ML20042E552
| ML20042E552 | |
| Person / Time | |
|---|---|
| Issue date: | 04/02/1990 |
| From: | Moist R, Potapovs U Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20042E548 | List: |
| References | |
| REF-QA-99900277 NUDOCS 9004230594 | |
| Download: ML20042E552 (24) | |
Text
.-. _ _
-~
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST.GRANBY, CONNECTICUT j
REPORT INSPECTION INSPECTION NO.: 99900277/90-01 DATES: Jan. 30 - Feb. 2. 1990 ON-SITE HOURS: 30 CORRESPONDENCE ADDRESS: The Rockbestos Company A Member of the Marmon Group Mr. S. Bond, President and General Manager Post Office Box Drawer 1102 New Haven, Connecticut 06504 1
ORGANIZATIONAL CONTACT: Mr. George Littlehales, QA Manager TELEPHONE NUMBER:
(203)772-2250 NUCLEAR INDUSTRY ACTIVITY:
Rockbestos conducts thermal aging, loss-of-coolant accident (LOCA) simulatiens, and post-LOCA sample evaluation testing for environ-mental qualification (EQ) as wel! ss flame testing for their nuclear adverse service Class 1E (safety-related) cables in the engineering laboratory at their New Haven, Connecticut factory. Nuclear service cable is manufactured at both the New Haven and East Granby, Connecticut, plants, with the majority of it being made at East Granby. Approximately 40 percent of Rockbestos' business is from nuclear utilities, i
4 ASSIGNED INSPECTOR: /dc4 d d 3 /// M J/
J[#9/90
- f. N. Moist,' Reagtive Inspection Section No. 2 Date (RIS-2)
OTHERINSPECTOR(S):
S. Alexander, RIS-2 M. Jacob s, Sandia National Laboratories (SNL) f APPROVED BY:
k
%%e 4 2-90
(
U. Potapovs, Chief, RIl-2, Vendor Inspection Branch Date 1
INSPECTION BASES AND SCOPE:
A.
BASES:
10 CFR Part 21 and 10 CFR Part 50, Appendix B P.
SCOPE:
Follow-up on 10 CFR Part 21 notification, follow-up on nonconfor-inances, open items and unresolved items from previous inspections, I
evaluation of quality assurance (QA) program implementation'in selected areas, and follow-up on an allegation.
1 PLANT SITE APPLICABILITY:
SanOnofre1,2and3(50-206/361/362); Haddam Neck (50-213);. Nine Mile Point 1 and 2-(50-220/410); Dresden 2 and 3 (50-237/249);
Millstone 1, 2 and 3 (50-245/336/423); Turkey Point 3 and 4 (50-250/251);
(continuedonnextpage)
.j 9004230594 900412 PDR OA999 EMVROCK 99900277 PDC x
-_,.,--,.-a n,
,.+.a
,.-,,s,.+,
n o,,w
1 ORGANIZATION: THE ROCKBESTOS COMPANY l
NEW HAVEN AND EAST GRANBY, CONNECTICVT
]
4 I
- REPORT INSPECTION l
NO.: 99900277/90-01 RESULTS:
PAGE 2 of 24 l
l PLANT SITE APPLICABILITY:
(continued) Palisades (50-255); Monticello (50-263);- Quad Cities '2 (50-265); ' Point Beach 1 and 2 (59-266/301); Peach Bottom 2 and 3 (50-277/278); Prairie Island 1 and 2 (50-282/306); Indian Pilg(rim (50-293); Zion 1 and 2 (50-295/304);50-309); Arkansas Kewaunee Point 3 (50-286);
(50-305); Maine Yankee Calvert Cliffs 1 and 2 (50-317/318); Fitzpatrick (50-333); St. Lucie 1 and 2 (50-335/389); McGuire 1 (50-369); LaSalle 1 and 2 (50-373/374); Medical College of Hanover (50-377); Catawba 1 and 2 (50-413/414); andWNP-3(50-508).
l A.
VIOLATIONS:
None B.
NONCONFORMANCES:
1.
Contrary to the requirements of Criterion 111 of Appendix B to 10 CFR Part 50, "Desi9n Control," and Section 1 of Rockbestos Procedure S-3, " Design Control Systems," the ' implementation of the Rockbestos program for approval and release of design data was inadequate in that Rockbestos provided cable ampacities and ampacity calculations (nuclear plant.dasign data) that were not formally reviewed and approved, and' which were prepared Using a computer code that was not formally reviewed and approved for Firezone R cable sold to the Haddam Neck Plant.(99900277/90-01-01) 2.
Contrary to the requirements of ' Criterion V of Appendix B to 2
" Instructions, Procedures and Drawings,"
Rockbestos test procedure RTP-0050 and TG test _ data sheets specified by RTP-0050 were not appropriate to the circumstances because RTP '0050 had been revised improperly and TDEs had~ not been revised to reflect current practice prescribed in_ RTP-0050.
(99900277/90-01-02) 3.
Contrary to the requirements of Criterion XI of Appendix B to 10 CFR Part 50, " Test Control," Rockbestos Qualification Report QR-6802, Revision 1, stated that certain post-test insulation resistance values were passing when they were (1) taken with an instrument incapable of valid measurements of the required magnitude, (2) in some cases not shown to meet the acceptance.
criterion and (3) in other cases actually below the required values.
(99900277/90-01-03) 0 l
wewua v
e up
=
- w is--
m a..---
9
. ah
---r J
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 3 of 24 l
l 4.
Contrary to the requirements of Criterion XVI of Appendix B to i
10 CFR Part 50, " Corrective Action," Rockbestos measures estab-lished to correct certain conditions adverse to-quality that were identified as nonconformances during previous NRC inspec-tions mere not fully effective in that some of the same types -
of nonconformances were repeated and identified in this inspec-l tion.
(99900277/90-01-04)
C.
UNRESOLVED ITEMS:
1.
Control of Off-Standards Reports I
During the conduct of this inspection, the inspectors identified weaknesses in the system used at the Rockbestos East Granby, Connecticut, plant to maintain control and accountability of Off-1 Standard Reports.
This issue is discussed in detail in section E.8 of this report.
(99900277/90-01-05) 2.
Activation Energy for Siliccne Rubber The inspectors reviewed preliminary data from the new aging and elongation testing program in progress at Rockbestos to redeter-mine the activation energy for their KS-500 formulation of silicone rubber.
The dat.a obtained thus far indicate that the activation energy may turn out to be lower than expected which could impact the qualified life of installed safety-related 1
cable. Therefore, this issue, discussed in section E.2.a of this report cemains unresolved pending completion of-the new test series and recalculatien of the KS-500 activiation energy.
(99900277/90-01-06)
D.
STATUS OF PREVIOUS INSPECTION FINDINGS:
1.
(Closed) Nonconformance 84-02, item.B.1:
(East Granby Plant)
Contrary to Criterion V of Appendix B to 10 CFR Part 50 and Sections 4 and 5 of the Rockbestos CaTpany's-Quality Procedure Q-10A, dated April 4,
1984, traceabiltty could not be demon-strated for a reel of single cunductor insulated wire, (Shop Order 90205-01).
The reels of cable which utilized the above mentioned wire for inclusion as one conductor of a two conductor cable are:
4A-536/034964, 4A-845/D349643, and 4A-846/034964A.
1 5
fhh
1 ORGANIZATION: THE ROCKBESTOS COMPANY-NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 4 of 24 s
The NRC inspectors reviewed records, notes and the Rockbestos response to the USNRC con'
.ing this nonconformance.
All reels of completed cable which sorated the suspect single conduc-tor were identified with rm mance (OSR) tags and were later 4
~
scrapped.
The inspectors ev1esd Rockbestos Quality Procedure Q-10A, Revision 6, dated July 7, 1989, which established the..
i method of traceability for cabling.
The. inspectors verified traceability of cabling by reviewing six purchase orders and associated supporting data such as Certified Test Reports, Test.
Data A&B (TDA and TOB) sheets, shop order numbers and reel numbers.
2.
(Closed) Nonconformance 84-02, Item B.2:
(East Granby' Plant) q Contrary to Criterion V of Appendix B to-10 CFR Part 50 - and Section 13 of the Rockbestos Company's Quality. Manual for tool and gage control, six timer units used for rework processes and one unit used for hot modulus testing had not been calibrated as required.
The inspectors reviewed Rockbestos Quality Manual, Section 13, dated March 18, 1988,- titled " Tool and Gage Control." Four-timer-units from the insulation rework area were selected by the inspector to verify implementation of the QA manual.
All units had current calibration stickers on the face of the units for the curing temperature, footage counter and curing time measurements.
Unit 812 did not have a calibration sticker for the curing time measurement; however, a red sticker was on ' the unit which -
displayed do not use until unit is calibrated..
The NRC inspectors also reviewed the calibration records of all four units to verify that information on the calibration sticker was correct.
3.
(Closed) Nonconformance 85-03, Item B.1:
(East Grandby. Plant)
Contrary to the commitment contained in the Rockbestos corrective action response letter, dated November-23,1982, with respect to paragraph E.1 in - the NRC Inspection Report No. 99900277/82-02, the procedure for the calibration of irradiation units-had not been followed.
This is evidenced by the lack of calibration stickers on the high voltrge resistor string. and the direct current tachometer digital readout, as required by the
" Procedure for Calibration of Irradiation Units."
}
T E
m i -
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT
~ INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 5 of 24 a
I The NRC inspector!s reviewed two high volthge resistor strings and verified that each had a current calibration sticker attached.
The calibration records were also reviewed to verify that the information on the calibration stickers was correct, i
4.
(Closed) Nonconfonunce 83-03, item B 2:
(East Granby Plant)
Contrary to Criterion V of Appendix B to 10 CFR Part 50. and paragraph 6 of Part C of Quality Procedere Q-19, dated May 15, i
1982, the Chief Technician or desi reviewed all-completed Test Data Sheets "B" (TDBs) gnee had not as evidenced by the lack of signature or initial and date.
{
The NRC-inspectors selected five current TDAs and verified that the test results were evaluated by the QA Analyst or designee to assure the test requirements had been met.
The NRC insectors also revieued five current TDBs and verified that the - Chief Technician or designee had reviewed the data sheets as evidenced by initials.
5.
(Closed) Nonconformance 83-03, Item B.3:
(East Granbj Hint)
Contrary to Criterion V of Appendix B to 10 CFR Part 50 and paragraph 4 and its subparagraphs b and c of Quality-Procedure Q-6A, dated January 12, 1983, the Quality Analyst Supervisor or designee had not revised TDB = G323AK, dated March 24, 1983, for Shop Order 81396-03, Reel No. 54093, which was completed June 29,
'i 1983.
The lack of revision is evidenced by an elongation requirement change responsible person o(r time. lined through) without identification of the i
The NRC inspectors reviewed Quality Procedure Q-6, Revision 12, dated December 19, 1988, titled " Procedure for Control of TDAs 4
and TDBs."
Paragraph 4.i of Q-6_ states in part "for Nuclear Orders, any change to requirements must be reflected in a new revision of the TDB. - Also, it is permissible for the laboratory to write in results of additional tests not mark certain tests N/A where necessary." part of the TDB or to The NRC inspectors verified implementation of the above by-selecting five purchase orders and associated TDBs.
ORGANIZATION:
THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 6 cf 24 1
6.
(Closed) Nonconformance 83-03, Item B.4:
(East Granby Plant) j Contrary to Criterion V of Appendix B to 10 CFR Part 50 and paragraph Ill and its subparagraphs 1 and 2 of Quality Procedure Q-9, dated June 7,1982, Quality Control Instruction No.114A, dated, August 10. -1982, did not contain identification of the individuals preparing and reviewing and/or-approving the instruction.
The NRC inspectors reviewed. Quality Procedure Q-9, Revision 9, dated, January 20, 1989,. titled " Document Control Procedure."'
This procedure lists different classes of documents and who is 6
responsible for approving these documents.
Also, the procedure l
requires the identification of the individuals preparing, reviewing and approving the documents via initials or signature.
Several documents were reviewed to verify the implementation of this procedure at both the East Granby plant and the New Haven plant during the course of this inspe: tion.
7.
(Closed) Nonconformance 83-03,- Item B.5:
(East Granby Plant)
Contrary to Criterion V of Appendix B to 10. CFR Part 50 and i
Ebasco Services Incorporated's letter dated, ' April 29, 1982, referencing the D60-01 cable insulation restoration procedure, records were not available which would ' substantiate that' the required thermocouple monitoring had been performed.
The NRC inspectors reviewed an internal memorandum dated, March 5, 1984, subject " Supporting Data 060-01 Test," laboratory Data Sheet dated, March 2, 1984, and a circular temperature chart -dated, March 1, 1984. This was a special test that was performed to show r'
objective evidence that heating approximately 10,000 feet D60-01 in an oven for 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> at 225'F would bring cable of 2/016 to 225'F.
l Temperature was monitored at top, middle, and bottom of reel at 32 and 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> with a Technoterm 5501 Pyrometer calibrated March 11, 1983.
The test data verified that the inner layers-on the cable reel reached 225* uniformly throughout the reel.
8.
(Closed) Nonconformance 83-04, Item B.1:
(NewHaven)
Contrary to the requirements of Criterion XVII of Appendix B to 10 CFR Part 50 and Section 8.3 of IEEE-323-1974, Rockbestos had not maintained sufficient records to furnish evidence that activities affecting quality were performed.
Specifically,
,. =
,n
-4
ORGANIZATION: THE ROCKBESTOL COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION.
NO.: 99900277/90-01 RESULTS:
PAGE 7 of 24 Rockbestos did not have the supporting data for test report F-C3798 entitled, " Qualification Tests of Electrical Cables Under i
Simulated Reactor Containment Service Conditions including l
Loss-of-Coalant-Accident While Electrically Energized," performed by the Franklin Research Center (FRC) for Rockbestos.
'(
During the NRC inspection of January 7-10, 1985, the NRC.
inspectors reviewed the latest _ revisions of-quality assurance 1
documents, procedures, inspection plans, etc., for compliance with regulatory requirements and verified by observation of test i
and QA activities and review of associated records that controls were being implemented except as noted 'in the nonconformance identified during that inspection.
Subsequently, these nonconformances were closed out during an NRC inspection conducted on April 22-24, May 22-24 and September 11-12, 1985.
During this inspection, the NRC inspectors reviewed' the latest revisions of selected QA documents, procedures,-inspection plans, qualification test reports and associated records and verified that controls were being implemented.
9.
(Closed) Nonconformance 83-04, Item B.2:
(NewHaven)
Contrary to the requirements of Criterion XI, Appendix B to 10 CFR Part 50, Rockbestos did not evaluate or' document the evaluation of 4 eable failures which occurred during the F-C3798 qualification test and 11 cable failures that occurred during the F-C3859-2 qualification test.-
Rockbestos concluded in both of the above instances that it is virtually impossible to ensure that damage to the cable samples can be eliminated at the point where they pass through the flange penetrations.
Additionally, both the expansion characteristics of the insulation and the thermal characteristics of'the jacket material exacerbate the penetration' problem.
During the requali-fication program, the NRC inspectors verified ' that Rockbestos evaluates and docurents' cable failures.
Also, several follow-up j
items in this report confirm that Rockbestos is evaluating and documenting cable failures.
i 10.
(Closed) Nonconformance 83-02, Item B.1, Example. (2) (New Haven)
Contrary to the requirement of Criterion 11 of Appendix ~ B to 10 -
CFR Part 50, Rockbestos failed to establish and implement an adequate QA program for its safety-related equipment qualifi-cation (F9) testing ef#~+.
Tnh is evidenced by the number of
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 8 of 24 i
i generic deficiencies identified during this' inspection.- Examples
)
are:
(1) numerous calibration system deficiencies; (2) use of inadequate test instrumentation; (3) the failure of engineering to develop, review, and approveJ test plans; (4) the failure of engineering to specify and describe test requirements; (5) the failure of engir.eering to identify and evaluate test nonconfor-f mances, variations, and deviations during testing and to document i
the same in test reports; (6) the failure of eng)ineering to perform adequate evaluations of test results; (7 technical inconsistencies between raw test data and final EQ test reports; and (8) the failure of the QA and control organization to audit and monitor EQ testing.
The instrument in question was a LOCA chemical spray flowmeter-as stated in NRC Inspection Report 85-01 D.7.C.(2). ' The maximum error recorded during before-use calibration of the flowmeter was l
6.3 percent, exceeding.the 5 percent allowed by paragraph 13.2.5 of the Rockbestos QAM.
A staff memorandum, dated December 11, 1984, indicated that the flowmeter accuracy specification should be changed to 120 percent.
However, procedures had. not been-revised accordingly.
A review of Section 13.2.6 of the Rockbestos QAM during this inspection indicated it now required that equipment for the Qualification and Test Department _ "shall not have an error greater than plus or minus 5 percent except as specifically excepted by the Technical Manual (TM) or applicable calibration procedures."
Section 4.2.1(b) of the Rockbestos TM states "the '
percent deviation from the standard for spray flowmeters shall j
not exceed 20 percent."
x 11.
(Closed)UnresolvedItem83-04,ItemC Rockbestos had used single conductor or single twisted pair test results to establish qualification for a vastly different cable, such as an 80-conductor (80/C) cable.
This was -a questionable use of similarity analy(sis.
The NRC inspector observed that requalification program RP) specimens now included 7/C and 2/C completed cable section samples as representatives of multicon-ductor (M/C) types.
Results of requalification testing indicated satisfactory performance of both the _ jacketed, multiconductor cabie
.id unjacketed single conductors of the Firewall III and Firewall SIS types of XLPE insulated wire respectively, i
,a, w
ww-,,-
,w-e-m-e-e--
r v
--a e----
,.~-w w
a,,, -,
1
+
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION 1
NO.: 99900277/90-01 RESULTsi PAGE 9 of 24 12.
(0 pen)bnresolvedItem83-03,ItemC The NRC inspector observed that.the " Oil Aged" section of work-sheet EGTL No. 4 indicated that calculations had been indicated:
complete by being stamped and dated.
However, data had not been entered in the " Average" blocks of the table.
Rockbestos stated that the spaces 'were not completed because additional data was.
required prior to perfor:ning the calculation.
Apparently, there were no requirements that addressed stamping / dating-of in-process test calculations.
This item was not addressed during this inspection.-
13.
(Closed) Unresolved Item 82-02, Item C. 2 (East Granby Plant)
It was not apparent that brazed connection of conductors had been measured, or a procedure existed for accomplishing the task. The validity of retests was questionable for accelerated water absorption and flamability properties.
During NRC inspection a
83-03, the NRC inspector deferred action on this item until staff technical evaluation was complete. - During this inspection, the NRC inspectors reviewad Rockbestos' brazing procedure.. dated March 1,- 1985.
The procedure delineated the steps.to be.per-
).
formed during this operation and the inspection characteristics to be inspected.
No brazing of conductors in the stranding area was being performed during this visit. to verify implementation.
However, the procedure appeared to be adequate to, perform this operation.
14.
(Closed) Follow-upItem81-01 This item involved the evaluation of the original qualification l
testing of the RSS-6-100 series coaxial cable.
In May 1981 General Atomics Company (now Sorrento Electronics Division of GA Technologies) submitted a 10 CFR Part.21 report on the failure of Rockbestos RSS-6-104 coaxial cable to. maintain sufficient insula-tion resistance during' main-steam-line-break- (MSLB ; and oven-Rockbestos responded to this failure.by (a) ) redesigning.
tests.
the construction of the coaxial and triaxial cable Jseries; (b) performing qualification-type tests o'n the " newer second and third generation coaxial and tricxial cables; (c) recognizing that the redesign causes-an additional technical concern for the i
larger diameter. members of the RSS-6-100 series cables (for example-RSS-6-109); and (d) performing-specialized qualification tests for customers on the RSS-6-109 cables.to alleviate qualifi-cation concerns.
This item was examined extensively during the l'
l I
'NY j
ORGANIZATION: THE ROCKBESTO:
NEW HAVEN AND LdT GRANBY, CONNECTICUT REPORT INSPECTION l
NO.: 99900277/90-01 RESULTS:
PAGE 10 of 24 l
inspection of June 6-10, 1983, and a number of questions concerning whether qualification had been demonstrated by the Rockbestos effort were raised.
These concerns were outlined in-inspection report 83-01 and remained open pending the NRC evalua-tion of the Rockbestos requalification test program being
~
conducted in response to inspection reports 83-01, 83-02,' and 83-04.
The results of the Rockbestos requalification program on coaxial, twinaxial and triaxial cable of the RSS-6-100 series were documented in Rockbestos-qualification report QR-6802.
QR-6802 indicates that the samples carried nominal / rated voltage and current successfully for 28 days of loss-of-coolant-accident (LOCA) and post-LOCA simulation exposure intended to ' follow a IEEE 323-1974 LOCA profile.
The published values of. insulation resistance during and af ter LOCA exposure, for these cables, while not necessarily unacceptable for many qualified applica-tions, may still be too low with respect to the values needed for continuous satisfactory performance of the GA high range radiation monitor under some design basis accident conditions as first discussed in the GA 10 CFR Part 21 report.
During this inspection, the NRC inspectors reviewed QR-6802, raised technical concerns described in section E of this. report, and identified the nonconformance cited in paragraph B.3 of this report.
15.
(Closed) Follow-upItemIR 99900277/86-01 paragraph E.4' The test failures of the triaxial cabla sample C2 and one conduc-I tor of the twinaxial cable sample D1 were further investigated by l
Rockbestos after the-86-01 NRC inspection.
T_he test report l
documents the failures and the failure analysis.
The failure
=
analysis concluded that extrusion of the cables through LOCA chamber penetrations caused severe stress on the-cables and probable mechanical damage.
The information in the test report was considered sufficient for -utilities to determine the suitability of this cable for their plant applications.
16.
(Closed) Follow-up Item IR 99900277/85-02 paragraph E.1.e.(4) l The analysis of chemically cross-linked polyethylene-insulated Firewall SIS sample failures had concluded that failures were attributable to handling damage which justified further testing.
According to records, Rockbestos repaired the handling-damaged sections of faulted samples and reperformed the voltage withstand test with satisfactory results.
l l
t t
N
-e, 4
~o.
~
ORGANIZATf0N:
THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 11 of 24 17.
(Closed) Follow-up Item IR 99900277/85-02, Paragraph E.2.b The NRC inspectors ha'd observed an incorrectly set power supply voltage prior to a LOCA test.
Rockbestos' normal pretest checks had not detected the problem.
The inspectors recommended that a pretest check procedure be instituted ' to prevent such problems.
The NRC inspector interviewed the test technician during this inspection and found that Rockbestos has made a practice of tagging each loading circuit with sample numbers and voltage and current requirements to preclude similar problems.
18.
(Closed) Follow-upItemIR 99900277/85-02 Paragraph E.4.b During inspection 85-02, the NRC team questioned the Rockbestos analysis showing that LOCA test conditions envelope a 2-minute main steam line break to -381'F based on a thermal lag heat transfer analysis.
The analysis has not been modified by Rockbestos.
They do not include the analysis in test reports when the actual test data is pre ided.
Rockbestos stated that if a customer requests analysis ivr a higher temperature than -was tested, Rockbestos provides information that includes Arrhenius equivalent degradation analysis, justification of _ peak tempera-ture based on short circuit rating and a report with information with which the customer can calculate insulation resistance at a given temperature above the LOCA test temperature.
19.
(Closed) Follow-upItem,IR 99900277/85-01 Paragraph E.11 During a previous inspection, the NRC inspectors identified some elements on the Rockestos personnel QA/QC. qualification _. record sheets of two test. personnel that were ' marked with arrows proceeding from elements above in the block for instructors' 4.
initials and date of completion and were not indicated as completed as required by quality procedure (QP) Q-25B, Revision "0", dated April 26, 1983. During this inspection, the inspector reviewed Rockbestos QP Q-25B, Revision 2 dated March 20, 1986.
Paragraph 2 of Q-25B now states, in part: "If each item is not separately initialed or signed, applicability of initials / signa-tures shall be clearly shown by suitable brackets, arrows or -
similar means."
20.
(Closed) Follow-upItem,IR 99900277/85-01 Paragraph E.12 During a previous inspection, Rockbestos could not produce the record of the survey reportedly performed by Rockbestos on the calibration laboratory they were using at that time. During this s
- - - - ~
n.
ORGANfZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 12 of 24 inspection, the NRC inspectors reviewed a Rockbestos survey. that had been completed on September 20,1988 on their current cali-bration laboratory, EIL Instrument, Incorporated.
No anomalies were identified during this review.
7 E.
OTHER FINDINGS OR C01;MENTS 1.
Rockbestos 10 CFR Part'21 Notification-on Shielded Cable A 10 CFR Part 21 report was submitted to the NRC by Rockbestos on November 6,
1989 concerning a Rockbestos electrical cable supplied to Consumers Fower Company for use in the Palisades '
i Nuclear Plant.
The Consumers Power purchase order required the special purpose, 600V, 90'C, shielded, 2/C, #12 AWG~ control cable to have a #34 AWG tinned copper braided shield, but it was supplied with a #34 AWG stainless steel braided shield instead.
The Rockbestos manufacturing product specification (MPS) for this order (revision A), dated -April 20, 1989, prepared by the specification engineer and approved by the supervisor of specifi-cation engineering, specified the shield material as iype 302 stainless steel.
To prevent recurrence, ' Rockbestos has now developed a checklist which is used to check the MPS against the requirements of the shop order, Rockbestos Order Form 441, which is supp0 sed to include all purchase order ~ requirements. 'The NRC inspector reviewed the checklist and determined that it would be adequate to verify that the - correct technical requirements are translated into the MPS. A Rockbestos internal memorandum, dated November 6, 1989, stated that Consumers Power had performed a calculation with which they determined that a problem would not exist with installed circuits 200 feet or less in ler,:*
Consumers Power reportedly believed that -all installet' cim;,
aere less than 200 ft and that no cable would have to be removec.
2.
Notifications to the NRC regarding Rockbestos Silicone Rubber insulated cables The Rockbestos Company notified the NRC on October 2,1989 of a l
concern relating to activation energy and ampacity values for Rockbestos silicone rubber (SR) insulated cable.
EcoTech/ RAM-Q Industries notified the NRC on October 9 1989 of their concerns relating to findings from their source v,erification and perfor-mance-based audit of Rockbestos which was performed for Northeast L
Utilities (NU).
On January 2,1990, NU notified the NRC of a related concern regarding -Rockbestos "Firezone-R,"
silicone-rvbber (SR)-insulated cable.
All the notifications pertained to i
9
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 13 of 24 the same technical concerns, SR activation energy and Firezone-R ampacity, and related QA concerns. Specifically:
a.
Silicone Rub'oer Activation Energy issue:
Rockbestos conducted an accelerated thermal aging program on their KS-500 formulation of SR in the mid 1970's.
The elongation test data generated from that program was used-to develop time-temperature curves at 200*C,190*C and 180*C, The original data supporting the plots of the curves is no longer available, becase it was apparently lost or inadver-tently discarded.
During the audit of Rockbestos mentioned above, EcoTech/
RAM-Q reviewed the data that Rockbestos had taken from'the curves to calculate the SR activation energy and found problems with (1) the basis for-the curves, (2)' the basis for selection of the end point line at 50% retained elonga-l tion, and (3) with the proximity of the points actually chosen for the regression analysis to.the actual intersec-tion of the time-temperature curves with the percent-l retained elo yation lines.
The activation energy value calculated by Rockbestos at 50%
t retained elongation was 1.805 ev.
However.-in cddition to the data supporting the Arrhenius plots not being available, the intersection points of the 190'C and 180'C curves with the selected 50% retained elongation. line were-not bounded by actual data points used to plot the curves, but were on L
sectors of those curves that were extrapolated from plotted I
elongation test results.
Furthermore, the points Rockbestos selected for the regression analysis did not exactly l
coincide with the curve-line intersections, but were skewed, resulting in an erroneously higher calculated activation energy than the actual intersections would have yielded.
f According to the Ecotech audit report, their own calcula-tions performed for comparison, yielded a value of 1.718 ev, at 50% retained elongation.
The lower value of activation energy reduces the qualified life of 125'C-rated SR from 95.06 as originally calculated, to 69.98 years at-125'C.
EcoTech/ RAM-Q then calculated activation energy at 55%
retained elongation in order to obtain intersection-points on all three curves bounded by actual data points and obtained a value of 1.622 av.
Using this lower activation-
.n m
.,-.,_.._..-.s
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.
99900277/90-01 RESULTS:
PAGE 14 of 24 energy, the Arrhenius formula yields a qualified life for SR further reduced. to 49.77 years for the rated ambient temperature.
During this inspection the NRC inspectors performed regres-sion analysis using data from the same plot.
At 55%
retained elongation, the NRC inspectors calculated the SR activation energy at 1.624 ev. - The result at 60% retained elongation was 1.611 ev,1.753 ev at 65 % and 2.092 ev-at 70% retained elongation.
These figures give a reasonable generic activation energy for the KS-500 compound in the. range of 1.61-1.62 ev.
Using this activation energy-value gives an acceptable qualified life for Firewall SR but the resultant qualified life f or Firezone R is only about 30 years because of the limited amount of accelerated theriral aging done on this type of cable. Nevertheless, the amount of actual aging. degradation that would be suffered by SR insulation in the Firezone R.at 4
90*C in an actual installation is insignificant relative to the thermal capability of SR. Less than 5% reduction in the elongationofSRwouldbeexpectedin40 years-at90*C(90*C because Firezone is arbitrarily derated to 90*C for it to be compatible with other cables with which it could share raceways). Accordingly, it would be appropriate to select a higher retention of elongation figure for calculating the activation energy for the Firezone R, thus yielding values higher than 1.62 ev, and hence obtaining longer qualified-life.
Calculations using the existing plots indicate that it is likely that there can be sufficient demonstrated qualified life for most plant applications of both the Firewall SR and the Firezone R if the data from which they are constructed are presumed to be correct.
However, in order to confirm the validity of the original curves used 'to calculate 4
activation energy for thermal aging purposes in. support of environmental qualification (EQ) of Rockbestos Firewall SR-and Firezone R cables insulated witt. the KS-500 compound, Rockbestos is conducting another aging program of the'KS-500 formulation SR compound which is scheduled to be completed by mid-April of 1990. Some initial data from the retests in progress appears to indicate a fairly significant deviation from previous data.
Dending completion of this aging and G
1 i
N
,e
.c r
,,-r--
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 15 of 24 z
testing program, _ and further review during a future NRC inspection, this issue is designated as Unresolved Item 99900277/90-01-06.
b.
Firezone R Ampacity Issue:
Rockbestos.had a contract with NU to provide them with some
- 4/0 AWG Firezone-R cable in accordance with Rockbestos sales specification RSS-5144.
RSS-5-144 provided ampacity values for commonly used cable sizes but act for #4/0 AWG.
Upon request for this information by NV, Rockbestos calcu-lated a value for NU using a computer code developed in-house and based on the standard methodology, and provided this value to NU. During the Ecotech audit mentioned above, it was determined that ampacity calculation had not been-reviewed and approved in accordance with design and document control QA procedures.
During this inspection, Rockbestos pointed out that standard ampacity tables such as those promulgated in the National Electric Code (NEC) can be used for most cables.
- However, due to its special application and design, Firezone R cable does not conform to the model used to calculate the nominal values in the NEC. Among other substantial differences from conventional cable design to enhance fire environment oper-ability, Firezone R uses a combination copper and nickel.
conductor.
Therefore, this is considered-special~ purpose cable and ampacity needs to be specially calculated.
In order to determine the accuracy of the Rockbestos figures and evaluate their calculation method, the NRC. reviewed a manual calculation of. the ampacity for 4/0 Firezone R cable-using the standard formula, but with the parameters and values unique. to Firezone R.-
The manual calculation _ gave the same answer as the Rockbestos compcter code that was the basis for the ampacity figure provided to NU.
The computer code had also been used to calculate the ampacities listed' in RSS-5-144 for the other available sizes.
The calculation input and methodology were also reviewed and no deficiencies were noted.
Thus, the ampacity calculations appear to be l
correct based on agreement between the computer code and the-manual calculation results.
Although no technical deficiencies were identified with either the methodology or the results of the ampacity calculations, nonconformances to QA requirements were
)
mvw y
ww--
p.-
g w
e
a 1
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT j
REPORT INSPECTION NO : 99900277/90-01 RESULTS:
PAGE 16 of 24 identified.
Rockbestos did not consider ampacity to be a cable design parameter, although it is a function of cable design.
However, vendor-provided ampacity-values
~
based on cable design and standard installation factors are used as nuclear design information by plant engineers in conjunction with plant-specific installation-related factors-that determine the final installed ampacity and are a cable sizing consideration.
Therefore, ampacity must be treated as other safety-related design information and the calcula-tion of ampacity is an activity affecting quality, requiring QA controle.
The failure to properly anC lormally document, review and document, review approve the computer code, and to formally (by someone other and approve the individual calculations than the person directly responsible) constituted a
nonconformance with respect to Rockbestos QA procedures on design and document control and Criterion III of Appendix B to 10 CFR Part 50.
This is designated as Nonconformance 99900277/90-01-01.
3.
Quality Assurance Program 4
The NRC -inspection team reviewed certain aspects of the implemen-tation of Rockbestos' QA program by evaluating corrective actions taken by them on previously-identified NRC findings at the New Haven and East Granby plants and also the findings identified by i
EcoTech/ RAM-Q during their audit for NU.
Also, the inspectors 1
reviewed current quality procedures, qualification test reports, qualification test plans, training-records, ~ audit reports, test data sheets, calibration records, engineering calculations, and certificates of conformance.
The inspectors found ~ that the portions of the QA program reviewed were being implemented in accordance with Appendix B to 10 CFR Part 50 with the exceptions noted.
The types of findings; identified in Section-B of this-report such as test control, design control, and -failure to follow procedures are repetitive of findings identified i r, previous NRC inspections, showing lack of adequate corrective action and are designated Nonconformance 99900277/90-01-04.
4 4.
Review of Test Report for Rockbestos Silicone Rubber KS-550 The NRC reviewed quelification report QR-8802 for silicone rubber KS-550 insulation and supporting data. Standard Rockbestos proce-dures and test parameters were used in the test.
4
,,-..r
=
.--.I
' ORGANIZATION: THE ROCKBESTOS COMPANY t
HEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 17 of 24 1
Iwanal aging consisted of 1400 hr. at 180*C' to give a qualified life of 40 yr. at 125'C.
The testing used 1-conductor, #14 AWG cables, one aged and one unaged.
The test plan called for a i
random selection from each of the two aged and two thermally-unaged specimens. Following irradiation of the 4 samples, one of the thermally-unaged specimens failed electrical tests. The only indication of mechanical damage from the visual observations was small indentations at 3,
9, and 12 feet along the sample.
Rockbestos believed that the failure was caused by (visually) undetectable mechanical damage in transportation from Isomedix, L
the irradiation facility.
During the LOCA test, the #14 AWG cable was loaded to only 10A.
Although this is acceptable, it is possible that this reduced loading could. create an unnecessary ampacity limitation for certain plant applications.
No findings were identified in this.
review.
5.
Unauthorized Repairs to Cable Armor NRC Region III referred to the Vendor Inspection Branch a report i
to them from Northern States Power Company (NSPC) regarding the i
licensee's discovery of damaged jacket material on some Rockbestos cable.
NSPC purchased Rockbestos 3/C #10 Firewall III cable with aluminum interlocked armor from the Rockbestos Company under Purchase Order NSP HIAWO2607, Rockbestos Shop-Order 72908 for use in Prairie Island Nuclear Generating Plant.
NSPC discovered while installing the cable that the armor had separated and that the jacket material was damaged; however, the conductor insulator was not damaged. During this inspection, the inspectors asked Rockbestos why the-jacket material was damaged.
Rockbestos stated that an unauthorized repair was made (TIG welding instead of soldering).
Not following the repair procedure for interlocked aluminum armor resulted in overheating the jacket, which caused jacket embrittlement and separation.
Rockbestos determined that the condition of the
- filler, flame-retardant polyethylene, indicated that the small area under the armor repair was exposed to a temperature of 160'C to 165'C.
This temperature degraded' the jacket material (Hypolon), but not
. i i
the crosslinked polyethylene. (XLPE) insulation.
Rockbestos determined that the insulation was not damaged based on the following:
t e
-=
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 18 of 24 a.
The XLPE insulation meetc the requirements of ICEA S-66-524 which requires survival and unaffected operation after exposure to 250*C short circuit conditions.
4 b.
The insulation had been qualified to the double transient of IEEE 323-74 f or combined PWR and BWR (340*F-171'C)- profile after being fully-aged and without a protective jacket.
c.
Heat aging studies have confirmed that this insulation will j
retain 85% of its original elongation properties after seven days exposure to 158'C.
j d.
Rockbestos confirmed that Firewall III meets the vertical-tray flame test without the added protection of the jacket.
As part of corrective action, the Rockbestos QA Manager sent a memorandum to his manufacturing manager and general foreman prohibiting repairs on aluminum interlocked armor.
The only other shipment of this cable was to Virginia Power.
A review of i
off-standard reports (OSR's) showed that nn repairs were conducted by Rockbestos on that order.
6.
Review of Rockbestos Qualification Report QR-6802 The inspectors reviewed Rockbestos environmental qualification report QR-6802, Revision.1, dated July 2, 1987,1for environmental qualification of Rockbestos coaxial, twinaxial, and triaxial cable for adverte nuclear Class 1E service.
Paragraph 12.1.B'of Data Section 2 of the report gives the acceptance criteria for insulation resistance (IR) measured per test pLn - TP-5803 in accordance with ICEA standard S-19-81.
Jheacceptancecriterion forIR,expressedintermsofgegohms(10 ohms) per 1000 feet of' cable len0th, was 100,000 (10 m
ms for types "A," "D,"
and-
"E" cable samples and 1,000,000 (1
) megohms for types "B" and "C"
samples.
The table of post-loss-of-coolant-accident (LOCA) sample evaluation IR values following Section 12.2 of the report j
shows all results to be passing. The post-LOCA insulation resis-tance values were listed as " greater than" -360,000 megohms (corrected for 1000 feet) for type "B"
and "C"
samples and 4,500,000 megohms (uncorrected as measured) for type "D" samples.
In order to determine the validity of these readings, the inspec-tors reviewed the relevant procedures, tech manuals, laboratory j
g.
- ~
ORGANIZATION: THE 4 A i.5TOS COMPANY 1
NEW HMEN AND EAST GRANBY, CONNECTICVT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 19 of 24 data sheets, interviewed the Rockbestos laboratory test equipment 1
calibration technician and reviewed his calibration records for Genrad 1864 Megohmeter Bridge, Serial Number 2311, indicated in 1
QR-6802 as having been the IR testor used for the-listed IR measurements.
The test plan for this test, TP-5802, called for.
use of this model instrument, set for 500 volts per ICEA S-19-81.
The Rockbestos general procedure for measuring IR on coaxial, twinaxial, and triaxial cables, RTP-0050-3, dated August 9, 1978, revision dated January.1,1986, was in effect at the time the post-LOCA irs were taken, but it prescribes use of the Kiethly 610C electrometer and a separate, 240-amp power supply -'or
" current set" in order to obtain valid IR readings of the high magnituderequiredonshort(inthiscase,lgfeet)gestsamples of coaxial type cables to assure that the 10 and 10 -megohm per 1000 feet acceptance criteria are met.
Review of the technical manual for the Kiethly 6100 electrometer indicated that its gable range, used as an ameter, extends down to as low as 10' ampercs.
Using the current set voltage setting of 100 volts specified by the type "B" cable-specific
(
laboratory test data sheets, or "TDBs" referenced in' RTP-0050 (suchasTDBG338ZforspecificationnumberRSS-6-100/LEcoaxig cable), the minimum current value would equate to an IR of 10 megohms.
However, review of the technical manual for the Genrad 1864 Megohmeter Bridge used revealed ghat, at 500 volts, its useful range is not greater than 20 X 10 megohms (20 terohms) at the greatest recommended scale and needle deflection. Therefore, l
when full scale readings of greater than 20 terohms were obtained during IR measurement of the 18-foot B and C samples, these were corrected for 1000 feet by dividing 20 terohms by a factor of 1000/18 (approx 55.56) yielding a figure listed in the QR-6802 data table as " greater than" 360,000 megohms per 1000 feet.
The indication in the table that these were passing values is erroneous in that a maximum reading of' " greater than 360,000 megohms" does not demonstrate that the acceptance criterion for the B and C samples of 1,000,000 meghoms is met.
The value of measured IR for the 18-foot D samples was listed in the report as 4,500,000 megohms, which, when corrected for 1000 feet by dividing by a factor of 1000/18 yields a value of 81,000 megohms. The 4,500,000 megohm result was ther efore characterized erroneously as " pass" in the table, when, in fact, the measured value corrected for 1000 feet is less than the specified accep-tance criterion of 100,000 megohms. per 1000 ~ feet for D samples, j
=
.., +,
,' ' ORGANIZATION: THE ROCKBESTOS COMPANY i
NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 20 of 24 In the interview of the Rockbestos laboratory test equipment calibration technician and review of the calibration procedures for the Genrad 1864 Megohmeter Bridge, it was learned that the procedure specifies that the instrument be calibrated'such that the maximum error does not exceed 10% of the full scale reading.
The calibration records for,Genr.ad 1864 Serial Number 2311 used for the QR-6802 IR measurements confirmed that tnis and other 3
Genrad 1864s are not calibrated above about 5 X 10 megohms.
This range is about.four orders of magnitude below the minimum rangeofaboutpX10 mogohms needeo to meet the lower IR speci-fication of 10 megohms-per-1000-feet with short sample -(18-FT) 1R measurements.
The calibration technician explained that this was the maximum range at which these instruments would have less than 10% error; so that it was their practice not to. attempt to calibrate them at any higher ranges and determine what the error was.
This meant, therefore, that Genrad 1864-Serial Number 2311 was uncalibrated for the range of IR values for which it was being used during post-LOCA sample evaluation for TP-5802 as published in QR-6802.
Thus, the values listed were (1) taken by an instrument that was-not only uncalibrated in the range of interest,: but was incapable of providing a valid reading of the required magnitude, (2) not demonstrated for samples B and C' to have passed as claimed,.and (3) were below specification for the D samples.
This failure to adequately evaluate the test results to assure that requirements are met constitutes a nonconformance with respect to Criterion XI of Appendix B to 10 CFR Part 50 and is designated as Nonconfor-mance 99900277/90-01-03.
7.
Review of IR Measurement Practices The QR-6802 report also-contained, but made no reference to, certified test reports (CTRs) of tests done by production - personnel on cable from -the -
reels from which the TP-5802 samples were taken using routine or standard IR measurement practices.
These reports indicated that satisfactory irs were obtained, e.g., on a Cy, megohms per 1000 dated February 27, 1986, for type B samples showing 1.53 X 10 feet.
Therefore, the inspectors reviewed the current procedures and those in effect at the time of the testing -in question, including the general procedures and individual test data sheets or "TDBs" referenced in the procedures.
The Rockbestos general procedure for measuring IR on coaxial, twinaxial, and triaxial cables, RTP-0050,. dated August 9,1978, revision dated September 24, 1986 (current revision), prescribes i
f
i 1
ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT l
REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 21 of 24 the use of a Genrad 1864 Megohmeter Bridge set for the voltage.
specified in the "TDB" (cable-specific laboratory test data sheet type "B").
This revision of the procedur' also still specifies the use of a separate 240-amp power supply which was used in accordance with previous revisions with the Kiethly 610 electro-meter, but which is inconsistent with tne use of, the Genrad 1864 which is AC powered and contains its -own internal test voltaga/
current power supply.
The procedure ther requires conversion of the measurement to units of megohms per 4000 feet of cable and reflects :the current practice of. measuring an entire cable reel at once by givir.g 2000 feet as the cable length measured in the sample calculation.
Thus, the procedure was improperly revised because the reference to the separate. power supply should have.
been deleted when use of the Genrad 1864 was instituted._
The laboratory test data sheets referred to in RTP-0050, such as i
TDB G3382, Revision E, dated December 17,.1874, for RSS-6-100/LE i
coaxial cable, however, specify a' length of 20 feet for a-specimen to be used for the test and specify a current set (power supply) voltage of 100 volts.
The sample length, current set, t
and voltage specified are not used with a Genrad 1864, but would i
have been used with a Kiethly 6100 electrometer consistent with.
previous revisions of RTP-0050 (such as the revision dated January 6,1986 in effect at the time of the testing in question) and requiring a separate power supply, i
While it appeared that the IR measurements given in the CTRs were taken in accordance with effective procedures, the practice since the later revision which established use of the Genrad 1864 on j
whole cable reels has not been in accordance with contradictory l
and out of date TDBs referenced.
It should be noted that the post-LOCA IR values listed in QR-6802, while not all meeting the j
stated. acceptance criteria, could be considered acceptable for i
some qualified applications, but since they.were taken in an l
uncalibrated and in some instances, invalid range of the instru-1 ment, the error associ~ated with those values is indeterminate.
I The deficiencies noted with respect' to revision of RTP-0050 and.
lack of revision to-associated TDBs constitute nonconformances with respect to the requirements of. Rockbestos quality procedures, and Criterion V 10 CFR Part 50 Appendix B in that they represent lack of procedures appropriate to the circumstances for activi-l ties affecting quality and are-designated as Nonconformance 1
99900277/90-01-02.
i b
i
. usu
> ORGANIZATION: THE ROCKBESTOS COMPANY NEW HAVEN AND EAST GRANBY, CONNECTICUT REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 22 of 24 8.
Followup on Allegation of Shipment of Damaged. Unrepaired Cable j
The NRC received an allegation that a 1,200-foot reel of 37-con-ductor (37/C), nuclear safety grade cable with damaged insulation p
on one conductor had been shipped from Rockbestos' East Granby plant to a nuclear utility without the damaged insulation being repaired.
The cable in question was identified-as 37/C Firewall Ill, with radiation cross-linked polyethylene (RXLPE}
j insulation.
The reported damage was a 1-inch section of insula-4 tion stripped from a conductor with blue insulation with a black stripe about 400 feet from the. end of the 1,200-foot reel, and was believed to have been caused by a burr on the flange of the E
metal feed. reel which snagged the insulation of the conductor as it was being unwound from this reel into the jacket extruder machine to be jacketed with the other conductors.
The incident allegedly occurred near Labor Day,1988, and the alleger believed i
that the cable had been sent to the Philadelphia Electric Company (PECO) licensee for the Peach Bottom and Limerick nuclear plants) (based on PECO's-being the only customer known by the!
alleger to have bought such unusual cable.in the past.
An OSR l
reportedly was initiated on the damaged insulation, but the dis-position of the OSR was unknown, and the alleger stated that he believed that the cable had been shipped without repair.
From the East Granby plant log of all shop orders for 1988, the i
NRC inspector identified all the 1988 shop orders.for PECO.
The inspectors then searched the files of all shop orders (all custo-mers) for 1988 and identified only one for 37/C cable for PECO and only one other one for 37/C cable.
Review of those shop orders revealed that the one for 37/C cable for PECO had been shipped in April of 1988, and the other one was shipped late in September,1988, but was for a commercia1 ' electrical supplier.
The "0SR Report," the computer printout of the descri status of presumably all OSRs ger.erated in a given year,ption.and was then
=
searched for all OSRs involving PEC0, damaged insulation, 37/C cable, and the time frame of interest.
The inspector identified only one OSR for 37/C cable that was issued during the time frame of interest, but the customer was the commercial electrical supplier mentioned above. The problem was with insulation on one conductor, but it was a striping discrepancy, not damage, and the color code marking was different than that alleged.
=
Review of related documentation, including the shop order, the OSR sheet for the shop order, and the OSR file copy itself, revealed that there indeed had been a 1,200-foot reel which was 4
y 5
I l
+
'~ ' '
(
ORGANIZATION: THE ROCKBESTOS COMPANY NEW tiAVEN AND EAST GRANBY, CONNECTICUT
\\
REPORT INSPECTION M
NO :- 99900277/90-01 RESULTS:
PAGE 23 of 24 l
L one of three on this order, but the OSR was on one of the other
[
two reels and had been satisfactorily dispositioned.
In addition, i
although the shop order. was for 37/C cable, it -listed the cable as "Pyrotrol 210," the Rockbestos trade name for their non-environmentally qualified commercial grade version. of the same formulation of RXLPE-insulated cable, and not nuclear safety grade Firewall III as the alleger had described it.
A search of all OSR file copies written during the time frame of interest was conducted. but none were found that either matched or resembled the allegers description of the problem.
During this activity, the 'QC supervisor described the details of the routine OSR processing practice, disclosing the facts that led the inspectors to identify a lack of adequate OSR accountability.
Rockbestos documents and tracks the disposition of the various conditions adverse to quality that may occur -during cable production, including damage, nonconformances, etc., using a.
system of three-part tags or chits called Off-Standard-Reports (0SRs).
According to Rockbestos' description-of this process, when a condition that requires some resolution is identified, the cognizant supervisor is notified by the identifying worker or quality control - (QC) inspector eM-then the QC inspector is notified, if not already informed, avid fills out an OSR, attaches one part to the affected item, as applicable, and turns in the copies to the QC office for disposition by the QC _ supervisor.
The QA manager's assistant at the East Granby Plant records the OSRs that come into the office, enters the information into'the OSR tracking computer data base and files the copies for review and disposition by the QC supervisor.
The QC supervisor deter-mines the corrective action and causes it to be taken, then-determines final disposition which is noted on the file OSR copy and - the status is entered in the data base.
When corrective action, such as repair, is completed, the OSR tag is removed from the affected item, noted as completed and returned to QC, where the status is noted by the QC supervisor, entered in the tracking i
data base, and the closed-out OSR is filed.
The weaknesses identified in the practice of controllin generation, tracking and disposition of OSRs were that (1)g the OSRs i
are not necessarily logged and/or copies retained by the identi-J fying worker and/or QC inspector and (2) although the OSR forms have printed serial numbers, no record is kept of which pads or packs of OSRs, by numerical series,. are issued to various plant -
i personnel; thus accountability for each form, by serial number, l
l
Q ORGANIZATION: THE ROCKBESTOS COMPANY
-NEW HAVEN AND EAST GRANBY, CONNECTICUT 1
REPORT INSPECTION NO.: 99900277/90-01 RESULTS:
PAGE 24 of 24 i
is not maintained. This condition creates the potential for OSRs to be lost, damaged, or improperly cleared without an auditable accounting system to detect such occurrances.
In response to i
these concerns, the Rockbestos QA Manager agreed to. investigate the matter and take appropriate corrective action.
The allegation follow-up indicated that if there had been damaged insulation shipped unrepaired, and if the OSR had been lost or l
even improperly removed, the cable was not shipped to PECO or any other nuclear utilities.
However, this allegation, while not j
substantiated, did serve to reveal the deficiencies in the
~
Rockbestos OSR system which could potentially allow such problems to go undetected with nuclear safety-related material. Therefore, this issue involving control.of OSRs remains. unresolved pending completion of corrective action by Rockbestos and further review during a future NRC inspection, and is designated Unresolved item 99900277/90-01-05.
F.
PERSONS CONTACTED:
New Haven
- H. Littlehales, QA Manager-
- G. Klein, Vice President of Engineering
- B. Gehm, Jr., Manager Electrical & Production Engineering M. Mennone, Test Technician J. Morganelli. Qualification and Test Administrator East Granby l
- G. Littlehales, QA Manager C. O' Conner, Chief Technician (Laboratory)
B. Jendza, QC Supervisor (East Granby plant)
D. Schuler, Manager of Radiation Services M. Bliss, QA Analyst
' " '