ML20133C095

Insp Rept 99900001/96-01 on 960318-0417.Weaknesses Noted. Major Areas Inspected:Licensee Activities at Engineering & Fuel Fabrication Facilities in Lynchburg,Va

ML20133C095
Person / Time
Issue date:	11/25/1996
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20133C080	List: ML20133C077 ML20133C095
References
REF-QA-99900001 NUDOCS 9701070032
Download: ML20133C095 (39)

Text

.

~*

U.S. NUCLEAR REGUL ATORY COMISSION OFFICE OF NUCLEAR REACTOR REGULATION l

Report No.: 99900001/96-01 Organization: Framatome Technologies, Inc. (FTI)

Framatome Cogema fuels (FCF) l Lynchburg, Virginia i l

Contact:

R. L. (Ronnie) Gardner l Manager, Quality FCF l

Nuclear Industry l Activity: FCF provides pressurized-water reactor (PWR) reload i core designs, safety analysis, and licensing, fuel  !

assemblies, and fuel-related core components to the U.S. nuclear industry.

Dates: March 18, 1996 - April 17, 1996 l I

Inspectors: Steven M. Matthews, DISP /PSIB David H. Brewer, DISP /PSIB ,

Dr. John F. Carew, Brookhaven National Laboratory Carl J. Czajkowski, Brookhaven National Laboratory Geoffrey R. Golub, DSSA/SRXB Rodney L. Grow, Par & meter, Inc.

Edward D. Kendrick, DSSA/SRXB Kamalakar R. Naidu, DISP /PSIB Billy H. Rogers, DISP /PSIB Approved by: Gregory C. Cwalina, Chief Vendor Inspection Section Special Inspection Branci:

Division of Inspection and Support Programs 4 9701070032 961125 PDR GA999 EMVFRAMA 99900001 PDR

y,a _ n,- - a +W> a --,n eu~A OL--s4 eL -~sk- 3+w--- --bal*A---A a ~ m 4 - - - - - wk m-

--_.A--.m A- a -

1 INSPECTION SUMARY:

From March 18, 1996, through April 17, 1996, the U.S. Nuclear Regulatory Commission (NRC) conducted an inspection of Framatome Cogema Fuels (FCF)

[formerly the Babcock and Wilcox (B&W) Fuel Company (BWFC)] activities at the engineering and fuel fabrication facilities in Lynchburg, Virginia.

The inspection bases were:

General Design Criterion (GDC) 10, " Reactor Design," and GDC 12,

" Suppression of Reactor Power Oscillations," of Appendix A. " General Design Criteria for Nuclear Power Plants," to Part 50, " Licensing of Production aid Utilization Facilities," of Title 10 of the Code of Federal Reagiations (10 CFR)

• Appendix B, " Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants," to 10 CFR Part 50 10 CFR Part 21, " Notification of Failure to Comply or Existence of a Defect" Section 4.2, " Fuel System Design," of NRC NUREG-0800, " Standard Review Plan" (SRP), Revision 2, July 1981, and its Appendix A. " Evaluation of Fuel Assembly Structural Response to Externally Applied Forces,"

Revision 0 FCF quality assurance (QA) program 56-1177617-03, " Quality Assurance Program for B&W Fuel Company," May 30, 1994 During this inspection, the team noted weaknesses and observations concerning FCF activities that affect quality. Neither the weaknesses nor the observations described in this report require any specific action or written response by FCF.

A list of acronyms used in this report is provided on page 37.

2 STATUS OF PREVIOUS INSPECTION FINDINGS l 2.1 Nonconformance 99900001/90-01-01 (CLOSED)

Contrary to Criterion V of Appendix B to 10 CFR Part 50 and Section 5.0 of the BWFC, Commercial Nuclear Fuel Plant (CNFP) procedure QC-506,

" Receiving Inspection - Lower End Fitting Mark C," Revision 1, October 27, 1980, CNFP failed to identify raised metal in the form of a burr adjacent to one grid pad bore hole on lower end fitting 15273-04.

FCF personnel stated that the burr had been removed and that prior to use, all lower end fittings were cleaned and reinspected for burrs. All receiving inspection personnel were instructed in the performance of visual inspection.

.. j s

The inspectors reviewed FCF corrective actions taken and determined that

, FCF had implemented actions to ensure that appropriate inspection standards were used to determine the acceptability of the lower end fittings.

2.2 Nonconformance 99900001/90-01-02 (CLOSED)

Contrary to Criterion V of Appendix B to 10 CFR Part 50, procedure QC-1414 " Retention and Storage of Quality Assurance Records," Revision  !

4, September 13, 1990, did not provide requirements for retention of l radiographs and did not identify whether the radiographs and the reader '

sheets were quality records.

The inspectors reviewed FCF corrective actions taken and determined that FCF had revised procedure 00-1414, Revision 5, August 7, 1991, to classify reviewer reader sheets and radiographs as lifetime c 'ality records.

2.3 M9Dconformance 99900001/91-01-01 (CLOSED)

Contrary to Criterion V of Appendix B to 10 CFR Part 50, and Section 6.3.6 of procedure 0C-802, " Fuel Rod Inspection (In process and Final),"

Revision 12, February 21, 1989, a 100% visual inspection was not performed as evidenced by felt cleaning plugs discovered in 22 fuel rods after visual examination and cleanliness acceptance were performed by quality control (QC). l The QC inspector responsible for performing the visual inspection was  ;

counseled on several occasions to ensure that the inspector understood '

that, by initialing the route card (RC), the QC inspector documented that the inspection had been completed to the requirements specified.

To ensure that instructions were fully understood and followed, the Manager of Inspection and each Inspection Foreman emphasized the importance of and the need for compliance with inspection instructions.

The inspectors reviewed FCF corrective actions taken and determined that FCF had implemented actions to ensure that appropriate inspection standards were used to accept fuel clad tubing. The inspectors noted that subsequent to the implementation of these changes, FCF changed its method for cleaning the inside diameter of the tubing and discontinued the use of felt cleaning plugs.

2.4 Nonconformance 99900001/91-01-02 (CLOSED)

Contrary to Criteria V, XV, and XVI of Appendix B to 10 CFR Part 50, CNFP did not issue a component discrepancy report (CDR) and a contract variation approval report (CVAR) to identify, document, evaluate, resalve, and process the noncon armance ider.tifi.d when felt cleaning plugs were found in 22 fuel rods.

1 A CDR and a CVAR were initiated in August 1991. The Manager of QA ,

conducted a series of meetings to identify and correct the causes of not '

reporting this nonconformance. In this case, the true cause for not issuing a CDR was that the responsible personnel viewed the recovery operation as an in-process reject that scraped each fuel rod containing l felt cleaning plugs. During that time, the in-process reject practice I was allowed without issuing a CDR. The current practice requires that a CDR be issued if a part or process deviated and the resolution of that deviation was not covered by a procedure or RC. 4 The inspectors reviewed FCF corrective actions taken and determined that l FCF had implemented appropriate actions to ensure that CDRs and CVARs would be issued as required.

2.5 Nonconformance 99900001/91-01-03 (CLOSED)

, Contrary to Criterion V of Appendix B to 10 CFR Part 50, procedure MA-450 did not provide guidance to the operator for the extent of QC '

inspection required or what notification was to be given to QC to ensure that the visual examination for cleanliness was performed.

Additionally, MA-450, Revision 16, March 8,1991, and QC-802, Revision l 13, April 29, 1991, did not provide the extent of the inspections to be  ;

performed by the operator or the QC inspector. l

The inspectors' review of the procedures referenced determined the following

• Procedure MA-450 was at Revision 23, December 18, 1995, and the I applicable paragraphs (4.2 and 4.3) were responsive to the l concerns.

l Procedure QC-802 was at Revision 20, August 1, 1995, and paragraph l 6.3.5 was responsive to the concerns.

2.6 Unresolved Item 99900001/91-01-04 (CLOSED)

The inspectors determined that BWFC had initiated actions to identify the root cause for the felt cleaning plugs left in loaded fuel rods. An evaluation in accordance with Section 21.21 of 10 CFR Part 21 may also be required. Pending receipt by NRC of a written response containing the evaluations, this issue was considered an unresolved item.

The inspectors reviewed the FCF's root cause evaluation 51-1205252-00 and concluded that FCF had adequately evaluated the event.

3 INSPECTION FINDINGS AND OTHER COMMENTS 1

This inspection included an evaluation of the FCF fuel operations management, engineering, and quality staff, and certain activities for providing reactor licensees PWR neutronic design information, reload core designs, fuel assemblies, fuel related core components, and fuel related field services.

l 1

l

o.

l The emphasis of the inspection was on the fuel design and engineering, organization, management, training and qualifications, corrective actions, and l compliance with the requirements of 10 CFR part 21.

3.1 10 CFR Part 21 Procram

a. Inspection Scope The inspectors reviewed FCF compliance with the NRC requirements in 10 CFR Part 21. The FCF procedures implementing the requirements of 10 CFR l Part 21 were BWNT-1707-01, " Processing Safety Concerns," Revision 24, April 13, 1994, and BWNT-1707-03, " Reporting of Defects and Failures to Comply (10 CFR 21)," Revision 1, 1994. The inspectors reviewed selected  :

10 CFR Part 21 evaluations, known as preliminary report of safety concern (PSC).

b. Observations and Findings I

b.1 PSC 5-94 1 PSC 5-94 concerned the fuel peak cladding temperature (PCT) during the reflood stage of the large break loss-of-coolant accident (LOCA). The i

inspectors discussed the P5C with the LOCA and safety analysis task and supervising engineers. The PSC addressed the use of nonconservative values for (a) the inlet enthalpy in the core thermal analysis and (b) ,

the assumed initial inventory in the core flood tank (CFT). The l nonconservative inlet enthalpy was due to the use of a coarse time step i in the reflood analysis, and an inadequate averaging of the nodal l enthalpy following node dryout and rewet.  !

The nonconservative initial CFT inventory was first observed on September 30, 1994, during the development of the new FCF RELAP 5/ MOD 2

- B&W methodology for application to the MARK-B11 LOCA analysis. At this time the licensed LOCA analysis was based on CRAFT 2 . A 1

subsequent CRAFT 2 analysis confirmed the nonconservatism in the assumed initial CFT inventory.

These nonconservatisms in the LOCA analysis resulted in a reduction in the LOCA kW/ft limits. FCF evaluated the effect of the reduction and determined that, because of available margin in the operating limits, no change was required in the FCF determined LOCA-based operating limits.

During these discussions, the inspectors noted that four months had elapsed from the time the LOCA analysis initial CFT inventory nonconservatism was first identified (September 1994) and the date NRC

'" CRAFT 2 - Fortran Program for Digital Simulation of Multinode Reactor Plant During Loss of Coolant," J. J. Cudlin, B&W, October 1982 received notification i. Although the inspectors recognized that'the 4

initial identification was made using the unlicensed RELAP 5/M0D 2 - B&W l

, methodology, the inspectors concluded-that a final confirmation of this nonconservatism CoJld have been made using existing CRAFT 2 models.

! ' Procedure BWNT 1707-01 required that the NRC be notified within_60 days i

after the PSC is initiated. PSC 5-94 was initiated on November 28, 1994, and was reported to the NRC on January 27, 1995; within the 60 day requirement. However,Section VIII, of this- procedure also required, in part, that a responsible engineer initiate a PSC for any concern i discovered during, and which is related to, the design and analysis that

! has or may have safety implications, i The inspectors determined that approximately 60-days had elapsed between

' the time the engineer had reasonable indication that the emergency core cooling accident (ECCS) LOCA analysis was nonconservative and the time that the PSC was initiated.

b.2 PSC 1-95 i

The inspectors discussed the identification and status of PSC 1-95 with the FCF responsible engineer. PSC 1-95 identified the concern regarding i 'a possible return to critical during the recovery stage following a j j small break LOCA (SBLOCA). During this period the plant may spend some i

time in the boiler / condenser mode and may accumulate deborated water in j the reactor coolant pump suction piping and downcomer. If, after this

accumulation, an operator performs a pump bump, or under natural j i circulation, this deborated water enters the core, the core boron '

i concentration may not be sufficient to ensure subcriticality. (This concern was also the subject of PSC 8-81 which was resolved in 1985.) l i

PSC 1-95 was a result of several analyses which indicated that the plant conditions assumed in the 1985 resolution of PSC 8-81 are no longer valid {or B&W designed PWRs. The NRC was notified of PSC 1-95 by i

letter . FCF recently issued a report' on the initial Phase-I investigation of the issues associated with PSC 1-95. This study was

, performed for the B&W Owners Group, and was not intended as a dntailed evaluation but rather as a path finding activity. The study proposed a follow-on Phase-II activity including a set of detailed transient l analyses and an evaluation of special effects for resolving PSC 1-95.

2 B&W letter to NRC, " Report of Preliminary Safety Concern Related to Large Break LOCA ECCS Analysis," January 27, 1995 3

B&W letter (0G-1521) to NRC, " Report, of Potential Core Criticality Concern," June 13, 1995

'FCF 47-1244436-00, " Preliminary Report to the B&W Owners Group on PSC 1-95 Investigation," January 1996 The inspectors concluded that the preliminary Phase-1 evaluation L adequately included the major issues associated with PSC 1-95.

c. Conclusions- -

Based on its review of PSC 5-94, the inspectors concluded that FCF could have initiated the PSC before November 28, 1995, and that FCF's failure to evaluate the PCT during LOCA concern when indications of the problem l

were first identified in September 30, 1994, was considered a weakness in the implementation of FCF's program.

3.2 Quality Assurance The inspectors reviewed FCF's QA program described in 56-1177617-03,

" Quality Assurance Program for B&W Fuel Company," May 30, 1994. The inspectors observed that'the FCF Manager of Quality had the responsibility, authority and organizational freedom to enforce and monitor compliance with the program.

The inspectors concluded that the QA program was generally implemented in an acceptable manner for the areas evaluated during this inspection.

3.3 Reload Core Desian. Safety Analysis and Licensina Process

a. Inspection Scope To evaluate FCF's reload core design, safety analysis, and licensing processes, the inspectors reviewed the performance, interfaces, and documentation of the reload process. The reload core design, analysis, and licensing activities consisted of determining the licensee requirements and the preliminary fuel cycle design; performing the steady-state and transient neutronic, thermal-hydraulic and fuel performance analysis; and updating the cycle-specific reload licensing analysis and the process computer databank.

b. Observations and Findings b.1 Licensee Requirements The inspectors observed that the process started with the FCF/ licensee reload fuel contract and the FCF reload project manager (PM) interacting with the licensee to define the requirements for a new reload core. The licensee specified the fuel assembly type and options for the new reload. Preliminary operating requirements and design data were documented in preliminary energy requirements.

The inspectors determined this was an iterative process between FCF and <

the licensee and found that the carrespondence dc.umented the request for the licensee to approve the requirements and design data.

l ,

l

The PM created a contract requirement document (CRD) which outlined the ,

scope of supply and special licensee requirements and used a document i release notice (DRN) to release the CRD to FCF fuel engineering.

The inspectors concluded that FCF's licensee requirements process did not use formal checklists or computerized design data. The inspectors, however, did not observe any instances where the lack of formality led to a design error.

b.2 Reload Core Design Process The inspectors reviewed the reload core design process and observed that the preliminary and scoping fuel cycle designs were based on the preliminary operating requirements specified by the licensee. This initial design effort determined the preliminary reload batch size and enrichment; the analysis utilized the two-dimensional (2D) lattice physics code (CASMO 3) results and the three-dimensional (3D) core simulation calculations from the NEMO computer program. This initial analysis was documented in the preliminary fuel cycle design (PFCD) report and this information was released via the DRN process. j According to FCF, the reload licensing analysis task engineer (RELATE) l was a key individual in the reload core design process and was  !

responsible for maintaining ths technology for the task. The inspectors I were also told that the reload analysis and licensing services (RALS) l inspectors maintained the quality of the service and ensured timely l interaction with the licensee. The cognizant RELATE generated the fuel l cycle design requirements (FCDR) following receipt of the CRD. Based on the requirements of the FCDR, the final fuel cycle design (FFCD) was produced and documented in a report to the licensee. The FFCD analysis consisted of CASMO 3 and NEM0 calculations which were documented in calculation packages.

The inspectors reviewed prior cycle FFCDs against the new reload FFCD calculation packages. From its review, the inspectors determined that the process was controlled by administrative procedures. FCF had not developed detailed engineering instructions. The inspectors also noted '

that the calculation packages did not contain documentation identifying what the reviewer checked or verified (e.g., no check marks or l checklists). The inspectors made similar conclusions regarding the '

reload safety analysis process discussed in Section 3.3.b.3 of this report.

b.3 Reload Safety Analysis Process The reload safety analysis process was performed to ensure that the updated version of the final safety analysis report (FSAR) remained valid for the reload core design. The accidents analyzed in the FSAR were evaluated to ensure that the reload core thermal performance during hypothetical transients was not degraded and that site doses were below the acceptance criteria of 10 CFR Part 100. In addition, the reload ,

safety analysis process included the determination of the core operating '

limits and reactor protection system (RPS) setpoints.

l The reload safety analysis consisted of both the non-LOCA and LOCA analyses. The safety analysis used the flux / flow limits determined by the NEM0}hermal-hydraulicanalysis,thepowerpeakinglimitsdeterminedby in the nuclear analysis, and the RPS offset limits determined in the NEMO maneuvering analysis. In the non-LOCA analysis, the core parameters that impact the FSAR transients were compared to the plant  ;

licensing basis values (LBVs) to ensure that the licensing basis remained valid for the reload core. The specific parameters evaluated  !

included the moderator and Doppler reactivity coefficients, initial baron concentration and inverse boron worth, maximum ejected and dropped  ;

rod worths at hot-full-power (HFP) and rod group worth at hot-zero-power  !

(HZP).

The transients considered included both overheating and overcooling events. The overheating events included moderator density dilution, rod withdrawal and ejection, loss of flow, loss of power, reactivity change l and the startup event. The overcooling transients considered were the steam line break, cold water event and the dropped or stuck rod event.

In the LOCA analysis, the linear heat rate (LHR) effects of fuel prepressure, enrichment, and plenum volume, and the cycle-burnup i dependenceofthefuelrodpressureandtemperaturewereevag*ated. The fuel rgd parameters were calculated with computer codes TAC 0 or GDTAC0 in the case of gadolinia (Gd) loaded fuel. Plant modifications ,

that can affect either the reload LOCA analysis (e.g., ECCS assumptions) or the plant transient analyses were reviewed against the licensing basis analysis. These changes were either verified to be within the LBVs or a re-analysis was performed to demonstrate their acceptability.

The power / imbalance / flow trip setpoints and safety limits were calculated using the cycle-specific offset limits determined in the NEMO maneuvering analysis.

5 BAW-10180-A, "NEM0-Nodal Expansion Method Optimized," Revision 1, March 1993

'BAW-10141P-A, " TAC 0 2: Fuel Performance Analysis," Revision 1, June 1983 7

BAW-10162P-A, " TACO 3: Fuel Pin Thermal Analysis Code," Revision 1, November 1989 sBAW-10184P, "GDTAC0, Urania-Godolinia Thermal Analysis Code," May 1992 l

l The reload evaluation also included a NEMO reactivity analysis to determine the refueling boron concentration required to ensure that the {

core remains subcritical following a refueling boron dilution event.

The NEMO analysis also provided core reactivity requirements for  !

determining (a) the minimum borated water storage tank (BWST) concentration to ensure that the core remains 1% shutdown following a LOCA and (b) the minimum volume of borated water in the BWST and the boric acid storage tank to independently ensure that the core can be hcid at cold shutdown.

The plant exclusion area and low-population zone boundary doses were evaluated for each reload to insure they were within the 10 CFR Part 100 4

acceptance criteria. The reload fuel and coolant radionuclide inventories were calculated and compared to the LBVs. A reanalysis was performed to demonstrate the acceptability of the boundary doses if the reload values were outside the LBVs, 4

i The inspectors found that the reload safety analyses were documented in formal cycle-specific calculational files. The safety analysis calculations were reported in calculational sumniary sheets.

i From its review of the reload safety analysis process, the inspectors determined that if a core parameter value was not bounded by the LBV for a given transient, then the value was justified by reanalysis of the transient. This reanalysis may consist of a simple sensitivity calculation or a complete reanalysis of the transient.

i i The inspectors reviewed prior cycle calculational files against the new reload calculation packages. From its review, the inspectors determined that the documentation and analysis technique tended to be guided by the prior cycle (s) documentation. The inspectors observed that the process was cunt, oiled by administrative procedures and that FCF had not developed detailed engineering instructions. The inspectors also noted I

that the calculation packages did not contain documentation identifying what the reviewer checked or verified (e.g., no check marks or checklists were used).

Based on these observations and similar observations made by the inspectors during their review of the reload core design process (see Section 3.3.b.2 of this report), the inspectors examined numerous calculation packages for the reload design process and the reload safety analysis process (see Section 3.3.b.4.(1)-(4) of this report).

! b.4 Reload Licensing Process The reload licensing process defined the specific tasks to be performed, licensing schedule, task coordination and interfaces, and required design reviews. The reload PM provded the primary licensee interface for the reload. The reload licensing process consisted of 18 tasks which were identified numerically (e.g., task 4 was the ECCS analysis and task 11 was the nuclear analysis). A task engineer was assigned to each task and was responsible for the basic technology and methods 4

4 i

A  !

employed in carrying out the task analysis. The task analyses were l documented in a series of FCF reports. The FCDR and the site sup  !

document were the primary output of the reload licensing process port i f (

i Each of the following reload designs was reviewed by the inspectors with  ;

the PM, the RELATE, and the engineers who performed the reload core >

l design and the reload safety analyses.

)

{ (1) Three Mile Island 1 i

The FCF licensing analysis provided the technical justification l

! supporting Cycle 10 operation of the GPU Nuclear Corporation I j (GPUN), Three Mile Island Unit 1 (TMI-1). The general reload

licensing information consisted of the fuel assemblies, burnable -

poison rod assemblies (BPRAs), and reactor rod cluster control assemblies (RCCAs) for TMI-1 Cycle 10 and was specified in the l GPUN/FCF reload fuel contract. The latest safety analysis and

! plant systems input for the Cycle 10 reload evaluation were

! . provided in the plant FSAR. The reload data checklist specified the initial number of fuel assemblies and plant operating conditions (e.g., fiows and temperatures). Those requirements l were finalized based on the FCF fuels engineering preliminary fuel cycle analysis of the cycle-length, number of fuel assemblies, i axial blanket, and Gd zoning.

j The reload licensing. analysis was initiated by the release of the j reload licensing schedule, which specified the tasks to be performed, schedule, and deliverables. The TMI-I Cycle 10 licensing analysis consisted of the standard FCF task-labeled t analyses.

As part of the evaluation of the FCF reload licensing process, the inspectors reviewed the nuclear analysis with the supervising engineer, the RELATE, and the engineer responsible for the analysis. The inspectors concluded that the calculations were complete and well documented.

The inspectors found that FCF's QA program for the use of computer codes in the reload licensing process required that approved code versions be used in licending analyses or, when approved versions l were not used, a justification must be provided in the calculation  !

files. During its review of nuclear analysis 32-1218970-01 and 32-1219600-01, theinspectorsnotedthatgnapprovedversionsof l NEMO (NEMO 3.lRS and NEMO 4.lRS) and SHUF (SHUF 3.1RS) were used  ;

in the nuclear analysis. However, no justification for the use of these unapproved versions was documented as required by the FCF QA program.

'The SHUF program is used to shuffle the fuel isotopic concentrations between core locations.

Oe When the inspectors brought this omission to the attention of the i FCF staff, revisions to these calculation files were provided that  ;

contained a justification for the use of unapproved computer codes on the basis that the errors in the unapproved versions had no impact on the results of the nuclear analysis.

4 The TMI-1 Cycle 10 reload safety analysis was reviewed by the

inspectors with the responsible task engineer and unit manager of Framatome Technologies, Inc. (FTI). The safety analysis compared the Doppler and moderator temperature coefficients (MTC), initial boron concentration, maximum ejected and dropped rod worths, and 4

rod group worths to the bounding values used in the FSAR. The values used in the FSAR were shown to bound the values for THI-1 i

Cycle 10. The safety analysis demonstrated that the generic set of LOCA analyses bounded the TMI-I Cycle 10 reload.

The TMI-l Cycle 10 reload fuel rod analysis was reviewed by the inspectors with the responsible fuel rod analysis engineer. The fuel rod mechanical design analysis used the power distribution ,

and power history data determined in the fuel cycle design analysis to evaluate the cladding creep collapse and the stress and strain for the Cycle 10 fuel. The design data for each of the Cycle 10 batches was taken from the applicable documents list (ADL). The axial locations of the fuel stacks, control rod poison stacks, burnable poison stacks and the axial blankets were determined from the core and vessel specifications and drawings. i

, The stress parameters for each of the four Cycle 10 fuel rod designs (MARK-B8, MARK-B8V, MARK-B9 and MARK-89-Gd) were bounded ,

by conservative licensing basis analyses. The creep collapse time l for both the MARK-88 and MARK-89 fuel assemblies was bounded by a  !

previous generic analysis. l As part of the inspectors' review of these generic analyses, the )

fuel rod growth calculation for the MARK-B9 fuel (32-1202547-00) was reviewed in detail. The growth analysis included a statistical combination of growth tolerances and used a conservative fluence growth model. The documentation was complete and included the necessary verification reviews.

The 28 MARK-B9-Gd fuel assemblies included Gd fuel rv. s The fuel mechanical analysis (32-1222994-00) concluded that the creep collapse analysis for the Gd fuel was bounded by the generic 4

analysis performed for the MARK-89-Gd fuel (32-1219532-00). The generic analysis was reviewed and it was noted that the analysis

employed an unapproved version of GDTACO.

However, upon the inspectors' request, the FCF staff provided a subsequent creep collapse analysis (32-1224702-00) which demonstrated that there was no significant difference in the results of the calculations performed by the approved versus unapproved versions of GDTACO.

1 .  !

i l i .

i The inspectors concluded that the documentation of all analyses

reviewed complied with the FCF documentation and verification i requirements.

l (2) Sequoyah I l

b j The inspectors selected the Tennessee Valley Authority (TVA),

1 Sequoyah Unit 1 Cycle 9 reload for review because Sequoyah was a

[ non-B&W designed plant (Westinghouse). Cycle 9 was the first 4

reload provided by FCF for Sequoyah. The inspectors were also ,

L

concerned about the handling of design inputs for a non-B&W plant and the analysis to support a mixed core reload. The inspectors were also told that the fuel design for Sequoyah, the Mark-BW, had l t

been used in fuel for Catawba and McGuire. (FCF had previously i

supplied fuel and LOCA analysis for McGuire, a sister plant to  ;

{ Sequoyah.)  !

The inspectors found that this reload was not c typical FCF reload j in that a first-of-a-kind (F0AK) analysis was performed prior to initiating the typical cycle specific analysis. The inspectors-

! found that the F0AK effort addressed changes to the FSAR, technical. specifications, core operating limits report (COLR) and

, also produced the topical report BAW 10220P, " Mark - BW Fuel

Assembly Application for Sequoyah Nuclear Units 1 and 2," Revision

0, March 1996. The inspectors were told that the FOAK analysis addressed the Sequoyah plant specific design and focused on the j plant safety analysis. The inspectors observed that the topical j report documented the LOCA and non-LOCA transient analysis,

thermal-hydraulic analysis, and mechanical and containment j analysis, i

j The neutronics analysis was largely addressed in prior Mark-BW 4 documents. The topical report documented the licensing basis for i FCF fuel in the Sequoyah reactor and identified the bounding physics parameters for the cycle-specific reload analysis. The l inspectors found that the Sequoyah design inputs were documented l

via correspondence between FCF and the licensee and the documentation showed that the inputs had been verified by the

licensee.

1 The inspectors found that the PFCD had been completed and FCF nad

received the final energy requirements from the licensee (TVA).

The inspectors reviewed the list of Sequoyah reload deliverables and contract documents list. The inspectors concluded that the handling of design data inputs and the documentation of FCF 4

activities through the preliminary design were consistent with the

FCF administrative manual procedures.

4

To assess the analysis process the inspectors examined calculation

. package 32-1228786-00, "Sequoyah Catchup and Accident Analysis."

The inspectors found that this package documented the benchmarking i of the CASMO 3/NEMO neutronics model for Cycles 1 to 8.

] t 4

T 4

.n. v .. . - -

- , - . ,w, . , , . . .-. , + - - - - - - ,

_______.7 4 ,

d i -

i Comparisons to INCORE (computer code used by TVA for technical  !

j specifications surveillance) power distributions, measured rod ,

i~ worths, isothermal temperature coefficients, and boron letdown I

values were documented. The calculation package also documented ,

the main steam line break analysit with comparisons to the FSAR  !

j results. The inspectors observed that this calculation package 1 was prepared consistent with the FCF administrative manual l 1

procedure BWNT-0402-01, " Preparing and Processing BWNT- '

i Calculations," Revision 29.

The inspectors concluded th'at the Sequoyah reload package was j adequately documented, the mixed core issues addressed, and the

design inputs were adequately controlled and verified by the j licensee.

) (3) Crystal River 3 i

i The inspectors reviewed the reload process for Florida Power l

Corporation, Crystal. River Unit 3 (CR3), Cycle 11 with the reload PM and cognizant RELATE. From its review of the CR3 Cycle 11 FCDR, FFCD, and reload report, the inspectors observed the j following unique characteristics about the CR3 Cycle 11 fuel 3 design:

1 i (a) the longest cycle length to date (670 effective full power

! days (EFPD))

{ (b) the highest enrichment to date (4.96 w/o Um)

(c) first time use of Gd 4

{ The inspectors found that the FFCD was documented in calculation package 32-1239620-00 and this package was reviewed by the t

inspectors relative to the Cycle 10 FFCD in calculation package 32-1224533-00. The inspectors noted that the Cycle 10 and 11 analysis packages were prepared by the same individual and the same reviewer.

Documentation for the MICBURN 3, CASMO 3, and NEM0 calculations j was reviewed by the inspectors. Based on its review of several

specific items including the use of Gd for the first time, the

change in axial nodalization in NEMO for the new cycle, the choice i

of the K-effective bias or critical K-effective for NEMO and the cross section generation process, the inspectors concluded that j 4 the design process was effective for CR3.

The CR3 Cycle 11 reload process was examined by the inspectors l with the major focus on the engineering analysis which supported i i reload report BAW-2262, " Crystal River Unit 3 Cycle 11 Reload l

Report," January 1996. The reload report was examined for each of l

the functional areas (fuel system, nuclear design, thermal-j hydraulic, transient and technical specification /COLR). '

i

~

l.

N

! The use of GOTAC0 for fuel performance, the application of the statistical core design methods for thermal-hydraulics, and the main steam line break analysis were among the topics reviewed by

the inspectors.

1 The inspectors observed that the quality of the calculation packages varied significantly in terms of readability and in overall content and in documentation of_the verification process.

j The inspectors concluded, however, that the reload analysis i prccess for CR3 was adequate.

4

, (4) Arkansas Nuclear 1 4

The Entergy Operations, Incorporated, Arkansas Nuclear Unit 1 (ANO-1), Cycle 13 reload core wa an FCF full-scope design. The j design included 57 MARK-B8 and 60 MARK-B8ZL reload bundles, and 60

- MARK-B9 fresh fuel assemblies. Unlike TMI-l Cycle 10, the fresh i

assemblies were loaded in a symmetric checker board pattern l throughout the core that reduced the potential for t.rud deposition

$ on the fuel (see Section 3.4.b.2 of this report).

l

! The inspectors reviewed the ANO-1 Cycle 13 nuclear analysis with j the PM and RELATE. . The specific Cycle 13 fuel loading did not require any modifications to the reload methodology and the j

standard FCF reload analysis codes and methods were employed. The j inspectors reviewed the design files documenting the results of 5

the Cycle 13 calculations for the on-line computer inpitc (86-1232957-00), the physics data supporting the Cycle 13 site-j operation (61-1232956-00), and the generation of key safety i parameters for the evaluation of transients, reactivity effects and control rod worths (32-1232950-00).

While the inspectors review of the calculational files showed that, in several instances, unapproved computer codes were' l employed, the inspectors also noted that each of these t

applications were justified by an evaluation of the effects on the

, results of the nuclear analysis. The inspectors determined that the calculations were well-documented and the necessary l

l verification was indicated. l 1

[ c. Conclusions i

The inspectors discussed with each individual the technique used to l perform certain design functions and learned that no design instructions existed for either performing the analysis or for review and

verification of the analysis. The inspectors found that the responsible individual follows the analysis process documentN for the prior cycle. '

! The inspectors also noted that tte PM, RELATE, W the RALS inspectors 4

i share the responsibility for the reload analysis. The inspectors determined that FCF performed these analysis processes based on the l 1 methodology documented in the prior cycle calculation packages in ,

, conformance with the guidance in the administrative manual procedures. ,

c

i 3

The lack of detailed design instructions to perform the reload calculations could result in a nonconservative value that may not be apparent from following the previous cycle methodology as a template for the current reload calculations. The inspectors, therefore, concluded that the lack of detailed design instructions was a weakner- in FCF's reload design process.

However, the inspectors did not identify any instance where the lack of 4

design instructions caused or contributed to an error. Specifically, the inspectors noted that for the CR3 Cycle 11 reload, the first time use of Gd was handled appropriately.

3.4 Fuel Assembly Mechanical Desian

a. Inspection Scope The inspectors reviewed recent FCF experience and activities related to  !

fuel assembly bow and the distinctive crud patter (DCP) following THI-l Cycle 10 and CR3 Cycle 10 operation.

b. Observations and Findings b.1 Fuel Assembly 3ow According to FCF, all fuel assemblies experience some assembly bow.

Fuel assembly bow affects local margin to departure from nucleate boiling (DNBR) as a result of reduced cooling as the assembly-to-assembly gap was decreased, and reduced the margin to LOCA F -limits as a result of increased power peaking as the assembly-to-assembly gap was increased.

Because of the stronger fast flux gradient on the core periphery, the differential growth-induced bow was larger for fuel assemblies near the core-edge and, therefore, the core reload pattern typically employs a cross-core shuffle. FCF post-irradiation examination data taken over three operating cycles indicates that most of the assembly bow occurs during the first cycle and that the increase in later cycles was not j large.

I Because of recent control rod insertion problems experiencM by NRC PWR '

licensees, the inspectors reviewed an assembly bow event at the Ringhals 3 and a subsequent related event at Ringhals 4. Both of these plants included second generation Framatome fuel assemblies. In the Ringhals 4 incident an RCCA jammed in the dashpot region and did not fully insert following a reactor trip. However, the drop times (measured to the top of the dashpot) for all other RCCAs was within the value used in the safety analysis. Only fresh or once-burned fuel assemblies were located in +he RCCA locations.

i As part of the investigation of these incidents, UT and Eddy current i tests wer performed on eight fuel assemblies to evaluate swelling and I wear. F 'y-eight fuel assemblies were also inspected for bowing,

._._._ _ _ _ _ _ _ __ _._ _ _ _ _ _ _ __- _ . _ . _ _ _. _ _ _ ~

$" elongation and the extent of rod-to-nozzle gap closure. These tests

indicated larger than usual (both "S" and " banana" shape) assembly  !

1 bowing. Based on subsequent measurements and calculations, it was concluded that the assembly deformation was the cause of the failure of

the RCCAs to insert. A review of the second generation Framatome fuel j

indicated that, while the fuel rod, guide tube and hold-down system

designs were identical, the decreased grid height resulted in reduced lateral stiffness which increases the susceptibility to assembly bow.

FCF analysis of the assembly bow included the effects of (a) differential growth due to flux gradients and (b) creep amplification of existing deformatic.w. The resulting fuel assembly design modifications resulted in incre sed guide tube stiffness and a

reduced fuel assembly hold-down force.

The inspectors concluded that FCF was actively evaluating the Ringhals bowing events and their relevance to FCF fuel designs.

i b.2 Fuel Cladding Distinctive Crud Pattern

! Examination of the fuel pins following TMI-l Cycle 10 and CR3 Cycle 10 i operation indicated a distinctive crud pattern (DCP), on more than 182 j

i fuel rods at TMI-l and on four fuel rods at CR3, which was significantly different than the " normal" corrosion pattern. Nine fresh (one-cycle

' burned) fuel rods at TMI-1 with tho DCP had through-wall leaks and many of the other rods with DCP had significant wall-thinning.

FCF provided a detailed description of their investigation into the cause of the DCP observed at TMI-1 Cycle 10 and CR3 Cycle 10. The inspectors reviewed this material in order to identify (a) the factors j that could have potentially contributed to the DCP and (b) the actions 3I that FCF may have taken to prevent the occurrence of the DCP.

i s

For TMI-l Cycle 10, the DCP tended to occur in the sixth-axial grid-span i

(second from the top of the fuel assembly) and on the outer-row of fuel

pins of the assembly. Most of the fuel pins with DCP were located in one of eight symmetrically loaded "T" loading patterns. These "T" loading patterns consisted of four fresh high enrichment (4.0 to 4.75 w/o U23s) fuel assemblies, arranged with three assemblies across the top of the "T" and one assembly centered below. These "T" patterns were located on the core periphery in the low-flow core region with one assembly facing the reflector (see Figure 1 on page 35 of this report).

Following the TMI-1 Cycle 10 fuel failure presentations by FCF, the inspectors decided to evaluate the DCP event based on the following possibilities:

(a) the presence of crud in the coolant was caused by inadequate water chemistry (b) the deposition of crud on the fuel pins was accelerated due to higher clad surface temperatures in relatively low-flow regions

- .. = - _ _ - . - _.

(c) the corrosion failures (crud / oxide spalling) were caused by local boiling in or under the crud layer The inspectors pursued investigations into various aspects of the failures and possible mechanisms.

(1) Reactor Coolant System Chemistry During operation of a 177 fuel assembly B&W plant, a minimum ,

coolant pH level is maintained in order to prevent any significant I increase in the deposition of crud on the fuel rods. However, the recent aggressive 74-month high-reactivity core designs, like TMI-1 Cycle 10 and CR3 Cycle 10 have increased boron levels (to ~1500 parts-per-million (ppm) at startup) which has resulted in a reduction in the coolant pH to below the target value of ~6.9.

This pH reduction is generally compensated for by maintaining an increased lithium (L1) concentration in the coolant which increases the pH.

I In its letter to GPUN, "1991 Revision of BAW-1385 - Water '

Chemistry Manual 177 FA Plants. Lithium / Boron Control," November  ;

11, 1991, BWFC provided GPUN the recommended Li/ boron control for  ;

TMI-l Cycle 10. This recommendation was to maintain a pH of 6.9  !

initially and then, if necessary, to increase the pH to a higher l value later in the cycle. Because of the high initial boron concentration in the coolant for this high reactivity core, this would have required a Li concentration higher than the usual BWFC l recommended limit of Li <2.2 ppm. (The upper limit on Li was designed to protect against fuel cladding corrosion.)

In a subsequent letter to GPUN, " Lithium / Boron Control 1 Requirements," December 11, 1991, BWFC recommended a second and different Li/ boron control. This recommendation was to startup at l a Li concentration of 2.2 ppm and hold at this value until the pH l increases to the desired value above 6.9 (as a result of the decreasing boron), and then to decrease the Li in order to maintain the desired pH. Responding to a request from GPUN for BWFC to clarify its position, BWFC indicated that the second recommendation should be followed. ,

The inspectors concluded that the final BWFC recommendation, to limit the maximum Li to 2.2 ppm, resulted in a pH less than 6.9 and significantly increased the tendency for deposition of crud on the fuel. The low pH was of special concern at TMI-1 Cycle 10 because of the high level of " circulating crud" believed to be present due to (a) the five years of plant shutdown that had occurred earlier and (b) the fact that the reactor coolant system (RCS) makeup flow was less than 50% of nominal. A similar situation existed at CR3 Cycle 10, although the lowest pH was 6.6 and the level of circulating crud was believed to be substantially less.

" I In its letter to GPUN, "TMI Cycle 11 Lithium Levels - Revision to Previous Recommendations," August 28, 1995, BWFC relaxed the Li limit to 3.0 ppm for THI-l Cycle 13 in order to maintain a pH above 6.9 and thus reduce the potential for fuel rod crud deposition.

i' During the initial months of operation of TMI-1 Cycle 10, GPUN did not add Li to the reactor coolant but rather allowed lithium-7 to  ;

" burn-in" as a result of the B-10 (n, a ) Li reaction. Because of the high Cycle 10 boron concentration, this resulted in an initial pH of 5.7, which increased to 6.9 after several months of operation as a result of the Li buildup and the decrease' in boron.

Based on the above evaluations, the inspectors made the following 4

conclusions:

1 (a) The FCF communication to GPUN concerning the THI-1 Cycle 10 Li/ boron control was not clear and may have contributed to the decision by GPUN to ignore the recommendation and initiate Cycle 10 without Li. The inspectors considered this a weakness in the FCF/ customer interface.

(b) The FCF Li/ boron control guidance provided to GPUN did not ensure a pH >6.9 and provide adequate protection from crud 3 deposition. The inspectors considered this a result of a weakness in the FTI water chemistry program to establish adequate Li/ boron control procedures.

(2) Nuclear and Thermal-Hydraulic Design i

The nuclear and thermal-hydraulic reload core design is an important consideration in preventing fuel rod crud, since both

' low-flow velocity and high fuel rod power enhance crud deposition; in part, due to the reduction of solubility with increased temperature. In addition, since several of the mechanisms suspected of causing the DCP failures involved boiling under the accumulated crud, the relatively small margin of 10-20*F to coolant boiling (in the sixth-axial span where UCP cccurred) makes this design relatively sensitive to small increases (decreases) in fuel rod power (flow veluity). This was especially true in TMI-1 Cycle 10, where a B0C power-tilt of +15% occurred at the sixth-axial span in the quadrant of the core where the DCP occurred.

(This tilt is believed to be due to boron " hideout" in the crud.)

In t 'Jel assembly designs for both the TMI-l and CR3 core reloae, the highest powered fuel rods were located on the outside row of the fuel assembly, where the lowest assembly flow velocity occurs. The placement of the fresh high e richment fuel assemblies in the "T" loading pattern on the core periphery resulted in the unusual situation in which the highest powered rods in the core (they were in the highest eight percent) were located in the peripheral (lowest velocity) locations. In i addition, the relatively low burnable poison loading of both the TMI-l Cycle 10 and CR3. Cycle 10 cores resulted in higher critical l boron concentrations and consequently, lower reactor coolant pH.

i

.It was well known that fuel rod crud deposition is enhanced by 3 high fuel rod power, reduced flow velocity, and low coolant pH.

i However, the inspectors determined that several features of the 3

TMI-l Cycle 10 core design resulted in a significant reduction in i the margin in these parameters to fuel rod crud deposition. These d' include (a) the highest.-to-date critical boron concentration which i lowered the coolant pH and (b) the location of the "T" loading pattern of high powered fuel assemblies (with the highest-to-date

! enrichment of 4.7 w/o Vas) in a low-flow velocity region on the

! core periphery.

1

The inspectors concluded that this design weakness together with i

the -(allowed) operation at low Li, which further lowered the

coolant pH, resulted in DCP on the high powered rods in the "T" 1 i pattern loaded fuel assemblies. Based on the above, the j inspectors concluded that there was a basic weakness in the TMI-l

Cycle 10 core design. The nuclear, thermal-hydraulics and j materials analysis of the THI-l Cycle 10 core were considered weak i by the inspectors because the reload core could not accommodate

! the expected lower coolant pH values and protect from deposition of crud on the fuel rods.

l

! (3) Fuel Assembly Bow I t

l- Fuel assembly bow, due to guide-thimble / fuel-rod differential j growth, occurs preferentially on the core periphery where the flux l gradients are strongest. As the fuel assembly-to-assembly gap

, increases, the power in the fuel rods located on the outer row of

the assembly increase (~3% per 50 mil of gap closure).

4 In the TMI-l Cycle 10 core design, the "T" pattern of fresh high enrichment high-powered fuel assemblies were placed near the core -

, periphery which maximized the flux gradient across these j assemblies and the resulting assembly bow. The inspectors determined that this resulting increase in rod power would j increase the potential for crud deposition in exactiv the location

! where the DCP occurred; the outer row of fuel rods. For TMI-l Cycle 10 operation, both the location of the DCP failures in the

' outer row of the fuel assemblies and on the core periphery was consistent with the assembly bow mechanism which would have j

resulted in increased fuel rod power and local boiling.

The inspectors concluded that FCF failure to perform an evaluation j (including measurements where practical) of assembly bow as a i' potential mechanism for the deposition of the DCP was a weakness in its DCP true cause analysis.

l l

i I

)"

t (4) Neutronic One theory pursued by the inspectors was that the high clad

- surface temperature and local boiling was caused by the inability 3 to correctly predict the heat flux or pin power distributions in i the fuel assemblies with DCP. The inspectors evaluated the

likelihood that the, accuracy of the neutronics as applied to the

TMI-1 Cycle 10 core design was a contributor or the potential

cause of the fuel failures.

j! The inspectors were told by FCF that the majority of the pin

failures had occurred in fuel assemblies located in the fresh fuel "Ts" (a four fresh fuel assembly group, which has 3 assemblies

, face-adjacent in a row forming the top of a T, see Figure 1 on  !

page 35 this report). Within these "T" assemblies, the majority 2

of the pin failures occurred in fuel pins on the outer row (along

the water gap) of the assembly and in the corner pin where the l fresh assembly faces meet. The failures occurred in the top one '

third of the fuel rod (typically 100- to 130-inches from bottom of ,

i the fuel). The inspectors evaluated FCF's ability to accurately I

predict the pin power in these locations, i i )

i. The inspectors reviewed the TMI-1 Cycle 10 " measured power distributions" (seven axial values for each of the 57 assemblies containing fixed incore detectors) versus those predicted by NEMO.

! Particular attention was given to the comparisons in the sixth and

seventh axial location which corresponds to the elevation with the i crud deposition and fuel failures. Comparisons from four points

! in time throughout the cycle were examinei l The inspectors observed that for the assemblies of interest, NEMO j consistently over-predicted the power in the seventh level (on the ,

! average, 5-10% over-predicted) and more closely predicted the l power in the sixth level (on the average 1% under-predicted to 3%

i over-predicted). This would result in the temperatures being l cooler in the actual core in these locations as compared to the predicted core.  ;

l

, In addition to the " measured" vs NEM0 predicted power i'

distributions presented in the NEMO topical report, the inspectors

reviewed TMI-1 Cycle 9 radial power distributions. The inspectors l observed that the radial power was typically predicted within 2%

, (root-mean-squared (RMS) or standard deviations were less than 2%

for various databases). The inspectors also observed that the 4

power in the face adjacent fresh fuel was predicted as well as 1 other radial locations.

l

. The inspectors reviewed analytical power distribution benchmarks j such as comparisons between NEM0 and other nodal codes or between i i NEMO and PDQ. Data from TMI-1 Cycles 8, 9, and 10 was examined and the inspectors observed that, compared to the other code ,

! prediction, NEMO tended to predict the same or higher power in the

] .

l

face adjacent fresh fuel locations. Based on the above evaluations, the inspectors concluded that for the specific "T" assembly locations, there was no evidence to sugge=t that NEMO under-predicted the assembly power.

The inspectors also observed that there was nothing unusual about the TMI-l Cycle 10 assembly power distributions relative to the benchmarked cycles (in terms of control rod presence or unusual burnable poison configurations). The inspectors did note that the fresh assembly "T" was placed in a core location which h9 a steep power gradient (due to presence near the edge of the core). Based on the inspectors conclusion that NEMO was predicting assembly power reasonably well, the inspectors next focused on the NEMO pin power reconstruction methodology and NEM0's ability to predict the local pin power.

The inspectors reviewed single assembly CASMO 3 calculations and NEMO quarter core with pin power reconstruction calculations to determine what the pin power distributions were for current FCF designs. The inspectors observed that the newer designs, e.g.,

CR3 Cycle 11, have a flatter assembly pin power distribution and the peak / average pin power ratio was typically less than 1.10.

The older designs may have peak /aversge pin power ratios of up to 1.2. The inspectors observed that the NEMO calculations for the pin power ratios in the fresh "T" locations show a very large gradient across the assembly (e.g., 1.4 to 0.6 or 1.5 to 0.8).

The inspectors also noted that the peak pin power often occurs in the corner pin where the fresh assembly face meet.  ;

i The inspectors reviewed multi assembly (2x2 array) comparisons of NEM0 pin power distributions vs McNP and comparisons of NEMO vs CASMO 3. The inspectors observed that the differences were less  ;

than 2% and often less than 1%. The inspectors concluded that NEMO was reconstructing the pin power distribution satisfactorily I for these analytical comparison; however, the actual core pin power distributions were more challenging (steep gradients). The inspectors also noted that their concern about the NEMO accuracy partially arises from the lack of directly applicable benchmarks.

That is, FCF had no benchmarks which duplicate the TMI-1 Cycle 10 core design and the pin power distributions gradients present in these core designs.

Based on the above review of the neutronic aspects of the TMI-1 )

Cycle 10, the inspectors concluded that the NEM0 methodology can accurately predict the assembly and pin power distributions. Thus there was no evidence to support the hypothesis that the inaccuracy of the neutronic models contributed to the fuel corrosion failures.

I 1

i ..

However, the inspectors also concluded that the TMI-I Cycle 10 design challenged the methodology more than other FCF designs via the "T" placement of fresh assemblies. The TMI-I Cycle 10 design resulted in large pin power peaks in the corner pins and steep l power gradients across the assembly and these local conditions can l not be found in any of the NEM0 benchmarks; thus, the methodology j was stressed to near its maximum capability. Other FCF core '

designs have either not used the "T" loading, or used an "L" loading which had lower pin peaking. The inspectors concluded  ;

that these factors resulted in a weakness in the TMI-1 Cycle 10 reload core design.

4

c. Conclusions Based on the review of the DCP and associated fuel rod failures, the inspectors concluded that the FCF design process was also a contributing factor to the DCP. The inspectors noted that there were many features of the TMI-I Cycle 10 core design which by themselves were not '

significant contributors to DCP, but when taken together, resulted in a '

reduction in the margin to crud deposition and localized boiling. These factors included:

(a) increased critical baron and the resulting decrease in reactor coolant pH (b) inadequate Li/ boron control (c) inadequate communication to TMI-l regarding the establishment of the Li/ boron control (d) low coolant pH was not accommodated by the nuclear, thermal-hydraulic, and material design (e) challenged the NEM0 methodology by using "T" loading pattern (f) both the fuel core and lattice designs placed the highest powered rods in the locations of lower-flow velocity and higher assembly bow The inspectors considered the lack of an overall multi-disciplinary review process by FCF to identify and evaluate the synercistic effect of the changes that occurred in the TMI-I Cycle 10 core design to be a weakness in the FCF design process.

The inspectors also considered 'he lack of an evaluation of assembly bow as a potential contributing mechanism as a weakness in FCF's true cause analysis of the DCP.

3.5 Enaineerina Computer Procrans and Database

a. Inspection Scope FCF's use of engineering compute programs (ECPs) was contingent upon a tiered approach to computer program certification. Programs (codes) can be fully certified, conditionally certified, interim certified, and uncertified. Procedure BWNT 0402-01, " Preparing and Processing BWNT Calculations," Revision 29, June 1,1994, proscribed the use of certified programs in performing calculations, and stated that uncertified programs must have an independent verification of accuracy.

To evaluate FCF's use of ECPs, the inspectors reviewed the certification process for new versions of ECPs and the procedures for error reporting, correction, and access control.

. b. Observations and Findings 4

b.1 Software Certification Process

, The inspectors reviewed FCF's recent modification to the certification status of NEMO 7.6 from fully certified to conditionally certified. The change was due to development of NEMO 7.7, which included notification messages regarding the use of NEMO. The warning messages were excluded from NEM0 7.6. FCF chose to downgrade the certification status of NEM0 7.6 for this reason.

FCF procedure BWNT 0902-06, " Software Certification," Revision 16, March 31, 1995, proscribed an affectivity statement for the procedure applied to computer programs changing certification status due to errors within the code. The procedure requires changes to the certification status to be placed in the certification file. However, the inspectors determined that the change to NEM0 7.6 was not in the certification file.

FCF agreed with the inspectors findings and added the change to the

certification file. The inspectors concluded that the procedure was ambiguous because certification changes for any reason, including code errors, should be recorded in the certification file. FCF also agreed

, that the procedure governing ECP certification was ambiguous, and committed to modify the procedure.

The inspectors concluded that the ambiguity in the software certification procedure and the resulting incorrect information on NEM0 7.6 were considered weaknesses in FCF's computer code certification process, b.2 Engineering Computer Program Error Reporting The inspectors reviewed calculational files for consistency in ECP error reporting procedures. The Inspectors found that FCF had no specific procedures to address error reports in calculational files.

Furthermore, the inspectors determined that procedure BWNT 0402-01 did not distinguish between the use of conditionally certified and fully certified ECPs. However, FCF stated that use of a conditionally certified code required justification in the calculation file, particularly if error reports exist for the code.

s

,' The inspectors examined two cases involving the use of the NEMO.4.lRS code for TMI-l Cycle 10. NEMO.4.lR5 was a conditionally certified code which contained an error in the calculation of certain core average values. The first use of this version of NEM0 was found in BWNT 32-1219600-00, "TMI-1 Cycle 10 Flex Nuclear Data," September 17, 1993. The calculation file did not include a justification for the use of the code, even though an error record existed for the code. FCF agreed that this was an oversight and added a reference to the error in the calculation file.

The second use of this version of the NEMO code was documented in BWNT 32-1218970-00, "TMI-I Cycle 10 Nuclear data /650 Cycle 9," September 17, 1983, that also did not contain justification for use of the code. FCF also corrected this oversight during the inspection.

1 Based on these findings, the inspectors concluded that inconsistency in addressing ECP error reports was a weakness in the use of ECPs.

b.3 Benchmarking CASMO 3 and NEMO The applicability of the CASMO 3/NEM0 neutronic methodology benchmarking to the new FCF reload designs was reviewed by the inspectors. The NEM0 topical report BAW 10180-A, "NEM0 - Nodal Expansion Method Optimized,"

Revision 1, March 1993, was used as a reference by the inspectors. The inspectors focused on the potential concern that the new fuel assemblies and/or core designs may have differed significantly from the range of fuel designs that were benchmarked in the topical report.

The inspectors reviewed core loading patterns for TMI-l Cycles 8 to 11, CR3 Cycles 10 and 11, ANO-1 Cycle 13, Trojan Cycles 12 and 13, and Sequoyah-1 Cycle 9. The inspectors noted that operating cycle lengths have increased over time and that some current reload designs were for 650 to 670 EFPD (24 month cycle). The longer cycle designs necessitated the use of higher U u3 enrichments (up to 4.96 w/o) with corresponding higher beginning of cycle (BOC) soluble boron concentrations (up to 2300 ppm at B0C, HZP), or with higher burnable poison loadings. The inspectors noted that the core designs, now being loaded or soon to be loaded, used up to 6 w/o Gd as the burnable poison and that some assembly designs used both Gd and the B 4C lumped burnable poison.

Having determined the past and upcoming fuel designs, the inspectors focused on reviewing the range of designs that have been benchmarked to validate the CASMO 3/NEM0 methodology. The inspectors's primary concern was the ability of NEM0 to predict core power distributions. In addition to the benchmark results published in the NEMO topical report, the inspectors reviewed power distribution comparisons from the following:

(a) NEM0 predictions vs measured incore detector data (b) NEM0 predictions vs PDQ predictions (two-dimensional (20) quarter core diffusion theory results)

(c) NEMO predictions vs SIMULATE-3 predictions (alternate 3D nodal code results)

(d) NEMO predictions vs MCNP predictions (2D multi assembly Monte Carlo code results)

Based on this review, the inspectors concluded that NEMO can accurately predict core power distributions within the accuracy stated in the NEMO topical report. The inspectors also concluded that the current and upcoming core designs were within the range of fuel design configurations and core parameters that were verified in the topical report. Additional evaluations were performed by the inspectors as part of the TMI-1 fuel corrosion review (see Section 3.4.b.2 of this report).

c. Conclusions The inspectors concluded that the ambiguity regarding ECP certification changes found in procedure BWNT 0902-06, " Software Certification,"

Revision 16, March 31, 1995, and the resulting incorrect information on ,

NEM0 version 7.6 were considered weaknesses in FCF procedures. l The inspectors examined a number of calculation files to determine adequacy and consistency in ECP error reporting and found examples where ECP error reporting was either inadequate or ncn-existent. Based on j these findings, the inspectors concluded that FCF's inconsistency in i addressing ECP error reports was a weakness in the use of ECPs.

3.6 Enaineerina - Manufacturina Interface

a. Inspection Scope i

The inspectors reviewed specific aspects of the reload analysis / fuel  !

fabrication process including the role of the RELATE, the engineering / manufacturing interface and the use of the release authorization / applicable documents list (RA/ADL).

I

b. Observations and Findings '

The inspectors found that the first stage of a reload contract involved development of the contract requirements document (CRD), which contained initial licensee requirements for the fuel cycle, including a description of work to be performed, the scope of supply, quantity, and special features and customer requirements for the contract. After the CRD was issued, the preliminary fuel cycle design (PFCD) process '

typically developed three fuel designs, with the final decision left to the licensee based on operating requirements. After customer input, the fuel cycle design requirements (FCDR) document was prepared by the RELATE engineer.

The FCDR contained operating parameters, batch descriptions, control component descriptions, and other special and cycle-specific features.

The inspectors determined that the FCDR can be revised as necessary and as more specific data was received from the licensee, such as exact

.. . . - = . _ _ = _ _ -- ._ .- .- - .

j" cycle lengths. Concurrent with development of the FCDR and the final fuel cycle design (FFCD), the RA/ADL was prepared by the fuel design control group.

The inspectors reviewed example RA/ ADLS. The ADL defines a specific base or contract design by listing the applicable design definition

, documents, part names, and part numbers and was used to release the 4

following:

I (a) product documentation as part of a procurement package  !

(b) documents to establish code applicability i (c) documentation for equipment that requires professional engineer certification The RA was sent to manufacturing to authorize fabrication of components i specified in the ADL. Fuel manufacturing began as the final fuel cycle i design was nearing completion. During fuel assembly fabrication, I communication between manufacturing and engineering can be in the form i of concurrence requests (CR), deviation reports (DR), transmittal records (TR), and design change requests (DCR).

The inspectors reviewed a number of these records from current core design contracts. These transmittals were maintained by the fuel design control group. Procedure QC-1434, " Submittal of Manufacturing and Quality Documents to Fuel Engineering and Customers," described the  !

methods by which manufacturing, quality, and supplier documents were '

transmitted to fuel engineering or to customers for approval (CR) or for l

information (TR) and the means by which such transmittals and approvals were documented. Design change requests (DCR) were sent from engineering to manufacturing to propose design changes to previously released nuclear fuel and core components based on contract specified designs. I

c. Conclusion l

l The inspectors concluded that the interface between engineering and manufacturing and the use of the RA/ADL were performed in accordance with written instructions and satisfactorily completed for those 1

activities affecting quality.

, 3.7 Procurement

, a. Inspection Scope

^

The inspectors reviewed FCF's activities related to the procurement of material and services for use in safety-related products, the documentation that FCF maintained supporting the purchases of material, equipment, and services used in ..he manufacture -T safety-related products, and QA specification 09-1212-05, " Quality Assurance Program Requirements for Suppliers," Revision 5, February 11, 1992.

S 1

1

}-

i'

j. .

A' b. Observations and Findings  !

FCF procedure BWNT-1719-25, " Quality System Surveys & Quality Audits of BWFC Suppliers," Revision 4, October 28, 1994, specified the i

requirements for qualifying vendors to supply safety-related material, j equipment, and services. BWNT-1719-25 required that FCF perform an on-site quality audit to. verify implementation of the applicable quality program (for initial approval and triennially thereafter) and also to perform an annual evaluation to review the vendor's recent activities  !

with FCF. The status of vendors was maintained by FCF on the supplier status list (SSL). ,

The inspectors reviewed several FCF audit reports, supporting placement on the SSL, for Suhm Coil Spring Works (BWFC 92-36), FCBC Romans Plant (BWFC 94-17), Gray Syracuse Report (BWFC 95-9), and Carpenter Technology (BWFC 95-7). The audit reports documented the basis for the quality audits, results, conclusions, and findings. In addition, the report files contained documentation between FCF and the vendors which adequately resolved any findings. The inspectors concluded that the quality audits had been adequately performed and that FCF had maintained adequate documentation.

The following purchase orders for materials were reviewed by the inspectors:

(a) P0 40028, March 25, 1996, to Zircotube, Paimboeuf, France, for fuel clad tubing (b) P0 21872, January 23, 1995, to Gray Syracuse, Inc., for stainless steel castings to make end fittings (c) P0.40027, March 25, 1996, to Zircotube for guide tubes ,

(d) P0 40051, March 25, 1996, to Cezus, Montreuil-Juigne, France, for l end plug bar stock (e) P0 27542, July 5, 1995, to Cezus, Rugles, France, for Zircaloy '

spacer grid strip (f) PO 36317, January 19, 1996 to Ulbrich Specialty Strip Mill, Wallingford, Connecticut, for Inconel spacer gird strip (g) PO 4030, March 27, 1996, to Suhm for cruciform holddown spring section manufacturing services for plenum springs (h) P0 2849, November 16, 1993, to Inco Alloys International, Inc.,

for nickel alloy sheet for plenum s,arings

c. Conclusions The inspectors determined that specification 09-1212-05 invoked the requirements of 10 CFR Part 50, Appendix B, and that 10 CFR Part 21 was referenced in each of the P0s reviewed. The inspectors concluded that the procurements were performed in accordance with written instructions I and satisfactorily completed for thc4e activities affecting quality.

3.8 Fuel Fabrication

a. Inspection Scope FCF produced PWR fuel bundle and control components for B&W- and Westinghouse-designed reactors. They also produced fixed incore detectors for B&W , Combustion Engineering , and Westinghouse-designed reactors. The FCF fuel bundle iod matrix design was 15x15 array for Mark-89, Mark-BIO, and Mark-Bil (the B&W designs), and 17x17 for Mark-BW, the Westinghouse design.

FCF control components consisted of Mark-B burnable poison rod assemblies (BPRAs), Mark-B control red assemblies (CRAs), Mark-B axial power shaping rod assemblies (APSRAs), Mark-BW BPRAs (15x15 and 17x17) and rod cluster control assemblies (RCCAs). For the Mark-BW17 fuel bundles, FCF produced reconstitutable top nozzle assemblies, intermediate spacer grid restraint system, intermediate spacer grid assemblies, end spacer grid assemblies, debris resistant bottom nozzles, and advanced debris filter bottom nozzles.

During this inspection, FCF was not fabricating fuel rods or fuel bundle assemblies. Therefore, this portion of the inspection was limited to evaluating the fabrication of certain fuel bundle components.

b. Observations and Findings b.1 Burnable Poison Rod Assembly During this inspection, FCF was fabricating MK-BW type BPRAs for use in Westinghouse-designed PWRs. The fabrication encompassed manufacturing burnable poison rods and thimble plugs, and fastening them to an upper structure assembly with appropriate hardware. The inspectors observed the fabrication of burnable poison rods and upper structure assemblies, and assembly of BPRAs, as well as inspections performed by product quality (PQ) personnel and reviewed selective records pertaining to the quality of components used to fabricate the BPRAs. The team identified no adverse observations in this area.

b.2 Rod Cluster Control Assemblies l

During tnis inspection, FCF was fabricating RCCAs intended for use in Westinghouse-designed reactors. The main components of an RCCA were the spider, the control rods and the hardware to attach the control rods to l the spider. The team observed the activities related to the fabrication i of the control rods, and the final assembly of the RCCA including inspections by PQ, and reviewed records attesting to the quality of selected components. The team identified no adverse observations in this area.

l 1

s b.3 End Caps The inspectors reviewed the end cap fabrication process as specified by RC 9110-20, " Fabrication of Fuel Rod End Caps." Receiving inspection of the end plug rod was performed in accordance with QC-595, " Receiving 4 Inspection - Zircaloy Bar Stock and Tubing," Revision 4, October 13, j 1994. Manufacturing machined the end caps to the dimensions required by 1 the applicable drawing and . inspected 100% of the parts. PQ inspected the caps for dimensions and cleanliness. Accepted product was forwarded to material control for storage. PQ reviewed the route card for  ;

completeness and forwarded it, along with inspection data, to quality i verification for retention.

The inspectors observed that cleaning operations were performed in accordance with MA-549, " Miscellaneous Hardware Cleaning," Revision 4,  !

July 28, 1994. Paragraph 3.6 stated that within the limitations specified in the product routing or elsewhere in MA-549, the combination ,

and sequence of cleaning methods were left to the discretion of the  !

cleaning technician, However, the inspectors reviewed the training records of the cleaning technicians and determined they had no recorded training regarding the selection and sequence of cleaning methods. The inspectors considered this a weakness in the FCF manufacturing system. (This issue is discussed further in Section 3.12.b.2 of this report.)

b.4 Spacer Grids The inspectors reviewed the spacer grid fabrication process as specified i by RC 8500-11, "MK-BW Zircaloy Intermediate Spacer Grid Assembly Fabrication." Receiving inspection of the zircaloy strip was performed in accordance with QC-586, " Receiving Inspection - Zirconium Alloy Sheet or Strip," Revision 3, August 14, 1995. Manufacturing received released l spacer grid strip material, recorded the applicable information for traceability on the route card (RC) and assembled the spacer grid strips. The inspectors considered the manufacturing and inspection activities suitable for the production of grid spacers for safety related applications.

c. Conclusions Other than the weakness identified in the cleaning of end plugs, the inspectors considered the manufacturing and inspection activities suitable for the production of the safety-related items.

3.9 Calibration

a. Inspection Scope The inspectors reviewed ?he FCF program for ensuring that equipment was properly controlled and maintained in calibration. FCF maintained a gage control laboratory staffed by a gage control technician who was t

+  !

l responsible for controlling the status and maintaining the records i associated with equipment calibration. FCF calibrated most of the '

equipment used in the facility with the exception of certain specialty machines, such as the Baldwin tensile tester and the standards used to perform the calibrations, such as gage blocks.

b. Observations and Findings i FCF maintained the calibration records for the approximately 2,500 items in a hard copy filing system arranged by date of calibration (month or month / day). The gage control technician was efficient in the use of j this system and was able to provide the inspectors with the supporting i documentation for all calibrations reviewed. The inspector's I calibration reviews included the gage envelope dial indicator (QC14-17), l water channel gage standard fixture (QC-1949), micrometer inspection l gage (QC-1578), guide tube alignment gage (QC-4), lower end fitting casting template (QC-1579), lower end fitting casting inspection gage (QC-1578), and lower end fitting true position functional gage (QC-53).

FCF was able to provide documentation supporting traceability of the calibrations reviewed to the appropriate standards including the National Institute of Standards and Technology (NIST).

FCF procedure QC-1405, " Control of Measuring & Test Equipment," Revision 5, January 31, 1995, paragraph 4.5 stated, in part, that calibration services were classified as commercial grade and the measuring and test equipment were classified as commercial grade items. However, when the equipment was used to verify the capability of safety-related items, the equipment had a safety-related function to verify the parameters defined for acceptability of the safety-related item. Accordingly, FCF took actions to qualify the commercial grade service provided by calibration suppliers, to ensure suitability for the intended safety-related applicatior, by performing an audit of the suppliers technical and quality program. FCF used MIL-STD 45662A, " Calibration System Requirements," as the basis for the audits.

After successful completion of the audit, the calibration suppliers were listed on the SSL. The inspectors reviewed the FCF audit reports, supporting placement on the SSL, for AA Jansson, Inc., (BWFC 95-1),

American' Electronics Laboratory (BWFC 95-18), and Gage Laboratory (BWFC 93-35), which indicated that tre audits were adequately performed and documented. In addition, the inspectors reviewed P0s to calibration suppliers for AA Jansson, Inc. (P0 36241, January 12,1996), American l Electronic Lab (P0 38981, March 5, 1996), and Gage Lab (P0 30301, August  ;

17,1995), and determined that they contained adequate technical and quality requirements to assure performance of the requested service.

The inspectors concluded that FCF had taken adequate actions to verify and document the quality of the calibration services used to support safety-related work.

- . - _ . - - - - . - - - . - . - - - - . . .= . . . - - _ _ - - - . - . -

!, f l

During an inspection of production line equipment, the inspectors noted that the Baldwin tensile tester, identified as FCF QC-519, exhibited a

)

! "Satec" calibration sticker which stated that the machine had been

! calibrated on March 8, 1995, and was due for calibration on March 8, j i

1996. There was no FCF calibration sticker present on the machine. The '

inspectors noted that the machine was not administrative 1y out of {

! calibration due to procedure QC 800, " Gage Control," paragraph 5.2, i whi d allowed for the next calibration to be performed at any time 1 2

d uing the calendar month that the previous calibration expired such the the calibration on the Baldwin machine was acceptable for FCF work j throuh March 31, 1996.

l i

Discussion with FCF personnel indicated that the Baldwin machine was intended to be used in its present form for only an additional two i months and at that time it was to be modified. Based on consideration L for the short period of intended use, FCF had extended the calibration interval to eighteen months and had documented the extended calibration

interval in QC 800. FCF indicated that the basis for the extension from

) twelve months to eighteen months was ASTH E-4-94, Section 20, " Time

Interval Between Verifications" which stated, in part, that it was

! recommended that testing machines be verified annually or more

frequently if required and in no case shall the time interval between j verifications exceed 18 months.

[ The inspectors determined that FCF had not established a technical basis s for the calibration interval extension. The inspectors also noted that

! since the previous March 8, 1995, calibration had been performed by 3 Satec, FCF may consider Satec's input when establishing the technical l

basis.

FCF contacted Satec who provided a letter dated March 27, 1996, that I

stated that the decision to extend the current verification interval  ;

rested with FCF, that the machine could probably be used without a  !

j problem, but that FCF should have the machine's calibration reverified

as soon as possible. However, the inspectors determined that the Satec

letter did not provide a technical basis for extension of the 4

calibration interval.

j c. Conclusions j

! The inspectors concluded, based on the review of the FCF documentation i and the Satec letter, that FCF had not established a technical basis for i using the Baldwin machine past its original calibration date (and its administrative extension of March 31,1996). This was identified as a weakness in the implementation of the FCF calibration program.

3.10 Field Services Field Services was divided into three functional groups; engineering, outage operations and equipment operations. The Field Services quality assurance activities were governed by the same topical report that governed QA activities in other operations at FCF, document 56-1177617-

~

~

03, " Quality Assurance Program for B&W Fuel Company," May 30, 1994.

Procedure MA-580, " Specialized Tooling and Equipment," Revision 0, October 12, 1995, set forth the guidelines and responsibilities for the design, implementation, documentation and control of specialized tooling and equipment bearing on the outcome of manufactured product quality.  !

On the basis of its review, the inspectors concluded that the QA program !

for field service activities was adequate.

3.11 Quality Action Recuests The inspectors reviewed the quality action requests (QARs) program and the status of several open QARs and concluded that FCF personnel were sufficiently responsive to QARs.

3.12 Trainina

a. Inspection Scope 1

The inspectors evaluated the FCF training program in both the '

engineering and fabrication areas.

b. Observation and Findings b.1 Engineering The inspectors evaluated the adequacy of training in the engineering and services units primarily through interviews of management and staff, including interviews with unit managers in the fuel management and .

operations analysis, thermal and performance analysis and engineering '

analysis services units. The inspectors considered training to be made up of the following two general areas:

(a) direct, job-related technical training (b) non-technical or administrative training The inspectors also evaluated training for new employees and training in problem solving techniques. Procedure BWNT-1702-22, " Employee Training," Revision 16, October 10, 1994, provided requirements for

. training. The procedure stated, in part, that all initial job and on-going quality-related training shall be documented on a personnel training report (PTR). On-going quality-related training was described in the procedure as training on new or revised QA program manuals, the records management program manual, working instructions and new or revised quality-related policies, procedures, and forms.

In general, the inspectors found that no formal requirements exist for direct technical training, such as the use of engineering computer codes. Training on administrative procedures was the only training that

~

was officially documented. The inspectors examined available training I

} records for one engineering unit. Although substantial records existed i on administrative procedure training and software training, only one  ;

i record was found for training directly applicable to job performance.  !

}

b.2 Fuel Fabrication

{ f, FCF had divided the training of employees into two distinct categories,  !

inspection personnel and non-inspection personnel. The inspectors found

that procedures BWNT-1702-?2, " Employee Training," Revision 16, October )

1

. 10, 1994, applied to all FCF eaployees and that QC-1440, " Qualification j of Inspection Personnel," Revision 9, May 31, 1995, applied to personnel performing inspections to verify conformance to specified requirements j for the purpose of acceptability.

The inspector reviewed BWNT-1702-22 that stated that employee training was the responsibility of the management of each organization and that i each supervisor should evaluate the training needs of the employees within the group on a continuing basis. .All initial job and on-going quality related training was to be documented on a PTR. The specific

! training required by BWNT-1702-22 included initial job training to be

! accomplished prior to beginning work and documented on a PTR. The on-

! going quality related training was to be provided by the applicable L manager.

l The inspectors reviewed the training records for a selected group of

employees from various areas including QC, PQ, final assembly, gage i control, metallurgical laboratory, assembly room, machining, and i manufacturing. FCF provided records for the personnel performing i inspection activities which had documentation of the initial and-on-  ;

4 going quality related training. The inspectors concluded that the '

inspection personnel training was in accordance with the requirements of .

-QC-1440 and adequately documented. '

]

4 However, for non-inspection personnel, FCF could provide no records

! documenting either initial training or on-going quality training. FCF  ;

j management indicated that this training had occurred but there had  ;

previously been no requirement to document the completion of the

training.

c. Conclusion i Through review of FCF procedures and interviews with engineering management and staff, the inspectors found that no formal requirements

, exist-for direct technical training, such as the use of engineering

computer codes. Training on administrative procedures is the only i training that was officially documented. The inspectors considered this j lack of formal requirements for technical training and the limited

! incorporation of technical training to be a weakness.

i L

4 i

O The inspectors did not identify any examples of production activities having been affected by the lack of documented training. However, the inspectors concluded that the lack of documentation for the training of non-inspection personnel was a weakness in the implementation of the QA program.

3.13 Continuous Imorovement Procram The inspectors also interviewed the manager of the FCF continuous improvement program (CIP), part of the total quality management program at FCF. Employees trained in CIP generally lead or were a member on a quality improvement teams (QIT). Currently, 17 quality improvement teams have been formed. The CIP program was based on a seven step program which included identifying, defining and analyzing a problem and determining and implementing a solution. Some topics of recent QITs include eliminating sources of fuel failures and improvements in fuel rod welding. The final report on QIT B5 on fuel integrity was examined as an example. The Fuel Integrity QIT evaluated approximately 200 ideas submitted as the root cause of the unknown fuel failures. Formation of the inspectors, its processes for finding the root cause, the results of that effort, and recommendations for corrective action were included in the report. The program appears to be an effective method for improving product quality.

3.14 Entrance and Exit Neetinas In the entrance meeting on March 18, 1996, the NRC inspectors met with members of FCF management and staff, and discussed the scope of the inspection. The team also reviewed its responsibilities for handling proprietary information, as well as those of FCF. In addition the inspectors established contact persons within the management and staff of the applicable FCF organizations.

During the exit meeting with FCF management and staff, on April 17, 1996, the inspectors discussed its findings and concerns, as well as FCF's weaknesses. l Figure 1 1

l TMI-1 Cycle-10 Arrangement of the "T" Loading Patterns 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 i

3 A

B C

i i

D 2

i E F

4

'l 4

G

. H K

i, M

i j N i

J 0

P i R 1

a a

I i

f i

k a

{

l D

. i e I PARTIAL LIST OF PERSONS CONTACTED Andrews, B. Vice President, Fuel Engineering Armentrout, C. Lead, Quality Audits and Programs Bohart, J. President Carr, C. W. Vice President, Manufacturing and Service Coleman, T. Vice President Cudlin, J. Unit Manager, Analysis Services Deveney, R.C.

Engelke, W. Benchmarking / Methodology (Supervisor Nuclear Technology)

Manager, Process Quality Ford, J. T. Manager, Fuel Manufacturing Gardner, R. L. Manager, Quality Hanson, G.E.

Benchmarking / Methodology (Manager Fuel Engineering)

Harlow, N. Technical Specialist l Hobson, G.H. Benchmarking / Methodology (Principal Engineer)

Lamanna, L. Manager, Chemistry Engineer Matheson, J. E. Manager, Design and Development Mayberry, J. R. Manager, Product Quality McPhatter, F. Principal Engineer Meyer, G. Manager, Thermal and Performance Analysis ITEMS OPENED AND CLOSED Opened none l

Closed 99900001/90-01-01 NON burr on lower end fitting 99900001/90-01-02 NON radiographs as quality records '

99900001/91-01-01 NON visual inspection of tubes with felt plugs 99900001/91-01-02 NON issuance of CDR and CVAR 99900001/91-01-03 NON adequate guidance to operators and inspectors 99900001/91-01-04 URI root cause for the failure to detect felt plugs left in tubes g

I*

49 ACRONYMS USED l

l ADL Applicable Documents List ANO-1 Arkansas Nuclear Unit 1 '

l APSRA Axial Power Shaping Rod Assembly BOC Beginning Of Cycle BPRAs Burnable Poison Rod Assemblies B&W Babcock and Wilcox Company BWFC B&W Fuel Company l BWST Borated Water Storage Tank CDR Component Discrepancy Report CFR Code of Federal Regulation CFT Core Flood Tank CIP Continuous Improvement Process l

CNFP Commercial Nuclear Fuel Plant COLR Core Operating Limits Report CR Concurrence Request CR3 Crystal River Unit 3 CRA Control Rod Assembly CRD Contract Requirements Document CVAR Contract Variation Approval Report DCP Distinctive Crud Pattern DCR Design Change Request A Delta (Differential)

DNB Departure From Nucleate Boiling DNBR Departure From Nucleate Boiling Ratio DR Deviation Reports DRN Document Release Notice ECCS Emergency Core Cooling System ECP Engineering Computer Programs EFPD Effective Full Power Days FCDR Fuel Cycle Design Requirements FCF Framatome Cogema Fuels FFCD Final fuel Cycle Design FOAK First-0f-A-Kind FSAR Final Safety Analysis Report FTI Framatome Technologies, Inc.

Gd Gadolinium GDC General Design Criterion GPUN GPU Nuclear Corporation HFP Hot-Full-Power HZP Hot-Zero-Power LBVs Licensing Basis Values LHR Linear Heat Rate Li Lithium LOCA Loss-of-Coolant Accident MTC Moderator Temperature Coefficient NIST National Institute of Standards and Technology NRC U.S. Nuclear Regulatory Commission PCT Peak Cladding Temperature l

PFCD Preliminary Fuel Cycl- Design s

o t

PM Program Manager PO Purchase Order ppe Parts Per Million PQ Product Quality PSC Preliminary Report of Safety Concern PTR Personnel Training Report PWR Pressurized-Water Reactor QA Quality Assurance QAR Quality Action Request  :

QC Quality Control QIT Quality Improvement Team RA/ADL Release Authorization / Applicable Documents List RALS Reload Analysis and Licensing Services RC Route Card 1 RCCA Rod Cluster Control Assembly i

RCS Reactor Coolant System l RELATE Reload Licensing Analysis Task Engineer RPS Reactor Protection System i SBLOCA Small Break Loss-of-Coolant Accident j SRP Standard Review Plan SSL Supplier Status List ,

3D Three-Dimensional '

TMI-l Three Mile Island Unit 1 TR Transmittal Records TVA Tennessee Valley Authority 2D Two-Dimensional I

w/o Weight Percent i

1 i

a_