
l Limitorque in a Reliance Electric letter dated November 15,1993. A Limitorque memorandum dated December 9,1993, which documented the Evaluation No. 47 Part 21 Meeting, stated that "the speed changes were inconsequential and therefore should not be a problem," but also state j "however, the Limitorque Nuclear Support 1 Group will be submitting the revised Performance Curves to any nuclear utility that has previously purchased the original Performance Curves from Limitorque Corporation? Limitorque Evaluation No. 47, dated December 9,1993, indicated that the deviation had been evaluated by the Part 21 Committee and determined that it was not a defect and not reportable to the NRC. However, the evaluation's corrective action stated 'this data to be fumished to affected Limitorque Corporation customers."
Limitorque had no record of the revised Performance Curves being issued to NRC licensees and indicated that the revised Performance Curves had not been submitted to the utilities; however, Limitorque did restate that the speed changes were insignificant and would not affect actuator performance.
8
i The team determined that for Part 21 Evaluations No. 45 and No. 47, although the evaluations were adequate, Limitorque was unable to provide documentation which l showed the completion of the corrective actions specified to be accomplished by Limitorque in the Part 21 evaluation files and indicated that the actions had not been accomplished. Limitorque's incompletion of corrective actions documented to be performed in Part 21 Evaluations No. 45 and No. 47 was identified as a Nonconformance to Criterion XVI, " Corrective Actions," of 10 CFR Part 50, Appendix B. (Nonconformance 99900100/98-01-02)
c. Conclusion
l The team concluded the Limitorque had incorporated the portions of 10 CFR Part 21 as required by 10 CFR 21.21(a) into the Part 21 implementing procedure, QAP 13.2,
! and had. in general, developed a procedure which could effectively implement the requirements of the regulation. However, the team did identify a weakness in the procedure concerning the terms used to identify deviations. In addition, Limitorque l had not adequately implemented its 10 CFR Part 21 program by not the completing l corrective actions which had been identified during the Part 21 evaluation process l
which was identified as a nonconformance.
l i 4 PERSONS CONTACTED l
l Henry Grilk, Vice President of Engineering l Patrick McQuillan, Manager Nuclear /Special Projects 1 Bill Miluszusky, Quality Assurance Engineer 1
l I
l 9
l l
l
ITEMS OPENED, CLOSED, AND DISCUSSED ltem Number ly.pg Descriotions Ooened !
99900100/98-01-01 NON No corrective action taken conceming the adequacy of the relaxation of the sizing equation criteria for ac-powered motor actuators 99900100/98-01-02 NON Corrective actions identified during the Part 21 process not completed GlQaed 99900100/93-01-03 URI SMB-000 actuators with incorrect hardware 99900100/93-01-04 URI Sampling frequency for bar stock and tubing 99900100/93-01-05 URI Errors in the Efficiency Table for SMB-3 actuators l
l 10 4
_g f %,i UNITED STATES l s
j g
NUCLEAR RE2ULATORY CCMMISSION CASHIN! TON, D.C. soneo coo 1 i
j
'+,***** ,d July 8, 1998 Mr. Thomas H. Zarges President and CEO Engineering and Construction Group Morrison Knudsen Corporation (Docket 99901329) 1500 West 3rd Street Cleveland, Ohio 44113-1406 Mr. Martin D. Cepkauskas, President The Steam Generating Team (Docket 99901334) 1500 West 3rd Street Cleveland, Ohio 44113-1406
SUBJECT:
NRC INSPECTION REPORT 99901329/98-01 AND 99901334/98-01 (NOTICE OF VIOLATION AND NOTICE OF NONCONFORMANCE)
Dear Gentlemen:
On May 20-22,1998, the U.S. Nuclear Regulatory Commission (NRC) completed an inspection at your facility, Morrison Knudsen Corporation, Engineering and Construction Group, Power Division (MK) and the Steam Generating Team (SGT), Cleveland, Ohio. The enclosed report presents the results of that inspection.
The NRC inspectors reviewed the implementation of selected portions of the MK and the SGT,
, quality assurance program, and reviewed activities associated with its welding, maintenance, cnd steam generator replacement services to the nuclear industry. During this inspection, the NRC Inspectors reviewed MKAGT correspondence of its steam generator replacement activities at Point Beach and St. Lucie Nuclear Power Station conceming MK/SGT's control of weld procedure specifications and procedure qualification records. The NRC inspectors were concerned about the capability of the MK/SGT quality assurance program used to evaluate the American Society Mechanical Engineers (ASME) Boiler and Pressure Vessel Code (Code) weld procedures after it had completed its review as discussed in the enclosed inspection report. It was found that certain of your activities appeared to be in violation of NRC requirements.
Specifically, a review of two 10 CFR Part 21 evaluations and several potential Part 21 issues identified that both 10 CFR Part 21 evaluations were not conducted in accordance with $21.21 of Part 21, and one potential Part 21 review by SGT/MK staff was found to have been stopped before the generic Part 21 implications were assessed by the SGT/MK staff. The inspectors determined that the two Part 21 evaluations should have been transmitted to applicable licensees within five days of discovery in accordance with 921.21(b) because the MK/SGT staff did not have the capability to determine whether a substantial safety hazard, as defined in
$21.3 of Part 21, existed. The NRC inspectors also determined that MK/SGT had not adopted adequate procedures to effectively implement the provisions of 10 CFR Part 21. This matter is cited in the enclosed Notice of Violation (NOV), and the circumstances surrounding the NOV cre described in detail in the enclosed report. Please note that you are required to respond to this letter and should follow the instructions specified in the enclosed NOV when preparing your response. The NRC will use your response, in part, to determine whether further enforcement cetion is necessary to ensure compliance with regulatory requirements.
Messrs. T. Zarges tnd M. C:pkruskas July 8,1998 in addition, the NRC inspectors found that the implementation of your quality assurance program failed to meet certain NRC requirements imposed on you by your customers.
Specifically, the inspectors determined that compliance with 10 CFR Part 50, Appendix B,
" Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants," and the ASME Code that was contractually imposed on MK/SGT by its nuclear utility customers was not met in certain areas. These nonconformances are cited in the enclosed Notice of Nonconformance (NON), and the circumstances surrounding them are described in detail in the enclosed report. You are requested to respond to the nonconformances and should follow the instructions specified in the enclosed NON when preparing your response.
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's Public Document Room (PDR). '
Sincerely, Suzan C. Black, Chief Quality Assurance, Vendor inspection and Maintenance Branch Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation Docket No. 99901329 & 99901334
Enclosures:
- 1. Notice of Violation
- 2. Notice of Nonconformance
- 3. Inspection Report 99901329/98-01 & 99901334/98-01 cc: Mr. Richard W. Bonsall Executive Vice President Energy & Environmental Group '
Duke Engineering & Services Post Office Box 1004 Charlotte, North Carolina 26201-1004
NOTICE OF VIOLATION Morrison Knudsen Corporation, and Docket No.: 99901329
, The Steam Generating Team Docket No.: 99901334 Cleveland, Ohio
' During an NRC inspection conducted on May 20-22,1998, a violation of NRC requirements was identified. In accordance with the " General Statement of Policy and Procedure for NRC Enforcement Actions," NUREG-1600, the violation is listed below:
Section 21.21, " Notification of failure to comply or existence of a defect and its evaluation," of 10 CFR Part 21, requires, in part, that each individual, corporation, partnership, dedicating entity, or other entity subject to the Part 21 regulation adopt appropriate procedures to (1)
Evaluate deviations and failures to comply to identify defects and failures to comply associated with substantial safety hazards as soon as practicable, and, except as prnvided in paragraph
- (a)(2) of this section, in all cases within 60 days of discovery, in order to identify a reportable dLfect or failure to comply that could create a substantial safety hazard, were it to remain uncorrected, and (2) Ensure that if an evaluation of an identified deviation or failure to comply potentially associated with a substantial safety hazard cannot be completed within 60 days from discovery of the deviation or failure to comply, an interim report is prepared and submitted to the Commission through a director or responsible officer or designated person. Section 21.21(b) requires that if the supplier of basic components determines that it does not have the capability to perform the evaluation to determine if a defect exists, then the supplier must inform th3 purchasers or affected licensees within five working days of inis determination.
Contrary to the above, as of May 22,1998, Morrison Knudsen Corporation (MK) and The
, St:am Generating Team (SGT):
- 1) failed to perform an appropriate evaluation of two different welding program issues that
, .were evaluated and approved on March 18,1997 and September 26,1997 as MK/SGT was found to not have an adequate capability to determine whether a defect existed, and
- 2) failed to perform an adequate evaluation of a potential defect regarding undersized welds at St. Lucie NPS, which may also have been applicable at other MK/SGT project locations (e.g., D.C. Cook and Point Beach nuclear power stations), but were not reviewed by SGT or MK during its disposition of the matter. MK/SGT Quality Execution Procedure (QEP) 15.1 did not contain adequate requirements or guidance to ensure that generic implications were addressed in the MK/SGT Part 21 evaluation. MK/SGT staff did not assess whether the suspect welding inspectors performed welding a
inspections at other SGT/MK sites where safety-related welding and inspection had been performed. Violation 99901329/98-01-01 & 99901334/98-01-01 This is a Severity Level IV violation (Supplement Vil).
e d
w i -
Pursuant to tha provisions of 10 CFR 2.201, Morrison Knudsen Corporation and The Stearn generating team are hereby required to submit 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, Quality Assurance, Vendor Inspection, and Maintenance Branch, Division of Reactor Controls and Human Factors, 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 a Notice of Violation" and should include for each violation: (1) the reason for the violation, or, if contested, the basis for disputing the violation, (2) the 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. Your response may reference or include previous docketed correspondence, if the correspondence adequately addresses the required response. Where good cause is shown, consideration will be given to extending the response time.
1 I
l Dated at Rockville, Maryland f this 8th day of July 1998
)
i l
I l
NOTICE OF NONCONFORMANCE Morrison Knudsen Corporation 1 Docket No.: 99901329 The Steam Generating Team Docket No.: 99901334 Cleveland, Ohio Based on the results of an inspection conducted on May 20 through 22,1998, l' appears that .
certain ov the Morrison Knudsen Corporation (MK) Steam Generating Team (SGT) activities w:re not conducted in accordance with NRC requirements.
A.
10 CFR 50, Appendix B, Criterion IX, " Control of Special Processes," requires, in part,
)
that measures shall be established to assure that special processes, including welding, and heat treating, are controlled and accomplished by qualified personi z' using qualified procedures in accordance with applicable codes, and standards.
Section 9.0, " Control of Processes," of the SGT Quality Assurance Program, April 26, 1996, states that the Construction Engineering manager is responsible for ensuring development and implementation of procedures and instructions which, as a minimum:
control or verify quality of special processes, such as welding, heat treatment, in accordance with applicable Codes and standards.
l Contrary to the above, and to the contractually imposed requirements of the American ,
Society Mechanical Engineers (ASME) Boiler and Pressure Vessel Code (Code), in i November and December of 1996, (weld procedure specifications) WPS GT-SM/1.1-1PB Revision 4 and WPS GT-SM-BU/1.3-1PB Revision 1, used to fabricate nine steam generator replacement project main steam and feedwater welds at Point Beach, did not meet the ASME requirements of QW-409.1, concerning the limitation of heat input over that which was qualified. The weld procedures allowed welding to be performed with an increase in heat input over that which had been qualified. (Nonconformance i i
99901334/98-01-02) i B.
Criterion V, " Instructions, Procedures, and Drawings," of 10 CFR Part 50, Appendix B l requires that activities affecting quality be prescribed by documented instructions and procedures and shall be accomplished in accordance with those instructions and procedures.
Section 5.0, " instructions, Procedures, and Drawings," of the SGT Quality Assurance ;
Program, requires that activities that affect quality shall be described by and accomplished through implementation of documented procedures, instructions, or drawings. Procedures, instructions and drawings shall include quantitative and qualitative acceptance criteria for determining satisfactory work performance and quality I compliance.
Enclosure 2 ,
I 1
Paragraph 9.2," Weld Procedure Qualifications", of Section 9.0," Welding and Heat Treating," of the MK Ouality Assurance Manual (OAM), August 20,1996, required that it was the responsibility of the Group Welding Engineer [GWE] to direct preparation and qualification of Welding Procedure Specifications (WPS) under the provisions of ASME Section lli and Section IX. Paragraph 9.2 also requires that welding procedures j previously qualified by MK may be utilized provided they are reviewed and accepted by j l
the GWE prior to use in production. This review shall be documented by signature and date on a cover sheet attached to the MK welding procedure.
MK Ouality Assurance Instruction (OAI) 11.2," Control of Welding Procedure !
I Specifications", September 23,1996, requiros that the GWE may designate the WPS '
qualification process to a PWR [ Project Welding Representative] by letter, but the GWE still retains the responsibility for this process. Designation to the PWR for procedure j qualification shall not be re-delegated by the designee.
SGT Ouality Execution Procedure (OEP) 20.2," Welding Procedure Qualification," j October 25,1996, requires that the GWE (Corporate Level) is responsible for '
qualification and approval of new ASME Procedure Qualification Records (POR) that are developed for use at the site level prior to their use. The GWE is also responsible for suppling POR's to the site that have been previously qualified. The GWE may delegate this responsibility for qualification to the Project Welding Engineer".
MK/SGT Point Beach Project steam generator replacement, Main Steam and Main Feedwater weld procedure, WPS GT-SM/1.1-1 PB, November 28,1996. Revision 4, did not prescribe a post weld heat treatment on weldments with a thickness of 1.5 inches and below.
Contrary to the above, and the corstractually imposed requirements of the ASME Code, the following examples were identified where MK/SGT did not accomplish quality activities in accordance with approved procedures, (Nonconformance 99901334/98 03):
- 1) Thirteen weld procedures were approved by an individual who had not been directly delegated authority by the GWE as required by OAM 9.2 and QAl 11.2 requirements.
- 2) On December 9-10,1996, post weld heat treatments were applied to two Point Beach Main Steam welds with a material thickness of 1.35 inches.
C. Criterion XVil," Quality Assurance Records", of 10 CFR Part 50, Appendix B requires in part, that " Sufficient records shall be maintained to fumish evidence of activities affecting quality."
l Section 17.0," Quality Assurance Records," of the SGT Quality Assurance Program, requires that records and/or indexing systems shall provide sufficient information to permit identification between the record and the item (s) or activity (ies) to which it )
applies. l Contrary to the above, the following examples were identified where SGT did not document adequate information to furnish evidence of welder activities in the following areas:
For the St. Lucie NPS steam generator replacement project, of the 36 we!ders entered on an SOT weld data card it was found that only six welders actually performed welding activities for the specific weldment.
. For Point Beach tt3am generator replacement weldments, it was found that SGT Weld Data Cards WDC 3081-A-2 and WDC 3081-B-7 did not record the appropriate documentation for the weld procedure revision used to fabricate the weld. Consequently, when Revision 4 of the welding procedure was determined to be an unqualified procedure, it could not be determined if those particular welds required requalification. (Nonconformance 99901334/98-01-04)
Pl:ase 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, Quality Assurance, Vendor Inspection and Maintenance Brancil, Division of Reactor Controls and Human Factors, Office of Nuclear Reactor Regulation, within 30 days of the date of the letter tiansmitting 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 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 proventive measures were or will be completed.
Dat:;o at Rockville, Maryland this 8th day of July 1998
. . - - -- , -. . . . _ . . . _ ~ . . . . . - - .
U.S. NUCLEAR REGULATORY COMMISSION i
OFFICE OF NUCLEAR REACTOR REGULATION !
Report No: 99901329/98-01 1 99901334/98-01 Organization: Morrison Knudsen Corporation (MK) !
Engineering and Construction Group l Power Division !
1500 West 3rd Street Cleveland, Ohio 44113-1406 The Steam Generating Team (SGT) 1500 West 3rd Street i Cleveland, Ohio 44113-1406
Contact:
Michael F. Hendricks, Group Quality Director (216) 523-8103 Nuclear industry: MK, Power Division, offers the nuclear industry services in construction, retrofit, maintenance and decommissioning of power plants.
SGT, a limited liability company owned by MK and Duke Engineering &
Services offers the nuclear industry steam generator replacement services.
Inspection Dates: May 20-22,1998 -
Inspectors: Joseph J. Petrosino, HQMB/NRR.
Katherine GreenBates, DRS/ Region ill Jerome J. Blake, DRS/ Region II Approved by: Robert A. Gramrn, Chief Quality Assurance and Safety Assessment Section Quality Assurance, Vendor inspection and Maintenance Branch Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation Enclosure 3 l i
_ _ _ . -i
1 INSPECTION
SUMMARY
During this inspection, the NRC inspectors reviewed the implementation of selected portions of ths Morrison Knudsen Corporation (MK) and the Steam Generating Team (SGT), quality assurance (OA) program, and reviewed activities associated with its welding, maintenance, and steam generator replacement services to the nuclear industry.
The inspection bases were:
o Appendix B, " Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants," to Pad 50 of Title 10 of the Code of Federal Regulations (Appendix B) o 10 CFR Part 21, " Reporting of Defects and Noncompliance" During this inspection, a violation of NRC requirements was identified and is discussed in Section 3.1 of this report. Additionally, instances where MK/SGT failed to conform to NRC and American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code requirements contractually imposed upon them by NRC licensees were identified. These nonconformances are discussed herein.
2 STATUS OF PREVIOUS INSPECTION FINDINGS This was the first NRC inspection of the MK/SGT Cleveland, Ohio facility.
3 INSPECTION FINDING"a AND OTHER COMMENTS 3.1 10 CFR Part 21 Program
- a. Insoection Scoce Tha NRC inspectors reviewed records related to the identification and evaluation of deviations to assess the overall effectiveness of MK/SGT's program that had been estab'ished to comply with the requirements of 10 CFR Part 21.
- b. Observations and Findinas The NRC inspectors reviewed two MK "10 CFR Part 21 evaluation" packages and several project specific nonconformances that were dispositioned by SGT procedures that addressed 10 CFR Part 21 issue dispositions.
b.1 Part 21 Evaluations The Part 21. evaluations were documented on MK's Quality Assurance Instruction (OAI) form 1,1-1, " Determination Checklist for 10 CFR Part 21 Applicability," which is referenced in MK procedure OAl 1.1," Reporting of Defects and Noncompliance," May 18,1998. Section 1.2.2 of 2
QAl 1.1 stated that the form was applicable to tha evaluation and rcporting of defsets and noncompliance discovered at the Cleveland office, and that defects and noncompliance discovered at project sites shall be handled in accordance with project procedures.
~
The evaluation packages regarded: (1) discrepancies in welding procedure specifications and procedure qualification records (PQRs) at DC Cook and Point Beach initially evaluated on i January 23,1997, closed on March 18,1997, and (2) Point Beach Welding Procedure :
Specifications (WPS') and PORs that were qualified through testing performed by an ;
unapproved testing agency, initially evaluated on September 25,1997, closed on September 26,1997.
The QAl 1,1-1 form had four parts: A. Description of deviation or Noncompliance, B. Initial !
Evaluation, C. Final Evaluation, and D. 'Reportability Evaluation. The NRC inspectors noted ;
that the initial evaluation (Section B) requires a determination of whether a violation of the Atomic Energy Act of 1954 or any applicable rule or regulation relating to substantial safety }
hazards, or whether the activity or component d. mil contain defects which could create a substantial safety hazard. The choices (blocks that may be checked) were "yes,""no," or 'l
" unknown by MK." Both of the Part 21 evaluations performed by MK/SGT were checked as l i
unknown by MK. However,in the final evaluation section of the QAl 1.1-1 forms, similar questions are asked regarding whether, on the basis of an evaluation, the issue could create a substantial safety hazard and, therefore be a defect, and whether the deviation could contribute to the exceeding of a safety limit. The inspectors noted that the reviewers checked "no"in the finti evaluation boxes, even though they had previously checked " unknown"in Section D, that is, the initial evaluation section. [
In the subject evaluation packages MK/SGT staff identified the deviation of its services as
" basic components," as defined in $21.3 of Part 21, and also identified in the initial evaluation section that it was " unknown" to them whether the questionable activities had violated any rule, regulation or license of the NRC. It was also identified by the MK/SGT staff that it was unknown ,
to the evaluators whether " defects" existed which could cause a substantial safety hazard. ,
However, in the final evaluation section of form QAl 1.1-1 the evaluator determined that defects did not exist, and also concluded that the deviation could not contribute to the exceeding of a safety limit. The NRC Inspectors discussed the evaluation process with one of the evaluators and found that the MK/SGT was not familiar with the failure mechanisms or consequences to other components in the area of the steam generator weldments, and was not aware of all of ,
the interactions and consequences for a design easis event accident. The NRC Inspectors !
determined that MK/SGT did not have adequate capability to perform the required evaluation as defined in Part 21.
Therefore, the NRC Inspectors concluded that MK/SGT did not have adequate plant specific knowledge to determine whether the matter could have caused a substantial safety hazard or
< exceeded technical specification parameters; therefore, MK/SGT was required to evaluate the matter in accordance with 6 21.21 of Part 21, The issue contained in evaluation package 2, appeared to be an issue that could have been completely dispositioned in accordance with the MK/SGT QA program instead of the Part 21 program.
l 4
b.2 Project Soecific Nonconformances A review of several MK/SGT nonconformances identified that all had been appropriately dispositioned with one exception. One of the several potential Part 21 issues that were dispositioned in accordance with the MK/SGT QA program was determined to be a deviation, es defined in @21.3 of Part 21, and should have been processed in accordance with the vendor's Part 21 program, and is discussed in Section 3.2 below.
C. Conclusions The NRC inspectors conclude that the MK/SGT Part 21 program and the procedures that it has adopted to implement the provisions of 10 CFR Part 21 does not ensure effective implementation of the Part 21 provisions. Violation 99901329/98-01-01, and 99901334/98-01-01 was identified in this area.
3.2 Evaluation of St Lucie NPS Weldina insoection issue
- a. Insoection Scoce The NRC inspectors reviewed several vendor Nonconformance Reports (NCR's) to verify that MK/SGT staff had appropriately evaluated and dispositioned reported problems or deficiencies.
had taken adequate corrective actions to preclude recurrence, and had evaluated potential deviations in accordance with Part 21.
- b. , Observations and Findinas The NRC inspectors reviewed selected NCR's and identified one instance on the Florida Power and Light company, St. Lucie NPS Steam Generator Replacement Project where SGT did not p:rform an adequate evaluation of an identified deviation. A review of SGT Nonconformance Report (NCR) 254, issued October 28,1997, identified that two undersize fillet welds had been found by the licensee's QA inspectors and brought to SGT's attention. The welds had previously been inspected and accepted by a qualified SGT quality control (OC) weld inspector.
A second OC weld inspector was sent out to reinspect a sample of nine welds previously inspected and found to be acceptable by the first inspector. All nine welds were found acceptable by the second QC inspector.
A subsequent inspection of the same nine sample welds by licensee QA staff determined that two of the nine welds inspected by the second SGT QC inspector were undersized. As a resu!!,
SGT generated NCR 262, issued October 31,1997, and instituted an inspection program to re-inspect 100% of the welds accepted by the two QC inspectors and a 10% random sampling of all St. Lucie NPS welds accepted by all of the other SGT QC weld inspection personnel. As a r:sult of the re-inspection effort, SGT identified a total of nine unacceptable welds that had be:n accepted by the first QC weld inspector; two of which were erroneously found as acceptable by the second OC inspector. The first inspector was subsequently removed from wsid inspection duties, whereas, the second OC weld inspector was retrained and retumed to
duty. Since the systems that contained the unacceptable (undersized) wCds had bien turned l over to the licensee, and the condition was a deviation, as defined in @21.3 of 10 CFR Part 21, l SGT was required to evaluate the deviation to determine whether a substantial safety hazard !
could exist.
The NRC inspectors determined that SGT processed the two nonconformances in accordance )
with its procedure adopted to implement the provisions of 10 CFR Part 21, SGT Quality Execution Procedure (CEP) 15.1, " Identification and Control of Deviations," St. Lucie NPS Project, Revised October 7,1997. Discussions with the cognizant St. Lucie NPS Project Quality Manager (POM) and a review of the procedure determined that OEP 15.1 did not contain adequate criteria to allow the POM to determine how the deviation should be dispositioned in accordance with 10 CFR Part 21, including whether SGT would have the capability to perform the required evaluation, as discussed in @21.21(b).
The inspectors met with the MK Engineering and Construction Group Quality Director (GOD) to discuss how the generic view of this situation had been handled and whether the possibility that undersized fillet welds existed at other MK/SGT nuclear work sites had been addressed. Since two SGT welding inspectors had failed to identify the undersize welds, this did not appear to be an isolated case. The NRC inspector questioned whether a root cause for why the qualified welders and two O.C. inspectors had failef to identify the unacceptable fillet welds (e.g., poor training, inappropriate measuring tools, insufficient knowledge of the applicable ASME coue),
had been considered.
Additionally, there appeared to be an overalllack of MK/SGT guidance on how to perform and document a thorough evaluation in OEP-15.1, that is, quality records were not present in the NCR packages to show what generic applicability activities had been completed. After the NRC inspectors questioned the inadequate information that was contained in the packages, the MK staff contacted its Personnel department to determine if either of the OC inspectors had worked at other MK work sites.
It was noted that MK/SGT satisfactorily identified the extent of the questionable welding inspections at the St. Lucie project and had taken appropriate corrective action regarding the two suspect inspectors. However, as a result of the lack of adequate procedural criteria to ensure that effective evaluations were performed, MK/SGT staff did not consider the generic implications of the identified deviation. That is, MK/SGT did not determine whether either of the suspect OC inspectors had performed welding inspections for MK/SGT on any of its other nuclear power station projects. The lack of evaluation criteria guidance contained within SGT quality assurance procedure OEP 15.1 resulted in an inadequate evaluaCon of a deviation.
Violation 99901329/98-01-01, and 99901335/98-01-01 was identified la this area.
Before the NRC inspector's exit meeting the GOD identified to the NRC Inspectors that the first inspector had not been employed by MK/SGT on any other projects. Subsequent to the inspection, MK informed the NRC inspectors that the second weld inspector worked for MK/SGT at both D.C. Cook and Point Beach projects. However, the second inspector did not perform any weldment inspection at the D.C. Cook project, only the St. Lucie NPS and Point Beach steam generator replacements. MK staif informed the lead inspector that no welding anomalies had been identified during the Point Beach steam generator replacement. However, the lead insp ctor point!d out that lack of ntgative weld inspection results at Point Beach did not necessarily indicate that the second inspector had not accepted undersize welds at Point Beach. As a result of the discussion between the GOD and the Lead Inspector, MK committed to review the circumstances of the issue and inform the licensee of those circumstances on or bsfore July 17,1998,if applicable
- c. Conclusions The MK/SGT procedures lacked adequate guidance to perform a proper Part 21 evaluation. As a result, a deviation relating to potentially unacceptab!e welds at other MK/SGT sites was not taken into consideration.
3.3 OA Proaram Review for Control and Qualification of Weldina Activities
- a. Insoection Scoce The inspectors reviewed welding documentatiori for the Point Beach and St. Lucie NPS steam generator replacement projects to verify that ASME Code and MK/SGT quality assurance program requirements were met. The inspectors review included procedure qualification records, weld procedure specifications, weld data cards, and associated quality assurance manual chapters.
- b. Obcervations and Findinos inspector review identified examples where MK/SGT activities, or the procedures used for activities, did not meet ASME Code requirements and/or were not in compliance with written MK/SGT quality assurance program requirements. Examples of these instances are provided !
in the following sections:
b.1 WPS GT-SM/1.1-1 PB Revision 4 Comoliance with ASME Code The NRC inspectors identified that in November and December of 1996, MK/SGT weld procedure WPS GT-SM/1.1-1 PB Revision 4, used to fabricate steam generator replacerr.ent Main Steam and Main Feedwater weldments at Point Beach, did not meet certain requirements of ASME QW-409.1. The ASME requirement, as implemented by QW-253 and QW-256, does not allow changes in current or polarity, increases in heat input, or increases in deposited weld mstal over that which was qualified.
However, the inspectors noted that this weld procedure allowed welding to be performed with an increase in heat input over that which had been qualified. Exceeding the qualified heat input for these welds has a potential consequence of lowering of the desired impact strength of the weldment.
The NRC Inspectors revi w:d MK correspond:nce No. M-QM-97-045, June 3,1997, concerning the re-evaluation of Point Beach WPS's and POR's at Point Beach's request. After the review, the NRC inspectors were concerned about the capability of the MK quality assurance program that is used to evaluate ASME Code weld procedures. The MK letter states at the conclusion of the MK re-evaluation that "SSW (Point Beach Site Specific WPS] heat input allowable were at, or below, that allowed by the POR."
The re-evaluation failed to identify that the Main Steam and Feedwater welds fabricated using gas tungsten arc welding (GTAW) in accordance with this WPS were fabricated using an unqualified weld procedure. These welds had a material thickness range between 5/8 and 1.5 inches and were not post weld heat treated (PWHT). The weld procedure allowed these welds to be fabricated with heat inputs of up to 47.4 kilo-Joule per inch (kJ/in). However, the qualification welds documented in the supporting POR GT/1.104, were welded with a maximum heat input of 29.5 kJ/in for material thicknesses between 5/8 and 1.5 inches that did not receive a PWHT. Therefore, the heat inputs authorized in WPS GT-SM/1.1-1PB Revision 4, exceeded the heat ;nputs irt the supporting qualification welds, which was contrary to ASME paragraph QW 409.1 requirements.
Increased heat input has the potential consequence of lowering the desired impact strength of the weld which bounds the materials ability to resis: crack propagation. Nonconformance 99901334/98-01-02 was identified in this area.
b.2 Point Beach Weld and Procedure Qualification Records The inspectors reviewed SGT Quality Execution Procedure (OEP) 20.2," Welding Procedure Qualification," October 25,1996, for Wisconsin Electric's Point Beach station. The procedure required that the new site PORs must be approved by the MK Engineering and Construction Group Welding Engineer (GWE), o. delegated to the Project Welding Engineer (PWE).
The inspector identified that POR GT SM/1.1-05 was signed and approved by an SGT staff member on November 27,1996, without a previously issued delegation, and therefore was not in compliance with the vendor's written quality assurance program requirements. The POR was for use in qualifying a weld procedure to fabricate Point Beach Unit 2 safety related Main Steam and Feedwater welds. Since the same SGT staff member had reviewed and approved the applicable SGT quality assurance program requirement chapter (OEP 20.2) for Wisconsin Electric on October 25,1996, the inspector concluded that the individual was aware of this quality assurance requirement when he signed POR GT-SM/1.1-OS on November 27,1996. l l
The inspector identified that the control of weld procedures (review and approval) for welding on safety related components / systems did not appear to be in compliance with the vendor's quality assurance program requirements. Inspector review of a sample of weld procedures I listed below, found 13 that were approved by an individual who had not been directly delegated authority by the GWE as required by the quality assurance program. Because the GWE did not i specifically delegate authority to the individual the inspectors concluded that the following steam generator replacement project WPS's signed by that individual were not in compliance with MK/SGT requirements, including (a) Section 9.2.1 of the MK Quality Assurance Manual, 7
I l
August 20,1996, "W;tding cnd Hnt Tr:ating - Wald Procidurc Qualifications", (b) Se etion 9.2.6 of the MK QAM, and (c) Quality Assurance Instruction (OAl) 11.2, September 23,1996,
" Control of Welding Procedure Specifications." Nonconformance 99901329/98-01-03 was identified in this area.
l
- WPS GTM/1.1-2PB Revision 1 December 21996, l
= WPS GTM/1.1-3PB Revision 0 November 27,1996
- WPS GT-SM/1.1-1PB Revision 4 November 28,1996 )
WPS GT-SM/1.1-1PB Revision 3 November 23,1996 e
- WPS GT-SM/1.8-1PB Revision 1 November 23,1996
. WPS GT-SM/43.43PB Revision 1 November 23,1996
- WPS GT-SM/8.81PB Revision 2 November 23,1996
. WPS GT SM/8.8-1PB Revision 1 November 5,1996
. WPS GTM/1.1-2 PB Revision 0 November 27,1996
- WPS GTM/1.1-3 PB .
Revision 1 December 3,1996
. WPS POWPS-04PB Revision 0 October 24,1996
. WPS WP-2-1PB Revision 0 December 4,1996 l
. WPS GT-SM-BU/1.3-1PB Revision 1 November 23,1996 in an associated area, the inspectors de' ermined through discussions with SGT and MK staff that some ambiguity existed between tne SGT and MK QA program control in the different ,
w;lding procedure specifications. MK management expectations were made clear to the l inspector in that the control of welding procedure specifications for the safety related steam ginerator replacement welding activities at Point Beach and other licensee sites, should be closely monitored and approved by the MK GWE. The MK quality assurance documentation r; viewed by the inspector supported that practice. However, the inspector's review of the SGT quality assurance documents and discussions with SGT personnel, determined that the MK quality assurance requirements may not have been carried through to the site quality assurance requirements in this area. That is, the quality assurance requirements for the control of weld procedures for welding on safety related components and systems appeared to be inconsistent, with differing review and approval requirements. This area appeared to have some weaknesses.
b.3 Post Weld Heat Treatment Reoardina ASME Code The NRC inspectors reviewed the steam generator replacement, Main Steam and Main FOedwater weld p' medure WPS GT-SM/1.1-1 PB Revision 4, and found that it did not prescribe a post ,teld heat treatment on weldments with a thickness of 1.5 inches and below.
The inspectors i entified that on December 9,1996, and December 10,1996, post weld heat tr atments wert applied to two Main Steam welds with a materia! thickness of 1.34 inches.
However, the QA program controls that MK/SGT used in this area had not identified the di;viation from ASME Code requirements.
The ASME requirements, ASME Section IX,1995 Edition, QW-409.1, as implemented by requirements QW-253 and QW-256, required that "A change in the type of current or polarity, an increase in heat input, or an increase in volume of weld metal deposited per unit length of w:Id, [is not allowed] over that qualified."
-8 inspectors id:ntified th:t in Nov:mb:r and D:cemb:r of 1996, MK/SGT w:Id procedure WPS GT-SM/1.1-1 PB Revision 4, used to fabricate steam generator replacement Main Steam and Main Feedwater weldments at Point Beach, did not meet the requirements of the above stated ASME requirement QW-409.1. This weld procedure allowed welding to be performed with ari increase in heat input over that which had been qualified.
Inspectors reviewed MK correspondence No. M-QM-97-045 dated June 3,1997 concerning the re-evaluation of Point Beach WPS's and POR's at Point Beach's request. After review, inspectors became concerned about the capability of the MK quality assurance to evaluate ASME Code weld procedures. The MK letter states at the conclusion of the MK re-evaluation that "SSW [PB Site Specific WPS] heat input allowables were at, or below, that allowed by the POR." The re-evaluation apparently missed the fact that the Main Steam and Feedwater welds fabricated using gas tungsten arc welding (GTAW)in accordance with this WPS were fabricated using an unqualified weld procedure. These welds had a material thickness range between 5/8 and 1.5 inches and were not post weld heat treated (PWHT). The weld procedure allowed these welds to be fabricated with heat inputs of up to 47.4 kilo-Joule per inch (kJ/in).
However, the qualification welds documented in the supporting POR GT/1.104, were welded with a maximum heat input of 29.5 kJ/in for material thicknesses between 5/8 and 1.5 inches that did not receive a PWHT. Therefore, the heat inputs authorized in WPS GT-SM/1.1-1PB Revision 4, exceeded the heat inpuo in the supporting qualification welds, which was contrary to ASME paragraph QW 409.1 recsements. Using weld procedures not in compliance with ASME Code is considered a nonconformance. This is identified as an example of Nonconformance 99901334/98-01-03.
b.4 WPS GT-SM-SU/1.311 PB Revision 1 Comoliance with ASME Code The NRC inspectors reviewed weld procedure WPS GT-SM-BU/1.3-1 PB Revision 1, that was used at Point Beach in November 1996. The inspectors determined that the procedure did not meet the requirements of ASME QW-409.1. The weld procedure allowed welding to be performed with an increase in heat input over that which had been qualified. Although this weld was later removed for other reasons at the request of Westinghouse, complete weld fabrication using this WPS took place without MK/SGT quality assurance program identification that an unqualified weld procedure was in use.
b.5 St. Lucie NPS & Point Beach Permanent Plant Records (Weld Data Cardt The NRC inspectors record review identified that as of March 9,1998, MK/SGT had not provided ecurate permanent plant records for certain welding processes used to assure the continued saNty, quality, and operability of the Point Beach Unit 2 steam generators, for example:
(a) The NRC Inspector found permanent plant welder records for the St. Lucie NPS steam generator replacement project that did not meet SGT QEP 20.1 " Control of Welding,"
October 31,1997, and SGT QEM 20.5, " Control of Filler Metal," October 24,1997, quality assurance requirements. Of the 36 welders entered on an SGT weld data card
-5 8-1 i
l quality docum2nt only six tctually perform w:Iding on the wsidm:nt. Tha licensee's steam generator project QA identified and brought the matter to SGT's attention (St.
Lucie NPS OR No. 97-6637) during a routine review of welding documentation.
(b) The NRC Inspector determined that an SGT Point Beach Weld Data Card, WDC 3081-A-2, and WDC 3081-B-7 did not have the applicable revision recorded for the weld procedure used to fabricate the steam generator replacement main steam weld.
Consequently, when Revision 4 of the welding procedure was determined to be an unoualified procedure, the licensee could not identify that these welds required requalification.
The quality documents for steam generator replacement welds did not contain sufficient records for the details of the welding that had been performed as required by 10 CFR Part 50, Appendix B requirements. The MK/SGT QA/QC programs had not identified the inconsistencies in l accordance with its OEP 20.1 and 20.5 quality assurance requirements. This matter is identified as an example of Nonconformance 99901334/98-01-04. l B.6 ASME Reauired Droo Weicht Testina
- a. Insoection Scoce The inspectors reviewed documentation of drop weight tests conducted to qualify ASME l welding procedure specifications (WPSs) for D. C. Cook and the Point Beach steam generator !
raplacement projects. i l
- b. Observations and Findinas The procedure qualification requirements for ASME WPSs involved Charpy V-notch (Cv) impact testing where the WPSs were to be used on materials for which ASME Section ll1 specified fracture toughness as a design requirement.
The testing of the Cv specimens required that an appropriate test temperature be established through consideration of the Lowest Service Temperature (LST) and/or Tuor, as defined by l ASME.
LST is the minimum temperature of the fluid retained by the component or, the l
calculated minimum metal temperature whenever the pressure within the component i exceeds 20% of the preoperational system hydrostatic test pressure.
Tuor is the temperature at or above the nil-ductility transition temperature NDT (ASTM E 208 81); Tuor is 10*F below the temperature at which two specimens show no-break performance.
The selected test temperature for the Cv specimens, Tev, in turn established the reference t:mperature (RTuor) for the qualified WPSs. RTuor is defined by ASME as the higher of Tuor and (Tev - 60*F) and establishes LST for materials welded using the WPSs.
- .- .- =. - __ _ . _ . - - _ _ - . - - _ .
Th3 Procedura Qualification Record (POR) for ths D. C. Cook WPS was POR No. 3-8 for the i Shielded Metal Arc Welding (SMAW) of A-302 Grade B base material to itself, (P3 Group 3 to P3 Group 3). For this POR, a series of drop weight tests was conducted which established the Tuor as -100*F for the weld metal zone and -60*F for the base material. The Cv specimens for POR No. 3-8 were tested at Tcv = +50*F which established the RTuor as -10*F for the qualified WPS. These tests met the requirements of the applicable ASME Section IX, and Section Ill.
1980 Edition Summer 1983 Addenda.
l The POR for the Point Beach WPS was POR No. FC/3.3-Q1 for the Flux Core Arc Welding I (FCAW) of SA-533 Grade B Class 1 to SA-533 Grade B Class 2, (P3 Group 3 to P3 Group 3). l For this POR, two drop weight specimens were tested at a test temperature of +5'F. After both specimens met the "no-break" requirement, the POR No. FC/3.3-Q1 Cv specimens were tested at Tev = +55'F which established the RTuor as -5'F for the qualified WPS. These tests met the requirements of the applicable ASME Section IX,1995 Edition, and Section Ill.1986 Edition.
- c. Conclusions Although the Point Beach, POR No. FC/3.3-Q1, drop weight test program was abbreviated.
when compared to the D. C. Cook POR No. 3-8, drop weight testing, the inspectors found that either method met the requirements of the ASME Code.
b.7 Procedure Qualification Record (POR) GT/3.3-C2
- a. Insoection Scoce The inspectors reviewed the documentation supporting special welding procedure qualification testing for the Point Beach I"it 2 steam generator replacement project.
- b. Observations anc f jnas A recent NRC Region til inspection of Point Beach related welding activities raised questions concerning the requalification of the Gas Tungsten Arc welding (GTAW) portion of WPS GT-SM/3.3-2PB. One of the questions involved the use of a single process POR to expand the qualification range of a multiple process WPS, On May 20,1997, a 1%-inch thick test plate was welded using the GTAW process and high heat inputs. At the time of the test, the test assembly was assigned the POR No. GT/3.3-Q2.
This test plate was welded using the GTAW process for the %-inch from the center of the plate to one surface; the opposite side of the plate was also welded with the GTAW but only to 3 balance the welding stresses and maintain the straightness of the test assembly. The test assembly portion of the weld was made using heat inputs of 61.4 kJ/in. through 105.9 kJ/in. On May 27,1997, Bodycote Taussig, incorporated (BTl), reported that the weld metal from this test assembly had been successfully impact tested as a welding procedure qualification coupon.
The inspectors determined that, having completed welding and testing of POR GT/3.3-Q2 assembly in May 1997, MK/SGT did not immediately follow through by documenting the test with a POR. The records showed that MK personnel considered canceling the POR No.
GT/3.3-Q2 b:cause thsy felt that tha t:st was not n:cd:d to support the WPS used at Point Beach. The records also showed that MK/SGT considered that POR GT-SM/3.3-Q1 was technically in compliance with ASME requirements. POR No. GT/3.3-Q2 was finally issued on April 16,1998. The inspectors also found that the documentation of the decision process involved in: the original decision to weld test assembly No. GT/3.3-02; the decision to not complete the POR process; and the decision to finally issue the POR were not clearly documented.
On the question of a single process POR being used to expand the qualification range of a multiple process WPS, the inspectors reviewed the original test weld data for POR No. GT-SM/3.3-01. This review showed that the original test weldment consisted of a 1%-inch thick tsst plate welded using the GTAW process for the %-inch from the center of the plate to one surface, and the SMAW process for the other %-inch from the center of the plate to the opposite surface. This method of welding the 1%-inch thick test plate essentially constituted two separate welding procedure qualifications which could then be used separately or in combination. ASME Section IX specifies that %-inch of weld metal deposited in a 1%-inch thick t:st plate constitutes a test. The inspectors also agreed that ASME Section IX allows the rcqualification of a single process to support changes to essential elements. (There is no requirement to retest both welding processes to support essential element changes in only one of the processes.)
- c. Conclusions The inspectors found that in the specific cases reviewed, the welding procedure qualification processes were in compliance with the requirements of the ASME Code.
b.8 D.C. Cook Main Steam and Feedwater Welds The NRC Inspectors reviewed D.C. Cook procurement records and an SGT/MK Work Package,
- 1531-A, and noted a disparity between licensee and vendor records. A September 18,1997, American Electric Power (AEP) letter, AEP NRC 1273, to the NRC Region 111 staff indicated that the Code requirements of ANSI B31.1 was the applicable piping Code used to fabricate the steam generator replacement main steam and feedwater final closure welds (new-to-old piping joints), at D.C. Cook. However the MK/SGT Work package provided by the MK/SGT staff l rppears to indicate that actual weld fabrication was performed to ASME Ill Code requirements.
The NRC Inspectors noted that MK could not easily identify the exact welde joints in question from the work package without extensive review, and the D.C. Cook Steam Generator Project Report was not present to determine whether the final closure weld joints had been reconciled b ck to the code of record. The inspector determined that this issue concems certain D.C.
Cook field welds and will be transmitted to NRC Region lit for its consideration for follow up.
l 1
l 1
3.4 Entrance and Exit Meetinas in the entrance meeting on May 20,1998, the NRC Inspectors discussed the scope of the inspection, outlined the areas to be inspected, and established interfaces with the MK/SGT management. In the exit meeting on May 22,1998, the NRC Inspectors discussed their findings and concerns.
- 4. PERSONS CONTACTED MORRISON KNUDSEN CORPORATION - ENGINEERING & CONSTRUCTION GROUP .
T.H. Zarges President & CEO M.F. Hendricks Group Quality Director (MK & SGT)
G.N. Hlifka Corporate Welding Engineer F.B. Kovacs Senior Quality Engineer L.E. Pardi Executive Vice President, Power Division THE STEAM GENERATING TEAM M.D. Cepkauskas President DUKE ENGINEERING AND SERVICES R.W. Bonsall Executive Vice President, Engineering & Environmental Group M.C. Scott Senior Engineer y 1 UNITED STATES j
- }
NUCLEAR RESULATORY COMMISSION WASHINGTON, D.C. 20666 0001
'o o,** September 4, 1998 Mr. Chris M. Powers, Vice President Quality and Regulatory Affairs Siemens Power Corporation - Nuclear Division P.O. Box 130 2101 Hom Rapids Road i Richland, WA 99352-0130
]
1
SUBJECT:
NRC INSPECTION REPORT NO. 99900081/98-01
Dear Mr. Powers:
On August 3-6,1998, the staff of the U.S. Nuclear Regulatory Commission (NRC) performed an i inspection of activities conducted by the Siemens Power Corporation - Nuclear Division (SPC),
ct the company's facilities in Richland, Washington. This letter transmits the report of that l inspection.
During the inspection, the NRC inspection team identified a concern where SPC did not request i NRC staff approval to extend peak pellet exposure limits for the 9x9-5 fuel design. SPC erred )
in using information from the 9x9-IX/9X Safety Evaluation Report (SER) to justify changes to '
the licensing basis for the 9x9-5 fuel. Prior to the NRC acceptance in 1995 of the SPC generic fuel design criteria, the staff wrote an SER approving each fuel type designed by SPC.
BIcause the SERs reference specific limits on peak bumup, staff review and approval would be necessary for SPC to exceed the peak burnup limits.
However, during the course of the inspection, SPC provided additional documentation to the tsam which provided a technical basis to extend the assembly average bumup from 40 GWd/t to 45 GWd/t and also considered extension of the peak pellet to 60 GWd/t. On the bases of that review, the team concluded that the issue was of low safety significance because the fuel 4 rods in the 9x9-IX/9X and 9x9-5 fuel are very similar. As a result, the team's concem was adequately addressed and closed during the inspection.
In accordance with 10 CFR 2.790, of the NRC's " Rules of Practice," a copy of this letter and its gnclosure will be placed in the NRC's Public Document Room. Should you have any questions conceming this inspection, we will be pleased to discuss them with you. Thank you for your cooperation during this process.
Sincerely, d @MX -
Suzann . Black, Chief Quality Assurance, Vendor inspection, and Maintenance Branch Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation
Enclosure:
Inspection Report No. 99900081/98-01 r . __ _ __ __ . . _ - - _ _ _ _ __
U.S. NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION l
Report No.: 99900081/98-01 Organization: Siemens Power Corporation - Nuclear Division :
- 2101 Horn Rapids Road Richland, Washington {
l
Contact:
Chris M. Powers, Vice President Quality and Regulatory Affairs Nuclear Industry Serves the U.S. nuclear industry by providing Activity: boiling- and pressurized-water reactors with fuel assemblies, reload core designs, and safety analyses.
Dates: August 3-6,1998 Inspection Team: Robert L. Pettis, Jr., HQMB/DRCH Anthony P. Ulses, SRXS/DSSA Carl E. Beyer, Consultant Approved by: Robert A. Gramm, Chief Quality Assurance and Safety Assessment Section Quality Assurance, Vendor inspection, and Maintenance Branch Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation Enclosure l
TABLE OF CONTENTS 1
1 1
l Ease l 1
1 INSPECTION
SUMMARY
...... .. .. ... .. ...... .... .. . ... . 1 2 STATUS OF PREVIOUS INSPECTION FINDINGS . . . . . . . . ..... . . .... 1 3 INSPECTION FINDINGS AND OTHER COMMENTS . ..... ... . . .. . .... 1 3.2 High Thermal Performance (HTP) Spacers . . . . . . ... . . . . .3 3.3 RODEX Quality Assurance . . . . . . . ... .. . . .. ...... .5 l 3.4 Review of 9X9-5 Fuel Design . . . . . . . . . . . . . ..... .. .. .. .6 !
3.5 Fuel Channel Growth . . . . . . ... .. . . . . .. .7 i 4 ENTRANCE AND EXIT MEETINGS . .... ...... .... ..... ... ... ... 10 '
i 4
1 I
l l
i
1 U.S. NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION Report No.: 99900081/98-01 Organization: Siemens Power Corporation - Nuclear Division 2101 Hom Rapids Road Richland, Washington
Contact:
Chris M. Powers, Vice President Quality and Regulatory Affairs Nuclear Industry Serves the U.S. nuclear industry by providing Activity: boiling- and pressurized-water reactors with fuel assemblies, reload core designs, and safety analyses.
Dates: August 3-6,1998 Inspection Team: Robert L. Pettis, Jr., HOMB/DRCH Anthony P. Ulses, SRXB/DSSA Carl E. Beyer, Consultant Approved by: Robert A. Gramm, Cnief Quality Assurance and Safety Assessment Section Quality Assurance, Vendor Inspection, and Maintenance Branch Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation l
Enclosure
- l
- - . - - - . ~ . . - . - . . = - . -. -.
1 1 INSPECTION
SUMMARY
From August 3-6,1998, representatives of the U.S. Nuclear Regulatory Commission (NRC) j conducted a performance-based inspection of the activities at Siemens Power Corporation - ,
Nuclear Division (SPC), in Richland, Washington. In conducting this inspection, the team Emphasized technically directed observations and evaluations of SPC's activities related to nuclear and mechanical engineering and manufacturing. As the technical bases for the inspection, the team relied upon the following:
- Part 21, " Notification of Failure to Comply or Existence of a Defect," as defined in Title to of the Code of Federal Regulations (10 CFR) 10 CFR 50.46, " Acceptance Criteria for Emergency Core Cooling Systems [ECCs] for Light-Water Nuclear Power Reactors"
- 10 CFR Part 50, Appondix B, " Quality Assurance Criteria for Nuclear Power Plants and ,
4 Fuel Reprocessing Plants" l
- 10 CFR Part 50, Appendix K, "ECCS Evaluation Models" SPC topical report EMF-1, " Quality Assurance Program for Nuclear Fuels and Services,"
Rev. 29, March 5,1998 (prepared by SPC's Engineering and Manufacturing Facility (EMF) and approved by the NRC on June 11,1998, as meeting the requirements of Appendix B to 10 CFR Part 50) ;
I During this inspection, the team did not review SPC's corrective actions taken to address NRC findings identified in NRC inspection Report 99900081/97-01. During this inspection, the team j identified an instance in which SPC failed to conform to NRC requirements However since adequate corrective actions were accomplished by SPC during the inspection, the issue was closed during the inspection.
2 STATUS OF PREVIOUS INSPECTION FINDINGS During this inspection, the team did not evaluate the current status of findings that remain open from the previous NRC inspection of SPC.
3 INSPECTION FINDINGS AND OTHER COMMENTS SPC performs core reload design analyses, fuel development engineering, safety and transient analyses, and other fuel-related services. This inspection included an evaluation of SPC Ectivities related to reload core design, neutronics and fabrication.
1 s
p 3.1 Debris Resistant Lower Tie Plate i insoection Scone I The inspection team evaluated the design adequacy of FUELGUARD," SPC's debris resistant
' lower tie plate. -The design adequacy J 7s based on the performance of the FUELGUARD*
lower tie plate in relation to Specified Acceptable Fuel Design Limits (SAFDLs) as required by General Design Criteria (GDC) 10 in Appendix A of 10 CFR Part 50 and Section 4.2 of the NRC Standard Review Plan (SRP). The inspection team primarily focused on the adequacy of the strength of the FUELGUARD lower tie plate.
Observations and Findinas ]
The FUELGUARD* lower tie plate is part of the structural skeleton of the assembly that i providee strength for the assembly to prevent significant deformation due to fuel handling and )
in-reactor accident loads. For pressurized-water reactor (PWR) fuel assemblies the guide' tubes are attached to the lower tie plate that makes up the structural skeleton for the lower part of the assembly. For boiling-water reactor (BWR) assemblies (with exception of the Atrium-10 design) the assembly structure is made of tie rods attached to the lower tie plate. For the - ,
Atrium-10 design the structure is made of the central water channel that is attached to the lower tie plate. The FUELGUARD lower tie plate is made up of a machined piece of stainless steel that has curved blades that are brazed in the tie plate. -
Several strength tests were performed by SPC on both prototypic and actual FUELGUARD lower tie plate designs used commercially for both PWR and BWR applications in the U.S. and ,
Europe. A PWR application in France (EDF FUELGUARD" Mechanical Strength Test, j EMF-92-193(P), January,1993) resulted in the most severe seismic-loss of coolant accident (LOCA) loads for a design basis accident (DBA) for this design to date. The testing results indicated that the FUELGUARD* lower tis ' late could withstand limiting DBA loads without cracking or significant deformation. Further testing (H. B. Robinson Robust FUELGUARD '
LTP Airbox and Strength Test, EMF-DTA-724 Revision 1, November 1997) at similar DBA load -
limits indicated that some of the prototype designs would result in cracks at the braz'ed joints of the curved blades but none of the blades were found to be broken or loose from the tie plate.
Further design adjustments were made to prevent this problem in the final design even though this was not a safety issue for DBAs within the United States. Further testing for a more robust designed FUELGUARD lower tie plate for application in H. B. Robinson indicated that the robust design would withstand dynamic loading (without failure) significantly greater than any ,
DBA loading. The BWR FUELGUARD lower tie plate was tested at a conservative upper >
bound design load limit and found to perform satisfactorily (Design Test Authorization for Atrium-10s and Atrium-10B FUELGUARD* Lower Tie Plate Prototype Evaluation, i EMF-DTA-648, Revision 4).
The team asked whether SPC had experienced significant damage or breakage in any of the FUELGUARD mechanical load tests. SPC responded that they had performed a special test for the Japanese in which a fuel bundle with a FUELGUARD* lower tie plate was dropped from a height of between 9 to 15 meters with the lower tie plate and bottom nozzle landing on a hard 2
l l
1 1
I
surface thtt resultId in cracking and dislodging parts of the curved blades from the FUELGUARD* lower tie plate. SPC calculated that this test resulted in impact loads on the lower tie plate that were several orders of magnitude greater than any fuel handling and DBA loads experienced in the United States.
Conclusions The team concluded that the SPC FUELGUARD lower tie plate will withstand bounding DBA and fuel handling loadings without breakage or significant deformation for current plant applications. The resulting DBA deformation will not reduce the coolable geometry of the fuel bundle or core as required by GDC 10.
3.2 Hiah Thermal Performance (HTP) Soacers insoection Scone The team evaluated the technical adequacy of SPC Zircaloy-4 HTP spacers used in PWR d: signs. There are three aspects to the performance of HTP spacers; the first is the adequacy of thermal hydraulic performance to maintain adequate departure from nucleate boiling (DNB) margins, the strength of the spacer grids to withstand seismic-LOCA loads,' and the spacer spring performance in relation to spacer spring relaxation to prevent fuel rod fretting failure due to flow induced fretting. If the spacer springs were to relax resulting in a large gap between the full rods and the spacer springs there could be a change in core dynamics during seismic loading but fretting wear and other problems would be observed during normal operation if th:se large gaps existed. This inspection concentrated on spacer spring relaxation with bumup and its impact on normal operation.
4 Observations and Findinos ,
The SPC PWR fuel assemblies with HTP spacers consist of six to seven HTPs that are equally spaced between the upper and lower bimetallic spacers. The HTP spacers are made of Ennealed Zircaloy 4 while the bimetallic spacers are made also made of annealed Zircaloy-4 cxcept for the spacer springs which are made of Incone!. The inconel springs generally retain (although not always) some small spring force, i.e.,90% or less of their original force, at high burnups. The Zircaloy-4 springs in the HTP spacers lose all of their spring force at high 4
bumups and due to cladding creepdown can result in a small gap between the Zircaloy spring ;
tnd the fuel rod. Fuel operating experience has shown that SPC Zircaloy spring-to-fuel rod g!ps on the order of 1 to 2 mils will not result in fuel rod fretting failures due to flow induce vibration of the rods against the springs or other spacer components. SPC further claims that gaps up to 6 mils will not result in significant fretting wear but this claim is based on out-of-r2 actor flow testing for only 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> while on cycle of in-reactor operation typically consists of 13,000 to 17,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> (Technical Basis for Design: High Thermal Performance Spacer and Intermediate Flow Mixing Assemblies for PWRs, EMF-S3100, Revision 2).
In addition, the out-of-reactor flow tests do not simulate the baffle jetting flow at the bottom and edge of the core in some plants that can result in additional fuel rod vibration and rod failure at the bottom spacer at the core edge. Although, SPC out-of-reactor flow tests do show that the 3
i 4
4
bimetallic spacers experience more frttting wzar than the SPC Zirciloy springs at equival:;nt gaps and flow conditions. The better fretting performance of the Zircaloy springs appears to be due to the design differences with the Zircaloy having a line contact and the Inconel springs having a smaller point contact surface.
The team focused on whether SPC had performed post-irradiation measurements of the spacer spring-to-fuel rod gaps and whether any fuel rod failures due to spacer spring fretting wear had been observed. Most of the SPC post-irradiation measurements that relate to spacer spring relaxation are pull tests that measure the force to move a rod vertically in the assembly. This measurement gives a qualitative measurement of the inconel spring forces in the top and bottom bimetallic spacers because most of the HTP Zircaloy springs have fully relaxed at high burnup. The exception was recent European post-irradiation examination that measured the spring-to-rod gaps from one lead assembly at a bumup of 46 Gigawatt-days per Metric Ton of I
Initial Uranium Metal (GWd/MTU) using a newly developed tool for this purpose. This examination showed that the spacer spring-to-rod gaps were on the order of 1 to 2 mils. These gaps are considered to be acceptable to prevent any significant rod vibration in the HTP spacer locations.
The team also focused on observed fuel failures in assemblies with HTP spacers. SPC has J demonstrated domestic operating experience with HTP spacers in four PWRs (nine reloads).
One plant had five reloads with three cycles (Cycles 9,11 and 12) with failures. The cause of failure in Cycle 9 was identified while the cause of the two failures in Cycle 11 were both at the '
bottom bimetallic spacer and believed to be due to baffle jetting. Two new failures were identified by ultrasonic testing in Cycle 12 and from examination were found to be due to fretting wear on the top bimetallic spacer. The fretting wear was due to o manufacturing defect in the spacer spring that resulted in a gap in the inconel spacer spring. Another failure was observed in another plant but this was believed to be due to debris because debris was found near the failed rod.
The team asked SPC whether they have used Zircaloy spacer springs in the top and bottom spacers of any of their PWR designs. SPC responded that they have recently placed lead test assemblies (LTAs) with Zircaloy springs in the top and bottom spacers but have not introduced this design change in commercial reloads at this time.
Conclusions TN) team concluded that HTP Zircaloy spacer springs appear to perform in a satisfactory manner based on the following observations:
- Out-of4sactor flow tests on assemblies with spacer spring gaps up to 6 mils in HTP spacers show significantly less fuel rod fretting than similar tests with bimetallic spacers.
. The small number of spacer spring gaps measured in HTP spacers from a high bumup assembly to date by SPC are only 1 to 2 mils.
4
e No frett' 1 failures have been observed in domestic p!st.ts that are due to HTP spacer l spring relaxation with 9 reloads while 3 rod failures due to fretting wear at the inconel spacer springs have been observed in either top or bottom bimetallic spacers.
, 3.3 RODEX Quality Assurance insoection Scooe The team reviewed Revision 2 of EMF-45047, Appendix C.3 of ANF-868(P), and EMF-2062(P) which control the use of different versions of RODEX. Furthermore, the team also reviewed several SPC Condition Reports (CRs) related to rnodifications made to the RODEX2 code. The tram focused on evaluating methods to ensure that the correct code was used for a particular analyses and that all relevant SER restrictions were met.
Observations and Findinos At SPC, RODEX is used primarily by Product Mechanical Engineering and the reload analysts.
Each group has its own version of RODEX and, until recently, also maintained their own version of the code. The decision to have two versions of the code was driven primarily by the fact that each group used a different computer platform for analysis. During the self-assessment activities which were undertaken as a result of corrective action to NRC Inspection Report 99900081/97-01, SPC discovered that there were several instances where an approved model was implemented by one group and not the other (Reference CR 6419). An example of this is 1 the latest NRC approved Gadolinia conductivity model which was introduced into RODEX2 but not RDX2LSE (Reference "RDX2LSE Gadolinia Conductivity Model," CR 6419 dated 1/7/98). ]
Based on discussions with the SPC staff the team leamed that this led SPC to have both VIrsions of the code maintained by one organization so that they will be modified consistently when new models are approved. Following the modifications to bring both codes up to the sIme modeling standard, SPC ran both codes on the same problem to confirm that they are !
giving the same results. This evaluation, documented in JGl:98:018," Comparisons of l RDX2LSE and RODEX2 for BWR Safety," dated July 31,1998, concludes that the codes did calculate consistent results.
Conclusions Although the tem e upressed concem upon discovering that SPC was maintaining two versions of an NRC app sved code, the team concluded that the QA program controls for RODEX was acceptable. The tser guidance for both versions of the code clearly distinguished which cptions are to be used for specific analyses and are consistent with the SER for RODEX. The tram considers the fact that SPC now has one organization maintaining both versions of RODEX to be an improvement over past practices and this should allow mistakes such as those docuraented in CR 6419 to be avoided in the future.
5
3.4 Review of 9X9-5 Fuel Desian insoection Scoce The team evaluated Supplements 1 and 2 to ANF-88-152(P), EMF-92-184, EMF-94-105(P),
KDH:94:041, and SLL:093:94 relating to the extension of the peak pellet bumup limit for the SFC 9x9-5 BWR reload fuel from 55 GWd/t to 60 GWd/t. This issue was originally identified by SPC during preliminary work related to Cycle 8 of Grand Gulf, Unit 1. The team reviewed the use of calculations from another fuel type as they were applied to the 9x9-5 fuel.
Observations and Findinas During a review of the calculation plan for upcoming Grand Gulf Cycic 8, SPC determined that the work for Cycle 7 did not explicitly evaluate the peak pellet exposure for the 9x9-5 fuel which would be discharged at the end of the cycle. Upon completing this analysis, SPC discovered that the peak pellet exposure limit of 55 GWd/t would be violated if Cycle 7 was operated as designed. Therefore, SPC needed to justify extension of the peak pellet exposure limit for 9x9-5 fuel to allow Grand Gulf to operate Cycle 7 as designed and to facilitate Cycle 8 design.
To justify this extension, SPC referred to the SER written for the 9x9-lX/9X fuel which was approved to a peak pellet exposure limit of 60 GWd/t.
Because the fuel rods in the 9x9-IX/9X fuel are similar in design to the rods in the 9x9-5 fuel SPC concluded that they could use the approval for the 9x9-IX/9X to justify extending the peak pellet exposure limits for the 9x9-5 fuel from 55 GWd/t to 60 GWd/t. This conclusion was documented in a July 1,1994, SPC leMer to Entergy and is also included in the references for the design documents related to Cycle 8 as justification for the bumup extension.
Conclusions The team concluded that SPC did not justify the extension of the peak pellet exposure limits for the 9x9-5 fuel because they did not request NRC staff approval of the change. Prior to the "
NRC acceptance in 1995 of the SPC generic fuel design criteria (" General Mechanical Design Criteria for BWR Fuel Designs," EMF-89-98(P)(A), Revision 1 and Supplement 1, dated May 1995), the staff wrote an SER, which prescribed limits on peak bumup, approving each fuel We designed by SPC. Because these limits were included in the SER, any changes to them would require staff review. Furthermore, as is stated clearly in all SER transmittal letters from the staff, SERs apply "...only to the matters described in the topical report." SPC cannot, therefore, apply information in an SER to anything other than the material reviewed in the SER.
SPC applied the information from the 9x9-IX/9X SER to justify changes to the licensing basis for the 9x9-5 fuel. Consequently, the team found that SPC's failure to submit the revised limits for NRC review and approval constituted a failure to comply with the SER peak pellet exposure limit of 55 GWd/t for the 9X9-5 fuel.
However, during the course of the inspection, SPC provided to the team Condition Report CR 6719, Revision 0, dated March 9,1998, which identified that generic design criteria was used to extend assembly exposure without NRC review and approval. The CR was identified by SPC during a compliance review associated with SER rest'ictions for fuel mechanical design 6
which was initiated, ir, part, as a result of corrective actions associated with NRC Inspection Report 99900081/97-01. On July 31,1998, SPC forwarded to the staff Supplement 2 of ANF-88-152, which mitigated the effect of the error relating to the extension of burnup limits for the 9x9-5 fuel. The report documented an evaluation using the generic fuel design criteria to extend the assembly average bumup from 40 GWd/t to 45 GWd/t and also considered extension of the peak pellet exposure to 60 GWt. The report, once included in the Grand Gulf Core Operating Limits Repoo will allow the currendy loaded fuel to be operated. On the bases of the review of the information provided by SPC, the team concluded that the issue was of low safety significance because the fuel rods in the 9x9-IX/9X and 9x9-5 fuel are vey similar.
3.5 Fuel ChanngJ Growth Insoection Scoce The team evaluated the quality assurance of the Siemens European channel growtt ' r used in the Atrium-10 assembly growth correlation and the technical adequacy of SPC calculations of rod and assembly growth based on calculated burnup and fast fluence.
Observations and Findinos Fuel rod and assembly growth for the BWR Atrium-10 design are calculated by SPC to 1) maintain clearance between the fuel rod upper end cap and the upper tie plate, and 2) maintain engagement between the lower tie plate seal spring and the fuel channel. Failure to maintain
~
adequate clearance between the upper end cap and the upper tie plate can sesult in fuel rod bowing which can produce two negative results: 1) reduce the thermal margins of the fuel rods, i.e., reduce the minimum critical power ratio (MCPR) margins, and/or 2) with sufficient bowing can result in fuel rod failure, howmr, the bowing would have to be severe. Failure to maintain engagement between the lower tie plate seal spring and the fuel channel will result in some coolant flow leaking from the assembly to the core bypass region which will also reduce the MCPR margins for the rods in the assembly.
Fuel rod and assembly growth are dependent on fast fluence (E>1 Mev) but SPC generally calculates assembly and rod-average burnuo values for their mechanical analyses. SPC has a correlation that calculates assembly average fast fluence from assembly average burnup and the adequacy of this correlation will be discusred later in this section.
The Atrium-10 assembly growth is based on assembly average bumups from European channel growth data. The central water channel that controls Atrium-10 assembly growth is made from the same material as the European channels are made (fully annealed Zircaloy) and, therefore, will grow at the same rate given the same fast fluence. Because the European channel growth data is correlated to assembly average burnup and is directly applied to the assembly average bumup of the Atrium-10 assemblies, the team was concemed whether the fast fluence values of the Atrium-10 water channel are the same as the fast fluence values of the European channels when both are at the same assembly average bumup. If there is a significant difference between the fast fluence of the Atrium-10 water channel and the European channels at equivalent assembly bumups, then the application to Atrium-10 would need to be corrected accordingly.
7 SPC provided a plot of the European channel data as a function of assembly average burnup with seven Atrium-10 assembly growth data points ranging from a burnup of 10 GWd/MTU up to 35 GWd/MTU that lie within the scatter of the European channel growth data (Letter, SPC to NRC Document Control Desk, " Atrium-10 Irradiation Growth Evaluation Model," February 24, 1997). If there were a significant bias in fast fluence between the Atrium-10 central channel and the European channels there would be a blas in the Atrium-10 growth data ! ; relation to the European channel growth data. The SPC plot shows that five of the seven Atrium-10 assembly growth data points lie slightly below the best fit to the European data while two data points lie above the line with one datum point near the upper bound of the data. However, there is some uncertainty with regards to the data point near the upper bound because a second measurement was made on this same assembly at a higher burnup and the assembly growth was less than measured at the lower burnup (previous cycle).
It is also noted that there is considerable scatter in the European channel growth data and that SPC applies upper and lower bound curves in a conservative manner. For example, the lower bound assembly growth curve is used to calculate rod to upper tie plate clearances and the upper bound assembly growth curve is used to maintain tie plate seal spring engagement with the channel. Based on the comparison of the Atrium-10 growth data to the European data and the large uncertainties that SPC conservatively applied to this data in their calculations, the team concluded that the European data appeared to be applicable to the Atrium-10 assembly growth without any corrections. However, SPC needs to continue to collect assembly growth data to verify that no correction is needed to maintain conservatisms in the growth ca:culation .
The team questioned whether the SPC correlation used for converting assembly average burnup to fast fluence values used for calculating rod growth is applicable to the Atrium-10 fuel rods. The SPC correlation between burnup and fast fluence is based on 8X8 and 9X9 fuel designs. In order to evaluate this concern, the team performed calculations to assess the correlation used by SPC to relate burnup to fast fluence. The staff used the New ESC-based Weighted Transport Code (NEWT) code with a 44 group ENDF/BN based library. NEWT is a discrete ordinates code using the method of characteristics. An examination of the NEWT calculated fast flux indicates that the presence of the water channelin the ATRIUM-10 fuel causes the flux to peak with a peaking factor of around 1.2 while the previous SPC designs had significantly lower fast flux peaking. Because the method used by SPC to relate bumup to fluence is based upon the assembly average fluence, the team estimated that the SPC method could be underpredicting the fast fluence and, therefore, the rod growth by as much as 20%. It is noted that SPC's calculation of rod to tie plate clearances conservatively uses their upper bound curve for o growth. This curve bounds all of their 7X7,8X8 and 9X9 fuel2 rod growth data out to an L .;mbly average fluence of approximately 10 x 102' neutrons /cm with the exception of one data point at low fluence. This data includes the variation in rod growth due to fast flux peaking for the earlier designs but as noted, the peaking is significantly less than the peaking for Atrium-10.
The team also considered the SPC generated correlation to relate bumup to fast fluence.
CASMO-3G was used to calculate the fast fluence as a function of bumup for SPC 8x8 and 9x9 fuel and the results were correlated to allow for quick bumup to fluence conversion. The team assessed SPC's evaluation of this correlation's applicability to 10x10 designs. The SPC calculated results indicated that up to the point of peak bundle reactivity, the SPC correlation 8
t- . - _ - - . - _ - _ _ - - _ _ _ _ _ _ _ _ _ _ __ _ _ _ ___ _______ __ _
underpredicted the fast fluence for the ATRIUM-10. Based upon a rough integration of these results, the team determined that this could introduce up to 10% additional error (for a total of , .
30% error) into the fast fluence and, therefore, the rod growth calculation. The SPC correlation considered rod peaking in the older fuel designs, but this does not bound its application to ATRIUM-10.
The team also examined the impact on the rod end cap to upper tie plate clearance for the Atrium-10 design if the SPC rod growth curves were biased non-conservatively. For example, the team assumed the upper bound rod growth was 25% greater than calculated by SPC at an assembly average bumup of 50 Gwd/MTU. This calculation demonstrated that there would be a .
very smallinterference fit and a very small amount of rod bowing that would not impact thermal margins. Also, if the lower . ' W assembly growth were 25% less than calculated by SPC a clearance would remain betwew the rod end cap to upper tie plate. In addition, even if the rod bow were large enough to nearly close the gap between an adjacent fuel rod the MCPR limit would not be exceeded because the rod powers at this high burnup are relatively low, in order to assess the impact of SPC underestimating the assembly growth in the calculation of lower tie .
plate seal spring engagement to the fuel channel for the Atrium-10 design, the inspection team also assumed a 25% higher assembly growth rate than calculated by SPC and found that engagement was still maintained.
The team reviewed inspection reports, prepared by Siemens Kraftwerk Union (KWU) AG, Germany, during the 1982 through 1994 period, to determine if adequate quality assurance had been applied to the data acquisition, measurement and inspection process associated with fuel channel growth data. Since the documentation was written in German, SPC provided a translation of the main portions of each report. The reports generated dur'ng the 1982-1992 period mention the use of a computerized data acquisition system and identified the inspection requirements and the methods of documenting the actual test results for the bow and dimensional measurements of the fuel channel. Although a separate measurement procedure did not exist prior to 1992, a 1982 report of measurements taken at the Brunsbuettel reactor site in Germany described the requirement to inspect the straightness of the fuel channel s inspection system, which was calibrated using a reference dummy fuel channel. The report also described the general methodology used to determine the length of the inspected fuel channels which was accomplished using a measuring tape attached to the top of the fuel channel. By lifting the fuel assembly with the fuel channel using the telescope mast of the refueling stage, the length of the channel could be determined by using a television camera and monitor.
SPC provided the team with Siemens Engineering Procedure F-AW B-T030, " Measurement of Fw Elements, Fuel Channels and Fuel Rods," Revision 1, dated October 25,1993, which described the procedure for length measurements of fuel assemblies, channels, and fuel rods.
The procedure also described that the technical characteristics of the measurements are relevant for use as quality acceptance criteria and references requirements for calibration, accuracy, and documentation. SPC also provided the team with Siemens Engineering Proce dure F-AW B-T045, " Interfaces and Data Management," Revision 0, dated November 19, 1993, which described the interfaces and data management related to the evaluation, '
acceptance, distribution, and storage of the data.
9
~
h Conclusions The team concluded that while SPC does not accurately calculate the rod and assembly fast fluences for the Atrium-10 fuel design in calculations of rod to tie plate clearances and the engagement between the tie plate seal spring to the fuel channel, there is minimal impact on fuel safety and fuel failure is not likely as a result of these imprecise calculations. The bases for this conclusion are as follows:
. The SPC uncertainties for their Atrium-10 assembly growth correlation are very large and this results in significant conservatism in their lower and upper bound curves for assembly growth. This conservatism most likely bounds the inaccuracies in their determination of fast fluence.
. The clearance and engagement safety concems in relation to thermal margins of the fuel rods are significantly reduced at high burnups due to the significant reduction in rod powers at these bumups.
The team also reviewed technical reports prepared by Siemens Power Generation (KWU),
which documented field measurements of fuel channels recorded during the 1982-1994 period.
Although a separate measurement procedure did not exist prior to 1992, a 1982 report of measurements taken at the Brunsbuettel reactor site in Germany provided the most complete description of the methods used to obtain field measurements. The team concluded that the reports provided reasonable assurance that fuel channel growtn data could be relied upon for input into SPC fuel channel growth calculations.
i 4 ENTRANCE AND EXIT MEETINGS During the entrance meeting on August 3,1998, the NRC team met with members of SPC management and staff and discussed the scope of the inspection. The team also reviewed its respe;uibilities for handling proprietary information as well as those of SPC. In addition, the team established contact persons within the management and staff of the applicable SPC organizations. The team discussed the results of the inspection with SPC management and staff on August 6,1998.
10
s PARTIAL LIST OF PERSONS CONTACTED R. Feuerbacher Vice President, Engineering l C. Powers Vice President, Quality and Regulatory Affairs J. Mallay Director, Regulatory Affairs R. Reynolds Director, Research and Technology M. Garrett Manager, BWR Safety Analysis T. Howe Manager, Product Mechanical Engineering J. Holm Product Licensing Manager H. Curet Product Licensing Manager C. Brown Staff Engineer, Mechanical Analysis M. Smith Senior Engineer, Mechanical Analysis N.
9
/
f .b 11
pnMau a t UNITED STATES s* g NUCLEAR REGULATORY COMMISSION
- WASHINGTON, D.C. 20666 c001 W4 $
%,,., July 29,1998 Mr. Nicholas J. Liparulo, Manager Equipment Design and Regulatory Engineering Department Westinghouse Electric Company P.O. Box 355 Pittsburgh, Pennsylvania 15230-0355
SUBJECT:
NRC INSPECTION NO. 99900404/98-02
Dear Mr. Liparulo:
This refers to the inspection conducted by Robert Pettis, Jr. and Joseph Petrosino of this office and Melvin Holmberg of NRC Region lit on June 15-17,1998. The purpose of the inspection was to review the implementation of the Westinghouse quality assurance (QA) program, including 10 CFR Part 21 reportability aspects related to ice condenser containment components, and to review Westinghouse corrective actions from a previous NRC inspection. At the conclusion of the ,
inspection, the findings were discussed with you and members of your staff identified in the enclosed report. ,
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). These violations are of specific concern to the NRC since they have the potential to affect equipment operability for those ,
plants employing an ice condenser containment.
You are required to respond to this letter and should follow the instructions specified in the enclosed Notice when preparing your response. In your response, you should document the specific actions taken and any additional actions you plan to prevent recurrence. Your response may reference or include previous docketed correspondence, if the correspondence adequately addresses the required response. After reviewing your response to this Notice, including your proposed corrective actions and the results of future inspections, the NRC will determine whether further Nf3C enforcement action is necessary to ensure compliance with NRC regulatory ;
requirements.
During this inspection it was also found that the implementation of your QA program failed to meet certain NRC requirements. The most significant was Westinghouse's failure to address in its June 22,1995 report to the Tennessee Valley Authority (TVA), the evaluation of broken and cracked ice condenser sheet metal screws, which was reported to Westinghouse in a June 2,1995 TVA metallurgical report for the Watts Bar Nuclear Plant.
The inspection team also identified several Westinghouse drawings which did not contain material specifications for ice condenser sheet metal screws which called into question the suitability of existing screws installed in domestic ice condenser containment plants, and may have contributed O,
2 to a Licensee Event Report submitted to the NRC by Duke Power Company in May 1990. The specific findings and references to the pertinent requirements are identified in the enclosures of this letter.
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.
On June 24,1998, NRC headquarters and regional staff participated in a conference call with you and members of your staff to discuss the status of corrective actions associated with the findings identified during the inspection and a schedule for performing metallurgical testing fcr a sample of sheet metal screws supplied to Westinghouse by the D.C. Cook nuclear plant. It was noted during our discussion that these actions are presently being performed under Westinghouse's 10 CFR Part 21 program. We also acknowledge receipt of your June 26,1998, letter which documented the schedule for completion of such actions.
In accordance with 10 CFR Part 2.790 of the NRC's " Rules of Practice " a copy of this letter, its enclosures, and your response will be placed in the NRC's Public Document Room (PDR). To the extent possible, your response should not include any personal privacy, proprietary, or safeguards information so that it can be placed in the PDR without redaction, if personal privacy or proprietary information is necessary to provide an acceptable response, then please provide a bracketed copy of your response that identifies the information that should be protected and a redacted copy of your response that deletes such information. If you request withholding of such material, you must specifically identify the portions of your response that you seek to have withheld and provide in detail the bases for your claim of withholding (e.g., explain why the disclosure ofinformation will create an unwarranted invasion of personal privacy or provide the information required by 10 CFR 2.790(b) to support a request for withholding confidential commercial or financial information).
If safeguards information is necessary to provide an acceptable response, please provide the level of protection described in 10 CFR 73.21.
The responses requested by this letter and the enclosed Notices are not subject to the clearance procedures of the Office of Management and Budget as required by the Paperwork Reduction Act of 1980, Pub. L No.96-511. Should you have any questions concerning this inspection, we will be pleased to discuss tnem with you.
Sincerely, f
Suzanne C. Black, Gnief Quality Assurance, Vendor inspection, and Maintenance Branch Division of Reactor Controls and Human Factors Office of Nuclear Reactor Regulation
Enclosures:
- 1. Notice of Violation
- 2. Notice of Nonconformance
- 3. Inspection Report No. 99900404/98-02
NOTICE OF VIOLATION ,
Westinghouse Electric Company Docket No. 99900404 During an NRC inspection conducted June 15-17,1998, violations of NRC requirements were identified. In accordance with the " General Statement of Policy and Procedure for NRC Enforcement Actions," NUREG-1600, the violation is listed below:
Section 21.21, " Notification of failure to comply or existence of a defect and its evaluation," of 10 CFR Part 21, requires, in part, that each individual, corporation, partnership, dedicating entity, or other entity subject to the regulation adopt appropriate procedures to evaluate deviations and failures to comply to identify defects and failures to comply associated with substantial safety hazards as soon as practicable, and, except as provided in paragraph 921.21(a)(2) of 10 CFR Part 21, in all cases within 60 days of discovery, in order to identify a reportable defect or failure to comply that could create a substantial safety hazard, were it to remain uncorrected.
Contrary to the above, Westinghouse Electric Company did not identify and evaluate the following .
deviations:
. Ice basket she et metal screws identified to Westinghouse Electric Company in Tennessee Valley Authority (TVA) Central Laborstories Services Technical Report 95-1021 " Watts Bar Nuclear Plant, Ice Condenser Basket Screws," dated June 2, 1995. Specifically this report, faxed to Westinghouse on June 8,1995, identified screws with manufacturing induced quench cracks and core hardness values in excess of design specification values.
. Shunt trip plungers that were stuck in two Westinghouse model DB-75 circuit breakers installed at the Ginna nuclear power station which Rochester Gas and Electric Company had informed Westinghouse about in November 1997.
(99900404/98-02-01)
This is a Severity Level IV violation (Supplement Vil).
Pursuant to the provisions of 10 CFR 2.201, Westinghouse Electric Company is hereby required to submit a written statement or explanation to the U.S. Nuclear Regulatory Commission, ATTN:
Documerjt Control Desk, Washington, D.C. 20555 with a copy to the Chief, Quality Assurance, Vendor Inspection, and Maintenance Branch, Division of Reactor Controls and Human Facto s, Office of Nuclear Reactor Regulation, within 30 days of the date of the letter transmitting this Notice of Violation (Notice). This reply should be clearly marked as a " Reply to a Notice of Violation" and should include for each violation: (1) the reason for the violation, or, if contested, the basis for disputing the violation, (2) the 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. Your response may reference or include previous docketed correspondence, if the correspondence adequately addresses the required response. Where good cause is shown, considaration will be given to extending the response time.
Enclosure 1 2
Because your response will be placed in the NRC Public Document Room (PDR), to the extent possible, it should not include any personal privacy, proprietary, or safeguards information so that it can be placed in the PDR without redaction.
If personal privacy or proprietary information is necessary to provide an acceptable response, then please provide a bracketed copy of your response that identifies the information that should be protected and a redacted copy of your response that deletes such information. If you request withholding of such material, you muit specifically identify the portions of your response that you seek to have withheld and provide in detail the bases for your claim of witnholding (e.g., explain why the disclosure of information will create an unwarranted invasion of personal privacy or provide the information required by 10 CFR 2.790(b) to suppcrt a request for withholding confidential commercial or financial information). If safeguards information is necessary to provide an acceptable response, please provide the level of protection described in 10 CFR 73.21.
Dated at Rockville, Maryland this 29th day of July 1998 2
7..
NOTICE OF NONCONFORMANCE Westinghouse Electric Company Docket No. 99900404 Based on the results of an NRC inspection conducted June 15-17,1998, it appears that certain of your activities were not conducted in accordance with NRC requirements.
A. 10 CFR 50 Appendix B, Ctiterion lli " Design Control" requires, in.part, that measures be established to assure that design changes, including field changes, shall be subject to de: sign control measures commensurate with those applied to the original design. Further, these meae.,res shall assure that the design basis for structures systems and components are correctly translated into specifications, drawings, procedures and instructions and provide for verifying the adequacy of design, such as by the performance of design reviews.
Section 4.5, " Design input" of Westinghouse Energy System Business Unit (ESBU)
Quality Management System (QSM) document (quality assurance topical report),
Revision 2, dated May 29,1997, states, in part, that Engineering is responsible for '
ensuring that the design output complies with design input requirements, customer and regulatory requirements, and considers the safe functioning of the designed items.
Contrary to the above,
- 1. Westinghouse had not adequately completed design reviews for the serviceability of the remaining ice basket sheet metal screws (potentially susceptible to in service failures) at the Watts Bar Nuclear Plant.
Specifically, the June 22,1995 Westinghouse assessment of this condition failed to include a design review of the cracking identified in TVA Central Laboratories Services Technical Report, " Watts Bar Nuclear Plant, Ice .
Condenser Basket Screws," dated June 2,1995, that had been provided to Westinghouse by fax on June 8,1995. Additionally, the Westinghouse assessment did not address the potential for failed ice basket segments to impact and damage the air handling units within the ice condenser, which
. could introduce glycolinto the sump potentially affecting emergency core
- cooling system performance. (99900404/98-02-02)
- 2. Westinghouse did not translate design basis information for material type, hardness and strength level for the ice basket sheet metal screws into ice basket drawings 1145E10 and 1191E57, Revision 7. Further, the goveming design specification (952745) stated that the material requirements for ice ' -
basket components would be described in these drawings. This condition appeared to be the basis for a Licensee Event Report, submitted to the NRC by Duke Power Company in 1990, which reported the installation of ice Enc'osure 2
I condenser screws cf the wrong mrts, rial due to a lack of material specification on the drawings. (99900404/98-02-03)
B. 10 CFR 50 Appendix B, Criterion V " Instructions, Procedures, and Drawings" requires, in part, that activities affecting quality shall be accomplished in accordance with documented instructions.
Section 5.0, " Document and Data Control" of Westinghouse ESBU QSM document, Revision 2, dated May 29,1997, states, in part, that all activities affecting quality of items and senrices are accomplished in accordance with controlled documents such as quality system manuals, procedures, work instructions, and controlled data such as customer order requirements.
Contrary to the above, potential conditions adverse to safety pertaining to failed ice baskets sheet metal screws at the D.C. Cook nuclear plant had not been entered into Westinghouse's 10 CFR Part 21 review program, as required by Section ill of Westinghouse procedure 21.0 " Identification and Reporting of Conditions Adverse to Safety," Revision 2. Specifically, on or about June 1,1998, Westinghouse received Gelles Laboratories Report "Metallurgl cal Evaluation of Screws From Cook Unit 1 ice Baskets," dated May 1998, which identified a number of potential causes, including quench induced cracks, for failed ice basket sheet metal screws. (99900404/98 04)
Please provioe 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, Quality Assurance, Vendor Inspection, and Maintenance Branch, Division of Reactor Controls and Human Factors, 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 s
Nonconformance" and should include for each nonconformance: (1) the reason for the nonconformance, or, if contested, the basis for disputing the nonconformance, (2) the corrective steps that have been taken and the results achieved, (3) the corrective steps that will be taken to avoid further noncompliances, and (4) the date when your corrective action will be completed.
Where good cause is shown, consideration will be given to extending the response time.
Dated at.Rockville, Maryland this 29th day of July 1998 2
s U. S. NUCLEAR REGULATORY COMMISSION
\
OFFICE OF NUCLEAR REACTOR REGULATION Report No.: 99300404/98-02 Organization: Westinghouse Electric Company '
Pittsburgh, Pennsylvania 15230
Contact:
Mr. Nicholas J. Liparulo, Manager Equipment Design and Regulatory Engineering Department Nuclear Industry Nuclear steam supply system design, components and Activity: services Date: June 15-17,1998 s
inspectors: Robert L. Pettis, Jr.,HQMB/DRCH Joseph J. Petrosino, HOMB/DRCH Melvin S. Holmberg, Region lll Approved: Robert A. Gramm, Chief Quality Assurance and Safety Assessment Section Quality Assurance, Vendor Inspection, and Maintenance Branch '
Division of Reactor Controls and Human Factors Office of Piclear Reactor Regulation Enclosure 3 1
1 INSPECTION
SUMMARY
The purpose of the insnection was to determine if quality activities performed by Westinghouse were conducted under the appropriate provisions of the Westinghouse 10 CFR Part 50, Appendix B, quality assurance program of record, and 10 CFR Part 21.
The inspection bases were as follows:
Appendix B, " Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants," to Part 50 of the Code of Federal Reaulations (10 CFR Part 50) 10 CFR Part 21, " Reporting of Defects and Noncompliance."
Westinghouse Energy System Business Unit (ESBU) Quality Management System (OSM) document (quality assurance topical report), Revision 2, dated May 29,1997.
During the inspection, the team identified the following instances where Westinghouse failed to conform to NRC requirements:
1.1 Violation Violation 99900404/98-02-01 was identified and is discussed in Sections 3.1 and 3.2 of this report.
1.2 Nonconformances Nonconformance 99900404/98-02-02 was identified and is discussed in Section 3.1 of this report.
Nonconformance 99900404/98-02-03 was identified and is discussed in Section 3.1 of this report.
Nonconformance 99900404/98-02-04 was identified and is discussed in Section 3.1 of this report.
2 STdTUS OF PREVIOUS INSPECTION FINDINGS Unresolved item 99900404/98-01-01 (Closed)
During the January 12-13,1998 NRC inspection at the Westinghouse Repair and Replacement Services facility in Cheswick, Pennsylvenia, the inspection team did not complete its review of an identified problem with a device used on Westinghouse 480-volt circuit breakers, direct trip actuators (DTAs). The DTA is typically used on model Type DB and DS breakers with either an Amptector or the newer Digitrip solid state overcurrent trip devices. The problem concemed a particular DTA model which incorporated a design change that used glued joints that held together 2
the core, permanent magnet and two pole piecas. The Browns Ferry nuclear plant informed Westinghouse that a DTA had failed after several cycles. Westinghouse sold Amptectors and Digitrip devices equipped with DTAs as safety-related components. Eaton Cutler-Hammer manufactures the DTAs as commercial grade components and supplies them to Westinghouse who dedicates them for use in safety-related applications.
The team reviewed the circumstances and held discussions with the Westinghouse staff and determined that the subject DTAs (with glued joints) were supplied only to the Browns Ferry and Vermont Yankee nuclear plants for use in safety-related breakers; that Westinghouse informed Browns Ferry and Vermont Yankee; and that the part:cular glued joint method is no longer employed by the manufacturer. The team concluded that Westinghouse appropriately informed the applicable licensees and satisfactorily executed its 10 CFR Part 21 responsibilities. As a result, this issue is considered closed.
Unresolved item 99900404/98-01-02 (Ocen) e During the January 1998 inspection, the team asked whether Westinghouse reconciled its original saismic qualification report for a model DB-50 problem at the Indian Point-2 (IP2) nuclear power plant to determine whether the reduction in mass or the extra winding on the trip pan spring could adversely affect the operation of the breaker. Westinghouse stated that there was conservatism in the design, but they had not yet reconciled the change to the qualification test report. They also stated that an upper limit of 31-ounces of force to raise the trip bar a quarter-inch is used as a limiting criterion at the factory and is published in vendor literature. However, Westinghouse did not publish a lower limit operational design criterion of 14-ounces which it uses intemally for manufacturing and refurbishment guidance at the factory to determine full operability. The Westinghouse staff stressed that the 14-ounce value is not a designed or analyzed number because they did not perform seismic testing to valida'.e the value. Westinghouse stated that it plans to perform seismic testing in the near future to substantiate its lower limit guidance and that it does not believe that either the removal of the trip pads from the trip bar or the extra winding on the trip pan spring is a major contributor to the failures of the IP2 DB-50 breakers. Westinghouse personnel informed the inspectors that it plans to perform seismic and evaluation tests to confirm that the trip pad removal and the extra winding on the trip pan spring do not contribute tc unreliable operation of DB type breakers, it is noted that although this potentially generic issue was identified to Westinghouse in October 1997, as of July 1,1998, it has not performed testing to substantiate the seismic adequacy of its circuit breaker functionality during a design basis event. Therefore, this matter will remain open and be reviewed during a future inspection.
Unresolved item 99900404/98-01-03 (Ocen)
On May 7,1997, in a letter to the NRC, Illinois Power Company's (IPC), stated that five arc chutes (part number 56F 17G02 for Westinghouse model 50DHP350,1200-ampere,4160-Vac circuit breakers) at its Chon Power Station had inadequate clearance between the rear are horn of the arcing contact and the lower edge of the molded case of the arc chute. In a letter dated July 2, 1997, IPC reported the issue to the NRC under 10 CFR Part 21.
3 s
Westinghouse Trip / Conference report Q&ES 98-0519, prepared April 6,1998, indicated that Westinghouse personnel traveled to the Clinton site and satisfactorily assembled the arc chutes assemblies with their associated breakers. An associated Westinp%use letter (RRS/DSE(98)-087, dated March 23,1998) indicated that the problem was that "whvmever attempted to assemble the arc chute to the breaker did not understand the assembly and had misconstrued the problem. The rear arc chute was never the problem, it was a bent seismic tie down bracket on the front of the arc chute." Subsequent to this inspection, the NRC inspectors contacted Clinton personnel and were informed that they did not agree with the Westinghouse conclusion. Additionally, Clinton personnel stated that when they first contacted Eaton Cutler Hammer, they stated they had never heard of a similar problem with the arc chute / arc hom assembly, but later (in person) indicated that they were aware of the problem. The Clinton personnel also stated that the Westinghouse vendor manual for the circuit breaker did not address the are chute problem. As a result, this issue will remain open and will be reviewed during a future inspection to assess the potentially generic implications of an inadequate vendor manual that does not address proper assembly.
Unresolved item 99900404/98-01-04 (Ocen)
The team reviewed the adequacy and effectiveness of records associated with circuit breaker issues,10 CFR 21.21.(a)(1) compliance, and related procedural requirements and guidance. The team also reviewed Westinghouse Energy Systems Business Unit (ESBU) Policy / Procedure 21.0,
" Identification and Reporting of Conditions Adverse to Safety," Revision 2, dated May 29,1997, and associated administrative instructions which included " retired
- Westinghouse Nuclear Technology Division's Administrative Manual (AM) F-3.0, " Potential Deviation or Failure to Comply (PD) Process," Revision 1, dated July 1,1995. The team revievi/ed several 10 CFR Part 21 related component issues identified by various licensees as potentially reportable.
The team identified previously that no references existed to PDs in the procedure. Westinghouse agreed that there was no section in the procedure that delineated the process used to handle PDs.
i During the last inspection the team noted that Westinghouse did not have any procedures or instructions which delineated the required actions for processing PDs. However during this inspection, the team was informed that Westinghouse had been using the retired Administrative Manual (AM) AM-F-3.0, for PD disposition. Additionally, the team determined that it had not reviewed a balanced sample of PDs and potential issues (Pls) to appropriately evaluate the effectiveness of the implementation of Westinghouse's 10 CFR Part 21 program. As a result, this matter will remain open and will be reviewed during a future NRC inspection.
Unresolved item 99900404/98-01-05 (Closed)
This issue concemed the failure of DS-416 circuit breakers as a result of the dry lubricant being removed from the mechanism during a refurbishment at a non-Westinghouse approved refurbishment facility. The team reviewed records, technical bulletins and conducted discussions with the Westinghouse staff and determined that the two technical DS circuit breaker issues were not related and were appropriately dispositioned. As a result, this issue is consioered closed.
4
-8 7-
3 INSPECTION FINDINGS AND OTHER COMMENTS 3.1 10 CFR Part 21 Prooram Imolementation
- a. Insoection Scone The team evaluated Westinghouse's 10 CFR Part 21 Program implementation by performing a review of information relating to identified concerns in domestic reactor plants such as recent ice condenser component problems.
- b. Observations and Findinos Ice basket segments (typically 12 feet in length) are coupled together using sheet metal screws arranged in pairs at 60 degrees intervals around the circumference of the ice basket. Twelve screws on each barket segment are installed at the coupling joint (24 screws total per coupling joint). NRC inspection reports 50-390/97-04 and 50-391/97-04 documented that broken ice basket sheet metal screws had been found in the ice melt tank at Watts Bar. Missing or defective screws at the intermediate or lower coupling joints could allow an ice basket to separate and become a missile hazard under blowdown loads created in a design basis loss of coolant accident. Therefore, defects and deviations from the design specifications for these screws represent a potential condition adverse to safety.
b.1 Evaluation Not Performed For Nonconformino Ice Basket Sheet Metal Screws On June 8,1995, Westinghouse received a faxed copy of Tennessee Valley Authority (TVA)
Central Laboratories Service Technical Report, " Watts Bar Nuclear Plant, ice Condenser Basket Screws," dated June 2,1995. This metallurgical report identified that failed and nonfailed ice basket sheet metal screws contained cracks at the screw thread roots. Among the potential causes for the failed screws identified in this report was that the pre-existing
" quench cracks" had been created during the heat treatment process step of manufacturing.
Additionally, this report identified measured core hardness values in excess of the 32-40 Rockwell C scale range identified in the original design specification 678956. A Watts Bar adverse condition report WBPER950246, closed on July 25,1995, listed corrective action No.
3 as completed, in which WA requested Westinghouse to evaluate the report. However, the NRC inspection team identified that no evaluation of the June 2,1995 report had been perfo,rmed by Westinghouse. Because this report identified conditions potentially adverse to safety that deviated from the original design specification for the sheet metal screws, the team concluded that an evaluat.on of these deviations as described in 10 CFR 21 was required.
Failure to perform an evaluation of the deviations has been identified as a violation of 10 CFR
! 21.21(a)(1). (Identified , s part of 99900404/98-02-01) b.2 incomotete Assessment of Broken and Missino Ice Basket Screws On June 16,1995, Westinghouse initiated Customer Blanket Order No. 94NNX-112361 pertaining to the disposition of a sheet metal screw issue at WA Watts Bar. This order required Westinghouse to evaluate field deviation report WATM-10356, dated June 15,1995, which l
5
=
identifi d that 162 screw heads and 32 screws had been found in the Le melt tank following ice basket loading and weighing operations completed in 1994. Westinghouse documented the assessment of this condition in a letter to TVA, dated June 22,1995.
This assessment concluded that the conditions identified with the screws did not affect the function or structural adequacy of the ice condenser containment at Watts Bar. The basis for this conclusion was: a statistical evaluation of the likelihood of one or more missing screws at a single coupling, a calculation of the minimum number of screws required at a basket coupling, and an evaluation of the potential projectile / missile hazards associated with a failed ice basket segment.
However, the team identified that the serviceability of the remaining screws, with respect to cracking identified in the TVA metallurgical report of June 2,1995, and the potential consequences of failed ice basket segments impacting the air handling units within the ice condenser, potentially introducing glycol into the sump affecting emergency core cooling systems, was not reviewed. As a result, the team considered the assessment incomplete for the areas noted. Failure to perform design reviews to complete the assessment of these areas is identified as a nonconformance of 10 CFR 50 Appendix B, Criteriun Ill. (99900404/98-02-02) b.3 Identification Of Conditions Adverse To Safety On or about June 1,1998. Westir.ghouse received Gelles Laboratories report, " Metallurgical Evaluation of Screws From Cook Power Plant Unit 1 Ice Baskets, dated May 1998. The report identified a number of potential causes for failed ice basket sheet metal screws including pre-existing cracks and hydrogen induced cracking, potentially related to failure to adequately bake the screws prior to plating, or the excessive hardness values measured for the failed sheet metal screws. Further, Westinghouse had initiated testing under contract with the D.C. Cook nuclear plant to evaluate the seceptability of in-service screws with these types of deviations from the original Westinghouse design specification 678956.
The team determined that the potential conditions adverse to safety (the failed screws at the D.C. Cook nuclear plant) had not been entered into Westinghouse's formal review program, as required by Section lli of ESBU procedure 21.0, nor had the condition been identified as a potential deviation in accordance with Westinghouse administrative instrucSons. Failure to follow ESBU 21.0 and enter this condition adverse to safety into the fornml review process, is a identified as a nonconformance of 10 CFR 50, Appendix B, Criterion V. ' WOO 404/98-02-04)
~
b.4 1,ack Of Material Soecifications For Screws In Ice Basket Desion Drawinas Purchase Order 546-AZG-221520-BEP, " Ice Basket- Ice Condenser," Change Notice No. 2, dated July 21,1975, identified design specifications 678956 and 952745 used in the construction of ice baskets for domestic reactors with ice condenser containments.
Westinghouse design specification 952745, dated January 10,1975, applied to all domestic Westinghouse designed plants other than the D.C. Cook and Sequoyah nuclear plants. Section 4 of this design specification required the material specifications for the ice basket components to be listed on the applicable ice basket design drawings.
6 l
The inspection team identified that " Ice Condenser Plant Ice Basket Detail and Assembly" drawings 1145E10 and 1191E57, Revision 7, dated February 1984, lacked material @
requirements, hardness values and strength levels for sheet metal screws required by original design specification 678956. Specifically the screws, described as piece No. 9, referred to
" Note Y" for material identification. However " Note Y" on the drawings lacked a description of the screw material required for fabrication, the required material hardness, and the required material strength levels.
This condition appeared to contribute to a May 30,1990, Licensee Event Report (LER),
submitted to the NRC by Duke Power Company (DPC), that identified installed ice condenser Eheet metal screws of improper material at the McGuire Nuclear Station (LER 369/90-07). DPC Problem Investigation Report 1-M90-0107, dated August 24,1990, proposed resolving the problem by initiating an editorial change to DPC drawing MCM 1201.17-0596, Sheet 001 (Westingho i drawing 1191E57, Revision D-8, dated December 1987, Sheet 1 of 3) to specify the ma' rial required for screws used in the ice basket couplings. The LER also concluded that the cause of the event was due to an installation deficiency because of imprcper material selection. The change consisted of revising " Note Y" on the drawing to specify the material specification for the screws which was omitted.
s Drawing Revision D-9, dated June 1%0, added "AISI 1022 carbon steel; minimum yield strength 130 ksi." Stainless steel screws of 10-16 type 410 material were installed in the bottom and intermediate couplings of the ice baskets instead of specially treated 10-32 AISI 1022 carbon steel screws, as specified by Westinghouse. The screws, installed in 1988, were identified in 1990 by DPC design engineering personnel to be a problem since the Westinghouse manufacturing drawing contained no reference as to the screw material; therefore, the wrong material was chosen when replacing the missing coupling screws.
The team also reviewed certificates of compliance for the vendors (LAMCO and Great Lakes Manufacturing Co.) that fabricated the ice baskets and screws respectively for dome: tic ice condenser plants. However, the certificates of compliance stated that they met the requirements of drawings 1145E10 and 1191E57. At the conclusion of the inspection, documentation for domestic ice condenser plants (other than D.C. Cook and Sequoyah) could not be located by Westinghouse which identified that the screws had been fabricated with serviceable material of the required strength and hardness. On June 17,1998, Westinghouse issued Corrective Action Request (CAR) CAR 98-1157 to document the lack of design requirements for the screws in the applicable drawings. Further, Westinghouse staff reportedly intended to locate the records which would demonstrate that the screws were fabricated using the appropriate materials or notify licensees to implement testing to confirm that the appropriate material was used. Failure to include design material and strength requirements for the screws is a nonconformance of 10 CFR 50, Appendix B, Criterion Ill. (99900404/98-02-03) b.5 Post-Insoection Information Sucolied by Westinohoust Following the inspection, Westinghouse supplied additionalinformation to the NRC staff regarding the ice condenser sheet metal screw issues identified by the team during the inspection. The information was documented in Westinghouse letter NSD-NRC-98-5728, dated June 26,1998.
7 l
Westinghouse is presently investigating the issues under its 10 CFR Part 21 program (Pl 98-016, dated June 17,1998), as delineated in Westinghouse procedure ESBU-21, and estimates completion by mid-August 1998. On June 24,1998, Westinghouse verbally notified all of its ice condenser containment customers of the issues identified during the inspection and the discussions between Westinghouse and the NRC staff. Westinghouse's action plan to address the issues will include the following:
- 1. Perform a metallurgical evaluation of 200 sheet metal screws supplied by D.C. Cook.
- 2. Review screw design including coatings and confirm the failure mode and root cause.
- 3. Review available reports, including previous metallurgical testing performed by Duke Power Company and TVA, and prepare technical position. ~
- 4. CAR 98-1170, dated June 17,1998, which will address the review of both 1995 WA ~
reports (June 2 and June 19) including the consequences of ejected ice baskets hitting air handling units. Based on the information obtained, a supplement to the June 22,1995 Westinghouse report will be prepared.
- 5. Cond',ct test program to establish load carrying capacity of screws, as applicable.
Westinghouse stated that preliminary information to date has identified quench cracks in only abou? 2 percent of the screws tested from D.C. Cook which corresponds to industry manufacturing data fer similar screws.
On June 17,1988, Westinghouse initiated CAR 98-1157 to review discrepancies between -
Westinghouse design specification 952745, dated January 10,1975, and drawings 1145E10 ,
and 1191E57, Revision 7, and determine if the correct screws were manufactured and supplied in accordance with Westinghouse specifications. The review will include a review of vendor purchase order documentation, a revision to the specification / drawing to include the required mechanical properties for the screws, and the issuance of a report.
3.2 Inadecuate Part 21 Proaram Recuirements During the January 1998 NRC inspection, the inspection team attempted to discuss an issue regatding a problem with a shunt trip device on a Westinghouse model DB-75 type circuit breaker installed at the Ginna nuclear power station. The problem was that the plunger in the ' '
shunt trip did not retum to the full out position following operation. When the licencee tested the assemblies in stock in the warehouse, the plunger would not retum to the full out position on 2 of the 10 assemblies. Westinghouse staff stated they had not been contacted by Ginna and as a result, the inspectors did not pursue discussions regarding the adequacy of Westinghouse's actions regarding 10 CFR Part 21 and closed the issue.
However, subsequent to the issuance of the report, the NRC was informed by the licensee, Rochester Gas and Electric Corporation (RGE), that on November 24,1997, RGE had -
contacted Westinghouse to discuss the issue. Westinght,use reviewed their records and 8
b
determined that RGE had informod them of the issue and provided the inspection team with correspondence that had been sent to another circuit breaker engineer and also to Westinghouse Nuclear Safety Department (NSD) personnel. The NSD personnel coordinate the execution of ESBU's program developed to implement the provisions of 10 CFR Part 21 and its implementing procedure, ESBU 21.0, and AM F-3.0.
The team determined that another retired ESBU Instruction Manual Procedure (ET B-1.0) states that its purpose is to define the process and responsibilities for assuring that a safety concem identified by, or referred to, Westinghouse Engineering Technology personnel are controlled and properly dispositioned in accordance with the QA program and to the criteria of i
10 CFR Part 50, Appendix B, and 10 CFR Part 21.
A review of the related records, procedures and discussions with the cognizant staff determined that RGE stsff called the Westinghouse engineer to inform him of the problem that had been experienced, ask whether similar occurrences had happened, and request guidance.
As a result, the Westinghouse engineer discussed the issue with another engineer, a senior circuit breaker shop technician and NSD staff to ask whether they were aware of similar problems. The issue was not pursued further since they were not aware of any similar problems. Additionally, although the NSD staff received a copy of the potential deviation, they did not formally document the issue in the PD/Pi process for disposition in accordance with Westinghouse procedures. This example of a potential deviation that was not adequately dispositioned or controlled by Westinghouse is a violation of 10 CFR Part 21. (Identified as part of 99900404/98-02-01)
- c. Conclusions The inspection team identified a violation and three nonconformances associated with failures to adequately implement 10 CFR 50, Appendix B, commitments and 10 CFR Part 21 requirements on issues pertaining to ice condenser sheet metal screws at the Watts Bar and D.C. Cook nuclear plants. The inspection team concluded that the issues represented an implementation breakdown of Westinghouse's QA and 10 CFR Part 21 program requirements.
3.3 Entrance and Exit Meetinas An entrance meeting was held on June 15,1998, in which the scope of the inspection was discussed with Westinghouse management and staff. On June 17,1998, an exit meeting was held with Westinghouse management and staff to discuss the inspection findings.
9
4 PERSONNEL CONTACTED Y The following represents a partiallist of persons contacted during the inspection: i Westinahouse Electric Comoany N. Liparuto, Manager G. Dillon, Manager J. Galembush, Senior Engineer C. Scrabis, Advanced Technical Engineer G. Rao, Senior Engineer R. Miller, Senior Engineer OPEN AND CLOSED ITEMS ITEMS OPENED 99900404/98-02-01 (VIO)
Failure to evaluate a June 1995 metallurgical report from TVA Watts Bar which identified quench cracked sheet metal screws, and the failure to process a potential deviation related to ice basket sheet metal screws in accordance with Westinghouse's Part 21 program.
99900404/98-02-02 (NON)
Failure to perform design reviews to complete the assessment of missing and broken sheet metal screws at TVA Watts Bar.
99900404/98-02-03 (NON) ,
Failure to follow Westinghouse procedure ESBU 21.0 and enter conditions adverse to quality into the formal review process.
99900404/98-02-04 (NON)
Failure to include on the design drawings design material and strength requirements for sheet metal screws.
ITEMS dLOSED 99900404/98-01-01 (URI)
Review to determine the reportability of problems in DTAs.
99900404/98-01-05 (URI)
DS-416 type circuit breaker problems.
10
ITEMS TO REMAIN OPEN 99900404/98-01-02 (URI)
Testing to be performed to substantiate seismic adequacy of circuit breaker functionality during design basis events.
99f,00404/98 01-03 (URI)
Potentialinadequate vendor manual narrative of assembly cautions.
99900404/03-Gi-04 (URI)
A seiL ction of Pls and PDs must be reviewed to assess the effectiveness of Westinghouse's Part 21 program.
I 11 l
___ _ __ _ _ _ - - _ - - _ _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - --~- --
NRC FoPM 338 U.S. NUCLEAR REGULATORY COMMISSION 1. REPORT NUMBER (249)
(Assigned by NRC, Add Vol., Supp., Rev,
"" "E 32o BIBLIOGRAPHIC DATA SHEET """""**"'4 (See astruc00ne on me rewrse)
- 2. TITLE AND SUBTITLE NUREG-0040 Vol. 22 No. 3 Ucensee Contractor and Vendor inspection Status Report 3. DATE REPORT PUBLISHED Quarterly Report uoNr.4 YEAR July-September 1998 l March 1999
- 5. AUTHOR (S)
- 6. TYPE OF REPORT Quarterly
- 7. PERIOD COVERED (incsusw Dates)
July - September 1998
- 8. enmon PERFORMING ORGANIZATION name andmadng adeess.) - NAME AND ADDRESS (#NRC, powde Dwmon, omce or Repan, ut Nucinar Reputatory commason, and madng acnesss, scontractor, Division of Inspection Program Mar gement Office of Nuclear Reactor Regulatic U.S. Nuclear Regulatory Commissi.
Washington, D.C. 20555-0001
- 9. and SPONSORING ORGANIZATION NAME AND ADORESS (#NRC, type 'same as aoomi # contractor, prowde NRC Dwsat omce orRegon, u3 NucsearReputatory commason mahng adsess)
Same as above
~'
to SUPPLEMENTARY NOTES
- 11. ABSTRACT (200 wonis or Arss)
This periodical covers the results of inspection performed by the NRC's Quality Assurance, Vendor Inspection and Maintenance Branch, that have been distributed to the inspected organizations during the period from July through September 1998.
- 12. KEY WORDS/DESCRIPTORS (but wcros or ptrases mat ud assat researchers a besang me report) 13 AVAH AssuTV STATEMENT Vendor inspection unlimited 14 SECURTTYCLAS$1FICATION (Thns Page) unclassified (Ths Report) unclassified
- 15. NUMBER OF PAGES
- 16. PRICE NRC FORM 335(249)
Ths form was electrorvcaHy produced by Elde Federal Forms inc,
i IL Printed
- on recycled paper Federal Recycling Program u
.__m_____.__ .-_-____-_
Z C
O 5 xx 3? $ 5 5]
BR g * =t kh .Z O ja 85'4 mi ~5%
58 $ 8
n 8 ki o gg "E
o Z
120555154486 1 1AN US NRC-0CIO DIV-INFORMATION MANAGEMENT TPS-POR-NUREG 2WFN-6E7 WASHINGTON DC 20555 3%
% o$8 h Ec m -
sog, zEas omE k EE n
er l
. .. _ _ _ _ _ _