ML20207A034

Insp Rept 70-7002/99-04 on 990222-0312.Violations Noted. Major Areas Inspected:Corrective Action Program & Implementation of Nuclear Criticality Safety Corrective Action Plan

ML20207A034
Person / Time
Site:	Portsmouth Gaseous Diffusion Plant
Issue date:	05/21/1999
From:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
To:
Shared Package
ML20207A014	List: ML20207A012 ML20207A027 ML20207A034
References
70-7002-99-04, 70-7002-99-4, NUDOCS 9905260093
Download: ML20207A034 (29)

Text

n U.S. NUCLEAR REGULATORY COMMISSION REGION lli j

Docket No: 070-7002 l Certificate No: GOP-2 Report No: 70-7002/99004(DNMS)

Facility Operator: United States Enrichment Corporation Facility: Portsmouth Gaseous Diffusion Plant Location: 3930 U.S. Route 23 South P.O. Box 628 Piketon, OH 45661 Dates: February 22 through March 12,1999 Inspectors: Timothy Reidinger, Senior Fuel Cycle inspector Jack Davis, Nuclear Process Engineer Yawar Faraz, Project Manager Approved By: Patrick L. Hiland, Chief Fuel Cycle Branch Division of Nuclear Materials Safety i

9905260093 990521 }

PDR ADOCK 07007002 i C PM j.

e EXECUTIVE

SUMMARY

United States Enrichment Corporation Portsmouth Gaseous Diffusion Plant NRC Inspection Report 70-7002/99004(DNMS)

The inspection involved a review of the plant staff's implementation of the Corrective Action Program and application of the Corrective Action Program to the completion and closure of actions prescribed in the Nuclear Criticality Safety Corrective Action Plan.

Corrective Action Proaram o The inspectors determined that the corrective action processes and procedures provided adequate guidance and were, in general, implemented accordingly to maintain an effective corrective action program. The procedures clearly articulated the corrective action process requirements; i.e., documenting and classifying problem reports, developing and documenting corrective action plans, and tracking action steps through to timely completion. Trend reports were provided to appropriate levels of management and were reviewed for appropriate corrective actions. In general, appropriate followup actions were taken to verify implementation of the corrective action steps. The program appeared to include the necessary processes to capture all significant issues at the plant. The program was successful in capturing significant issues resulting from an intemal self-assessment in order to correct identified program deficiencies.

(Section C1.1)-

Imolementation of the Nuclear Criticality Safety Corrective Action Plan e The inspectors identified a violation involving incomplete or inappropriately completed activities associated with the implementation Nuclear Criticality Safety Corrective Action ,

Plan, Task 3. The violation included inadequate identification or documentation of j process upset conditions, controls relied upon for nuclear criticality safety, and the I verification of relied upon calculations and assumptions. . ? violation indicated inadequate rigor in the performance of some corrective actions and inadequate oversight or review of other corrective actions. (Section N1.1, Task 3) e The inspectors identified a violation, in that, the plant laboratory staff conducted analyses of nuclear criticality safety-significant samples without a prior nuclear criticality safety evaluation of the operations and without the identification of controls necessary to ensure the proper implementation of the double contingency concept. (Section N1.1, Task 3) e The inspectors identified a violation, in that, the plant staff failed to identify and control a physical feature relied upon for double contingency. (Section N1.1, Task 3)

-* The inspectors identified an Unresolved item related to the handling of issues developed  !

as a part of an independent review of implementation of the Nuclear Criticality Safety Program. (Section N1.1, Task 8) ,

2 i s

• The inspectors identified a violation, in that, the plant staff made significant changes to the Nuclear Criticality Safety Program Corrective Action Plan without: 1) determining if the changes were warranted; 2) determining if the changes would still ensure that the original root causes would be corrected; and 3) securing a review and approval of the changes from the Corrective Action Review Board. (Section N1.2) l I

3

4 Report Details

1. Plant Support C1.0 Corrective Action Process C1.1 lmolementation of the Corrective Action Proaram

a. Insoection Scooe (40500)

The inspectors reviewed the Corrective Action (CA) Program processes and procedures l used to identify, track, and correct problems and deficiencies. The inspectors j interviewed plant staff to determine the effectiveness of the CA processes and whether I the reporting of plant problems had management support. The inspectors observed activities in progress and interviewed responsible operators and first-line managers (FLM) in the Cascade Buildings. In addition, the inspectors reviewed the following procedures relating to the CA Program to determine whether the procedures were consistent in definitions, responsibilities, actions, and references.

• XP2-BM-Cl1031, Rev. O," Corrective Action Process," dated June 15,1998; e XP2-BM-Cl1030, Rev. O, " Problem Reporting," dated June 15,1998; )

e XP4-BM-Cl1002, Rev. O, " Problem Report Screening Process," dated June 15, 1998; e XP4-BM-Cl1001, Rev. O, " Process for Combining and Closing 'Same As' or

'Similar To' Problem Reports," dated June 15,1998; e UE2-QA-Cl1036, Rev. O, " Root Cause Analysis," dated December 4,1997; '

e XP4-SF-SF1110, Rev. O, " Plant Shift Superintendent Actions on Problem Reports," dated June 15,1998; and e UE2-TO-RM1030, Rev. O, " Records Management Program," dated March 19, 1999,

b. . Observations and Findinas in response to various inspection findings, Intemal self-assessments, surveillance program findings, and self-identified issues, the senior plant managers and associated ~

plant organizations initiated a comprehensive and extensive action plan to enhance the overall effectiveness and quality of the CA Program in June 1998. A Quality of Operations Plan (QOOP) initiative was implemented with the overall objective of having a comprehensive CA Program that was pro-actively utilized by all plant organizations.

Senior plant managers informed the inspectors that the CA Program was redesigned to )

better focus plant resources and to improve problem reporting and identification within all plant organizations. Changes included:

• Re-engineering the CA Program to make the program more responsive to plant staff; e improving the trending program to ensure adverse trends were identified properly; ,

e improving the screening of problem reports (prs) to identify significant conditions; e Focusing the CA Program on the reduction human errors; and e improving the rigor of investigations of significant prs.

4

)

l l The inspectors reviewed each of the elements and the proposed CAs for the CA Program which detailed approximately 60 action steps required for completion and closure. The inspectors reviewed the implementation of the following elements of the CA Program action steps contained in the QOOP:

1) Corrective Action Management reorganization;

2) Problem Reporting and Record Keeping;

3) Problem Report Screening;

4) Root Cause Analyses and Corrective Action Plans;

~5) Trend Analyses;

6) Problem Report Resolution Timeliness; and

7) Corrective Action Review Board Meeting Minutes.

Corrective Action Manaaement Reoraanization An administrative change was approved in June 1998 to move the functions of the Commitment Management (CM) manager from the Administrative Support Organization to a senior management position reporting directly to the general plant manager. As a part of this change, the position of CA manager, reporting to the CM manager, was created to assist in the technical oversight of the CA Program. The CA Program organization also included five support staff members who coordinated administrative activities and processed prs into the Business Prioritization System (BPS), a computer

' database. The CA Program staff also generated associated trend charts for compilation into a "CA Program Report" which was provided to senior plant managers on a monthly basis. Discussions with the CA and CM managers indicated that improved problem reporting performance, particularly in the areas of problem identification and resolution, was noted since the QOOP initiative was implemented in 1998.

The inspectors noted that the revised CA Program was implemented in late 1998 and was undergoing a maturation process. The senior plant managers informed the inspectors that the CA Program still required short term management attention and refinement to be fully effective.

Problem Reportina and Record Keepina l The CA Program required that a PR shall be written for safety or quality issues. The implementing procedures required, in part, that plant staff identifying an issue, meeting the criteria for the issuance of a PR, shall document the issue in a PR to ensure senior plant management review.

The inspectors randomly selected and discussed the problem reporting system with  ;

plant staff from the Commitment Management, Operations, Nuclear Safety, and  ;

independent Assessment Group (IAG) Organizations. The inspectors noted that the i plant staff were familiar with the procedures and requirements for problem reporting,  !

and that the plant staff wrote prs for identified deficiencies. A review of prs from these j groups confirmed that plant staff were knowledgeable of the implementing requirements  ;

of the problem reporting system. Inspectors noted that prs had been documented on a variety of plant issues identified by plant staff. The plant staff informed the inspectors that the CA Program was easy to use and that senior plant managers were supportive of ,

the plant staff's efforts to identify problems in the plant. The ins ~p ectors encountered no

[

l 5

J

n

]

l l

evidence from plant staff which might Indicate any suppression of or resistence to problem reporting by senior plant rnanagers.

The inspectors reviewed approximately 50 PR evidence packages (records) that were maintained in the PR system in three categories; open status (meaning partially completed actions steps within a CAP); closed status (meaning all action steps were completed); or partial status (meaning CAP action steps were under developmerit). The inspectors noted that the PR records were complete (with appropriate information and signatures) and were maintained in accordance with plant quality-record requirements and PR procedures. For example, the inspectors noted that the Plant Shift

]

i Superintendent (PSS) reviewed all prs and documented a review of the as-found conditions by signing the " Justification / Comments / Actions" section of the PR, as required. The inspectors noted that critical PR information was entered appropriately into the BPS.

Problem Report Screenina The inspectors reviewed PR screening procedures and determined that the procedures )

contained definitive criteria for identifying and classifying prs. The procedures stated, j in part, that any plant staff or subcontractors discovering hardware or activity problems shall write prs. Examples of problems included, but were not limited to: problems relating to accidents or events associated with nuclear safety, industrial safety, production, or production-support systems, structures, components, activities, or services. In addition, the procedures stated that the plant staff shall initiate a PR in situations when any question existed as to whether or not to initiate a PR.

The inspectors noted that the procedures required a review of the prs by a screening committee. The PR Screening Committee (PRSC) was comprised of a PR Screening Administrator (PRSA) and representatives from Nuclear Regulatory Assurance, and Safety, Safeguards and Quality (SS&O) Organizations. The PRSC classified prs or reevaluated previously submitted prs as either Not Quality Related (NOR), Conditions Adverse to Quality (CAQ), or Significant Conditions Adverse to Quality (SCAQ), as appropriate. In addition, PRSC made recommendations for closing prs that were determined to have been sufficiently resolved or which required no further action. The CM and CA managers indicated in discussions with the inspectors that prior to the QOOP initiative, the plant staff classified many issues as SCAQs based on an incomplete understanding of the limited guidance provided in earlier CA Program procedures.

The inspectors also reviewed approximately 40 PR packages designated as SCAO issues and 60 recently closed prs designated as either CAO or NOR issues and did not identify any recent situations where a SCAQ issue was incorrectly classified as either a CAQ or NQR issue. The inspectors reviewed PRSC recommendations and noted that PR classifications were consistent with procedural requirements. The inspectors' review of the revised procedures indicated that the procedures provided adequate guidance to ensure that the SCAO classification was correctly applied to "significant" or programmatic issues only. The CA and CM managers indicated that the revised procedure ensured that PRSC identified and prioritized all"significant" problems while at the same time allowing non-significant problems to be dealt with on an as-needed basis.

6

~

l To assist in the screening process, the CA Program Organization developed a procedure to combine "same as" or "similar to" prs for significant issues, whereas, I previous procedures required each PR to be treated individually. The revised procedures allowed senior plant managers to combine SCAQs dealing with the same problem so that the investigation and associated CAP addressed more than the individual ' problems. The procedure appeared to allow the senior plant managers to better manage, assess, and resolve repeat SCAQs.

The inspectors reviewed 50 prs written in 1998 and 1999 to assess the effectiveness of revised procedures in combining similar prs. The inspectors determined that similar prs were consolidated appropriately and the procedures were effective in eliminating the need for handling multiple prs for similar significant issues.

In addition, the inspectors reviewed another 50 selected prs written from August 1998 to the present to determine whether any prs that warranted stopping work were identified by plant staff. Approximately 5 of the 50 prs were written to address one of the following: 1) unsafe work practices or conditions that could result in serious injury or death to an employee, contractor, or member of the public; 2) conditions or practices that could render a safety system inoperable; and 3) conditions or practices that could )

result in damage to plant safety equipment. The inspectors noted that for those cases l when safety or operability was in question, or when unsafe work practices were noted; j the applicable process and work practices were immediately stopped or the associated equipment was declared inoperable.

Root Cause Analyses and Corrective Action Plans:

J The CA Program required a Corrective Action Review Board (CARB) to review and approve the Corrective Action Plans (CAPS) that were prepared and submitted by the responsible PR initiators. In addition, the CARB ensured the CAPS adequately identified and addressed the root cause(s). The inspectors noted that the plant staff used the Taproot guidelines and methodology for conducting root cause analyses.

During the month of January 1999 approximately 55 percent of the submitted CAPS were rejected by the CARB. The inspectors reviewed the rejected CAPS to determine the root causes or contributing factors that precipitated the CARB rejections. The inspectors determined that the rejection methodology used by the CARB was valid. The inspectors noted that the rejections were mostly due to inadequate Taproot evaluations. The CARB rejected Taproot evaluations which failed to include complete root cause analyses or that identified incorrect causal factors. For example, the inspectors reviewed a PR entitled, " Compliance Plan Concerns (PR-PTS-98-06326)."

The inspectors noted that the CARB rejected the associated CAP, in part, for the following reasons: 1) causal factor (CF) No. 4 failed to match contributing cause;

2) CF No. 3 was located on the wrong location on the Event and Causal Factor (E&CF)

Chart; and 3) the CAs for CF No. 2 were too narrowly focused. The CAP for the PR entitled, " Operations Concems," (PR-PTS-98-07651) had been rejected, in part, for the following reasons: 1) the root cause was incorrect; and 2) the CA stops needed revision.

The CA and CM managers discussed the high number of CAP rejections from January 1999 with the inspectors. The inspectors were informed that the CARB recently had placed an increased emphasis on the quality and the extent of SCAQ investigations and the root cause analyses, requiring a higher quality product from Taproot 7

i

investigators. As a result, the Taproot investigators were still gaining experience in conducting effective investigations, particularly when conducting the more rigorous root cause analyses demanded of SCAQs. The senior plant managers indicated that the high number of CAP rejections were expected to taper off as the Taproot investigators became more proficient in the methodology by conducting more investigations. The senior plant managers were asked whether the CARB had conducted a Taproot analysis to identify all of the root causes related to the high number of CAP deficiencies.

At the conclusion of the inspection, the plant staff wrote a PR to address this issue.

The inspectors reviewed several training records of the plant staff assigned Taproot investigation responsibilities and noted that Taproot training was recently provided in February 1999 to a selected number of plant staff to augment the core group of Taproot investigators. The training included: 1) the completion of a two-day Taproot training class; 2) the completion of an advanced five-day Taproot training program; and 3) the completion of a two-day assessment training course. in addition, the CM manager provided a guidebook to the Taproot investigators, entitled, "SCAO Investigation, Corrective Action Plan Development, and End Point Assessment Guide," as part of the overall training effort. The inspectors reviewed the guide and determined that the document provided appropriate guidance to the investigators for analyzing prs, for I identifying root causes and related causal factors and for developing diagnostic event i flow charts. The guidelines provided step-by-step instructions to assist the investigators in determining the CAs necessary to address and mitigate the root causes. The CM manager indicated that each plant organization had future plans to identify an additional l

cadre of technical staff to perform evaluator or investigator functions.

Trend Analyses The CA Program, by procedure, established a process for trend analysis to identify, document, and report quality trends and to serve as an oversight tool in assessing plant regulatory performance. Discussions with the CM manager indicated that trend analyses were conducted using the BPS. The number of human performance errors, process deficiencies, and equipment defects were tracked and analyzed over consecutive periods of time to determine the presence of positive or negative quality trends. In addition, the CM manager informed the inspectors that issues identified in the trend program included, but were not limited to, those discovered through intemal and extemal audits, appraisals, assessments, self-assessments, prs, event reports, organizational surveillances, negative performance indicators, equipment failures, and other issues pertaining to regulatory requirements, as directed by management.

Trend charts were developed on a monthly basis to identify a quality trend in issues which may be based on items such as root cause, conditions, processes, human performance issues, and requirements. The CM manager incorporated the applicable '

trend charts into a monthly CA Program Report to assist senior plant managers in identifying groups or departments where adverse trends were occurring. The CA Program Report listed types and sources of problems and provided information on problem reporting by organization or department, with overdue actions listed, and data on the " Top Focus Areas." The " Top Focus Areas" outlined current trends in some of the following areas: 1) PR extensions; 2) training deficiencies; 3) uranium hexafluoride releases; and 4) nuclear criticality safety approval (NCSA) deficiencies. Another area of interest outlined in the CA Program Report was a category characterized as " Tier 1" Human Performance Errors (HPEs). This category included: 1) SCAQs relating to  ;

8

1

. o HPEs; 2) Technical Specification Requirement violations; 3) nuclear criticality safety violations; 4) regulatory violations; and 5) lock-out or tag-out violations.

The inspectors noted that senior plant managers used the BPS database to help identify an adverse trend in HPE in January 1999. The number of HPEs had increased from ten in November 1998 to approximately 14 in both December 1998 and January 1999.

The plant managers determined that contributing factors for the increased number of HPEs were a failure to follow directions and inadequate procedures. As a result, the senior plant managers initiated a site wide standdown (full-staff meetings with all plant organizations) in February 1999 to heighten the plant staff's awareness of the need to conduct operations in accordance with plant procedures and regulatory commitments.

The standdown addressed recent significant human performance errors including:

1) in correct safe mass calculations; 2) improper valve actuator installations; 3) improper modification of load-alarm relay tests; and 4) inadequate nuclear criticality safety approvals. In addition, the senior plant managers provided goals and performance expectations to the plant staff to reduce the number of HPEs and to improve regulatory performance.

The inspectors reviewed the adverse trend reports and other performance indicators and determined that: 1) trending charts included input from all problems, including minor  ;

issues; 2) repetitive issues or problems were being identified; 3) repetitive PR issues j were being escalated to ensure further analysis for improvement, replacement, or modifications; 4) trends included systems, equipment, and personnel performance issues; 5) results of PR trending were reported to senior plant managers for action; and j

6) PR trend reports provided clear conclusions and analysis of data for senior plant managers review. The inspectors also identified that some adverse trend chart analyses were not sufficiently comprehensive to provide an overall bases for the identified adverse

. trend. The CM manager agreed with the observation and wrote a PR to address the issue. The inspectors determined that the use of adverse trends and CA Program Reports appeared to help build site-wide awareness and understanding of the CA Program.

To confirm the accuracy of various adverse trend charts, the inspectors requested that q the CA staff process the BPS database using variable BPS parameters specified by the inspectors. The resulting information confirmed the monthly CA Program Report data and the consistency of conclusions drawn by the senior plant managers.

Problem Report Resolution Timeliness The CA Program established processing requirements and due dates for prs. The procedures required that senior plant managers focus appropriate resources on investigating and correcting SCAQs first, and allowed responsible organization managers to review SCAQs and CAQs, and establish the date for completing necessary actions.

The responsible organizational managers were required by procedure to provide a response due date for SCAQs and CAQs not to exceed 28 and 45 calendar days, respectively (no required due dates were specified for NORs). The responsible plant managers also must report specific CA steps that were completed or request an extension from the CM Managerin order to reschedule a completion date beyond the original due date. An End-Point Assessment was required (typically six months after closure of a SCAQ) to determine the effectiveness of the CAs.

9

n

~

The inspectors reviewed selected CAQs and SCAQs for timeliness and completeness.

Problem reports categorized as CAQs were generally completed and closed in a timely manner, often within a day. The inspectors noted that the time taken to close SCAQs improved over the past seven months. A positive trend was noted in that completion timeliness went from an average of 20 months in August 1998 to less than 10 months in January 1999. This trend was identified in the February CA Program Report.

However, the inspectors also noted a slight increase in the number of SCAO End-Point Assessments that were overdue in January 1999, as well as an increase in the number of open SCAO-Actions (Individual action steps within an open CAP). The inspectors determined that these negative trends would ultimately have a corresponding negative impact on resulting SCAO closure timeliness. The inspectors noted that the average time to SCAQ-Closure could potentially increase, reversing the previously noted positive trend.

The CM manager informed the inspectors that the senior plant managers were striving for a goal of zero "Past-Due End-Point Assessments." Discussions with the CM manager indicated that the negative trend in overdue SCAQ-Actions and Past-Due End-Point Assessments was primarily due to management focusing attention on increasing the effectiveness of Taproot investigations and improving the quality of the CAPS. The senior plant managers informed the inspectors that the focus was expected to result in more accurate problem-solving. The net result was expected to improve both effectiveness and timeliness of PR resolution.

Correction Action Review Board (CARB) Meetina Minutes Per the CA Program procedures and CARB charter, the CM manager was required to implement the CARB charter and corresponding requirements to document and retain CARB meeting minutes for review by the senior plant managers.

The inspectors reviewed selected CARB meeting minutes from June 1998 to the present to compare the reported activities with the requirements stated in the procedure and charter. The inspectors noted that the meetings were attended by the required personnel and satisfied the quorum requirements. The meetings included discussions of SCAQs and the corresponding CAPS which were developed from prs, audit assessments, and surveillance findings.

The inspectors noted in the meeting minutes that the CARB members appropriately questioned proposed action items. The inspectors later questioned the CARB chairperson and determined that the CARB chairperson was knowledgeable of the issues discussed. The inspectors determined that the minutes were consistent with the procedure and satisfied the requirements of the charter.

The inspectors identified that the meeting minutes clearly specified actions to be taken, the plant staff responsible for the identified actions, and the final decisions for previously identified actions, with one exception. The CARB meeting minutes indicated a {

recognition by the CARB members of tne negative trend in the increasing number of i deficient CAP and Taproot assessments. However, the minutes did not indicate any pro-active measures by the CARB to deal with the causal factors associated with the negative trends.

10 ,

J

c. Conclusions The inspectors determined that the CA Program, processes, and precedures provided )

adequate guidance and were implemented accordingly to maintain an effective CAP.

The inspectors noted that the procedures clearly articulated the CA process requirements; i.e., documenting and classifying prs, developing and documenting CAPS, and tracking action steps through to timely completion. Trend reports were provided te appropriate levels of management and were reviewed for appropriate CAs. Appropriate followup actions were taken to verify implementation of the CA steps. The CA Program appeared to include the necessary processes to capture all significant issues at the plant.

The program was successful in capturing significant issues resulting from an internal self-assessment in order to correct identified program deficiencies.

N1.0 Implementation of the Nuclear Criticality Safety Corrective Action Plan N1.1 Nuclear Criticality Safety Corrective Action Plan

a. Insoection Scope The inspectors reviewed the certificatee's implementation of CAs as specified in the Nuclear Criticality Safety (NCS) CAP. The inspectors assessed the status of CAs i relative to the information provided in NCS CAP Quarterly Status Report, dated )

February 1,1999 (GDP 99-0020). The following ten NCS CAP tasks were reviewed: j Task 3: Priority 1 Nuclear Criticality Safety Evaluation and Approval Upgrades; Task 4: Enhance Nuclear Criticality Safety Approval Training; Task 6: Complete a Comprehensive Root Cause Analysis; Task 8: Vertical Slice Review of the Nuclear Criticality Safety Program; Task 11: Outside and Independent Assessments; Task 13: Revise Training Program for Site Personnel;  ;

Task 15: Plant Changes and Configuration Management; l Task 19: Continued Use of Nuclear Criticality Safety Review Board;  !

Task 20: Nuclear Criticality Safety Field Operational Assistants; and j Task D.8: Nuclear Criticality Safety Oversight on the Operating Floors. I b, Observations and Findinas Task 3: Priority 1 Nuclear Criticality Safety Evaluation and Acoroval Uoarades i

The inspectors reviewed two of the eight Priority 1 Nuclear Criticality Safety Evaluation and Approvals (NCSE/A) which had been completed and implemented. The NCSE/As were:

1) NCSE/A 0705-015, " Waste Water Treatment (Microfiltration System)," dated April 2,1998; and

2) NCSE/A Plant 079, " Opening Equipment Containing Greater than a Safe Mass of Uranium-Bearing Material," dated November 7,1997.

Nuclear Criticality Safety Evaluation and Approval 0705-015 was prepared for operations involving solutions, a type of operation that normally had a higher probability of abnormal operations or operator errors which could lead to a critical excursion. Nuclear Criticality 11 J

l l

l Safety Evaluation and Approval Plant 079 involved the handling of fissile material in greater-than-safe-mass quantities, an operation that was classified as being singly contingent.

The inspectors reviewed the evaluations and approvals concentrating on the double contingency control schemes and logic flow since the evaluation portion of the analyses  ;

established the safety basis of the operations. Although an improvement was noted in the written analytical discussions and the overall clarity of the documents, the inspectors determined that not all credible assumptions or scenarios were fully documented or analyzed for either of the nuclear criticality safety evaluations (NCSE).

The inspectors noted that NCSE 0705-015 referenced a calculation (POEF-520-94-129) j which demonstrated the suberiticality of the microfilter membrane modules. However, l the inspectors determined that the calculation failed to consider the neutron poison 1 effects of the chlorine component in the polyvinyl chloride (PVC) piping on the K-effective of the system and failed to consider whether the chlorine component could be bounded by the representative model developed. As such, the chlorine component was  !

relied upon for sub-criticality even though the chlorine's reactivity worth was unknown. In addition, the NCSE/A did not include controls to ensure the minimum chlorine concentration or intended distribution.

In followup discussions with the NCS manager, the inspectors learned that the chlorine worth was not determined for this evaluation because a previous calculation, not referenced by NCSE 0705-015, evaluated experiments using PVC and determined that a positive K-effective bias in the computational code resulted. The inspectors reviewed the previous calculation, POEF-520-94-36," Criticality Safety Analysis of a Preliminary Design for an F-Can Secondary Container," and determined that the reasoning for and applicability of the calculation to NCSE 0705-015 was flawed. Although the analysis compared calculation results to actual experiments involving PVC, no conclusions could be made from the evidence presented that the positive bias was due solely or in part to i the chlorine cross-sections. The referenced calculation was concemed with the validity of the chlorine cross-sections in the ENDF/IV cross-section libraries being used rather l than the chlorine worth in PVC as related to neutron absorption characteristics. Also, since the chlorine neutron absorption cross-sections were energy dependent, a comparison of the similarities or differences between the energy spectrums of the two representative cases would be necessary to demonstrate the applicability of the study to NCSE 0705-015.

Technical Safety Requirement 3.11.1 requires, in part, that the NCS program shall be established, implemented, and maintained as described in the Safety Analysis Report (SAR) and shall address process evaluations and approvals. Safety Analysis Report Section 5.2.2.3, " Process Evaluation and Approval," required, in part, that a documented NCSE shallinclude a determination of the credible process upset conditions, and an identification of the assumptions and equipment needed to ensure NCS. The failure to perform an evaluation of or document in the NCSE the presence of and applicable controls for materials (e.g., polyvinyl chloride) involved with the operations which could affect nuclear criticality safety a Violation (070-7002/99004-01a).

Based upon the incorrect reliance in the NCSE on the PVC calculation, the inspectors reviewed actions taken by the plant staff to ensure that other relied upon calculations were correct and appropriate. NCSE 0705-015 identified at least ten separate 12

p calculations that directly supported the conclusions. The referenced calculations included:

1)- POEF-520-94-129, " Criticality Safety Analysis of the Waste Water Treatment Facility, X-705;"

2) POEF-520-94-55," Criticality Safety Analysis of the Effluent Microfiltration Bag Filters;"

3) POEF-520-94-54," Criticality Safety Analysis of the Microfilter Feed Pumps;"

4) NCS-CALC-97-004, " Summary of Calculations Report, pH Controller in the X-705 Building";

5) NCS-CALC-97-012, " Supplemental Calculations for Microfiltration;"

6) POEF-520-95-151, "X-705 Microfiltration Material Flow;"

7) NCS-CALC-98-005," Calculations for Determining an Acceptable Leak Rate from Safety System Valves in the Microfiltration System;"

8) POEF-340-98-113, " Dissolved Uranium Concentration Versus pH;"

9) POEF-SH-22, " Nuclear Criticality Safety Calculational Analysis for Fissile Mass Limits and Spacing Requirements for 55-Gallon Waste Drums;" and

10) NCS-CALC-96-027, " Calculations for Small Diameter Container Carts."

A review of these type of calculations was completed as a part of Tasks 2 and 26 of the NCS CAP. As part of Task 2 of the NCS CAP, the plant management committed to review NCSE/As completed by engineers that were not properly qualified, in accordance with plant procedures, to develop the NCSE/As. Subsequent to the development of Task 2, the plant management determined that the same unqualified engineers may have developed or reviewed calculations which supported the NCSEs. The plant management initially planned to address this second issue as a part of Task 2; however, as a result of documentation problems identified during the closure of Task 2, the plant management established a new task (Task 26) to resolve this issue.

As part of Task 26, the plant staff determined that at least four of the supporting calculations for NCSE 0705-015 had not received an appropriate review and verification.

Specifically, the plant staff determined that NCS-CALC-97-004 and NCS-CALC-97-012 were being used without an appropriate peer review. However, the inspectors noted that the plant staff did not perform a re-verification of these supporting calculations until more than a year after the original issue was raised and almost seven months after NCSE 0705-015 was authorized for use in the field.

The inspectors also determined that evidence for one of the referenced calculations (POEF-340-98-113), could not be located and the plant staff did not identify the calculation as one requiring a review to determine if the involved engineer was qualified to perform the calculations. Further, the inspectors noted that the plant staff had not reviewed the document to determine whether the conclusions adequately supported the NCSE 0705-015-assumed chemical behavior of uranium concentration for various pH operational ranges. However, the conclusions from POEF-340-98-113 were used to demonstrate that the passing of undetected, dissolved uranium to the geometrically-unfavorable T-105 Effluent Tank was highly unlikely. The inspectors also noted that Task 26 did not list POEF-520-94-54 as an evaluation requiring review even though NCSE 0705-015 referenced the calculation to support the subcriticality of waste water treatment components under fire sprinkler discharge.

13

O Technical Safety Requirement 3.11.1 requires, in part, that the NCS program shall be established, implemented, and maintained as described in the SAR and shall address process evaluations and approvals. Safety Analysis Report Section 5.2.2.3," Process Evaluation and Approval," required, in part, that a documented NCSE shall include a 1 determination of the credible process upset conditions, and an identification of the '

assumptions and equipment needed to ensure NSC. The failure, in the NCSE, to perform or document reviews, for adequacy and applicability, of referenced and relied i

upon calculations and assumptions related to the fissile material operations described in NCSE 0705-015 is a Violation (070-7002/99004-01b).

The inspectors also identified that NCSE 0705-015 credited "unlikely events" as controls, supporting the double contingency logic, without an appropriate justification for the unlikeliness basis and without identifying the implicit control which made the events unlikely. For instance, the controls provided to prevent uranium solids from passing into an unfavorable geometry tank, T 105, were an unlikely event (Control A - microfilter  !

membrane failure) and the microfiltration permeate effluent bag filter system (Control B).

However, the evaluation did not provide a basis for the assumed membrane availability and reliability (e.g., an indicator down stream that would alarm when the membrane I failed). The inspectors were also informed during discussions with the Building X-705 management that a membrane failure had occurred as recently as 1995. In addition, the inspectors identified a May 1994 PR (SSR-94-02-06) which documented a failure of the ceramic membrane module. The PR description indicated that the ceramic module l failure (Control A) resulted in a pressure rise which activated the bag filter safety system

[ Control B]. Within this same NCSE, the inspectors found that other unlikely events were identified as controls for double contingency (such as the unlikely event that a solution would have a pH low enough to allow dissolved uranium to move past the microfilters) without appropriate justification.

Based upon the results identified as a part of the review of NCSE 0705-015, the inspectors performed a cursory review of several other NCSEs to determine if unlikely events were also used as a part of the double contingency logic. The inspectors determined that many of the NCSEs used unlikely events as a control for at least l one leg of the double contingency logic. However, the NCSEs did not appear to include i an appropriate justification for the unlikeliness basis and did not identify the implicit controls which made the events unlikely. In several of the cases, the inspectors noted that other, more robust, controls appeared to have been available.

During discussions with the NCS Manager, the Nuclear Safety Manager, and other NCS staff, the inspectors were informed the management and staff believed that guidance of ANSI /ANS 8.1, " Nuclear Criticality Safety in Operations with Fissionable Materials Outside Reactors," allowed the double contingency to be demonstrated by a combination of controls and unlikely events. In addition, the management and staff indicated a belief that the NRC recognized this interpretation as acceptable. As evidence of their position, the plant staff provided the inspectors with an NRC Inspection Report 070-7002/97203.

The inspectors reviewed the inspection report and noted that the inspectors had questioned an apparent inconsistency between two sections of the SAR, wherein only one section allowed the use of unlikely events. The inspectors noted that the SAR discussion on the use of "unlikely events" was included in a section describing areas covered by the Compliance Plan. The Compliance Plan described non-compliances with the NRC regulations that existed prior to and at the time of the NRC's assumption of 14 J

regulatory authority for which the plant staff were required to take specific actions to correct. During the time that the plant was in non-compliance, a Justification for Continued Operation (JCO) provided an interim safety basis. Since the plant was committed, by the Compliance Plan, to revise the NCS controls using passive and active engineered barriers, the inspectors determined that the conclusions documented in NRC Inspection Report 070-7002/97203 were consistent with the current findings. The inspectors also noted that once the non-compliance was considered corrected, the involved section of the SAR was deleted.

The inspectors noted that the current SAR required the NCS program to meet the double contingency principle by implementing at least one control on each of two different parameters or implementing at least two controls on one parameter. The SAR further defined controls to include passive barriers, active engineered features, and administrative controls, without explicitly discussing the use of unlikely events as controls.

The plant staff noted that although the double contingency matrix (Part C) for i NCSE 0705-015 identified an unlikely event failure as a control for double contingency, the evaluation section (Part B) showed that, at least for the membrane failure scenario, double contingency could be demonstrated without the use of the unlikely event as a control. The plant staff also indicated that although not specifically listed in the NCSA Part B as an administrative control, pH control was an integral part of the operation. The plant staff also indicated a belief that double contingency was still maintained without reliance on the unlikely event due to the redundant pH safety system. However, the inspectors noted that the pH safety system operated as one control using two out of l three logic to automatically close two isolation valves on a low pH signal.

Following these discussions, the plant staff reviewed all the NCSEs to determine if the double contingency logic relied upon the use of unlikely events. No NCSEs were identified which relied upon unlikely events for both legs of the double contingency logic.

However, the NCS staff did identify 44 NCSE/As that relied upon an unlikely event as one leg of the double contingency logic; As a result, the NCS staff initiated a more comprehensive review of the 44 NCSE/As to determine if: 1) the NSCE included sufficient appropriate justification to demonstrate the availability and reliability of the implicit control function for the identified unlikely event; or 2) a more applicable control could be identified.

Technical Safety Requirement 3.11.1 requires, in part, that the NCS program shall be established, implemented, and maintained as described in the SAR and shall address process evaluations and approvals. Safety Analysis Report Section 5.2.2.3, " Process Evaluation and Approval," required, in part, that a documented NCSE shall include a l "Jetermination of the credible process upset conditions, and an identification of the assumptions and equipment needed to ensure NCS. The failure to perform evaluations for or to document in the NCSE the controls and justifications, in terms of the availability 1 and reliability, necessary to support the use of "unlikely events" as a part of the double '

contingency principle is a Violation (070-7002/99004-01c).

The inspectors noted that several of the NCSE 0705-015 controls appeared to have been used in lieu of analyses which potentially could have eliminated the extra controls. For instance, the NCSE included controls to preclude an over-batching situation rather than analyzing for the situation in the filter press. In addition, the revised NCSE included 15

I 29 administrative controls versus the 16 included in the original NCSE. The inspectors questioned whether the increase in administrative controls added to the reliability of microfiltration operation or whether the control schemes could have been streamlined with appropriate analysis. A positive feature of the revised r.nalysis was that the passive and active engineered controls were specif%lly identitied and controlled under the appropriate maintenance programs.

l The inspectors determined that NCFa Plant 079 also included undocumented assumptions or conflicting double contingency controls. For instance, NCSE Plant 079, Section A.7, stated that interaction was not controlled even though Section A.10 required a ten foot separation between uranium deposits or other fissile material containers. The l double contingency arguments for moderation and interaction control did not appear to

! consider the need to cover all containers (not just the enuipment containing the deposit)

! with pre-staged moderation covers in the event of fire spnsler discharge. Section B.10 contended that removing fissile material from the equipment deposit and placing the fissile material in properly spaced containers reduced the criticality potential because the material would be spread out in the containers. However, since more than a critical mass

, (maybe sevxal critical masses) could potentially ha contained in the equipment, the possibility existed that the amount of material removed and placed in any one container l

could contain sufficient mass to result in a critical excursion, if appropriately moderated.

Therefore, moderation control appeared to be required for the removal containers as well as the equipment. The failure to loentify and document all of the assumptions related to double corningency control for the fissile material operations described in NCSE Plant 079 is another example of Violation 070 7002/99004-01c.

The inspectors reviewed the NCSA controls implemented for the microfiltration system as a part of NCSE 0705-015. The microfiltration system included five-inch diameter (safe geometry) vertical storage columns from which a liquid batch could be transferred to an unsafe geometry filter press for the removal of the entrained uranium solids. Prior to transferring the solution to the filter press, the NCSA required a determination of the liquid batch uranium-235 concentration. The concentration was analyzed through the collection of two independent liquid samples from each individual watte water batch by two different plant operators. After the samples were collected, the plant procedures allowed sample preparation by a single laboratory technician. The samples were then  ;

analyzed for uranium content by the following two wet chemistry methMs: 1) dye me'. hod in conjunction with a spectrophotometer instrument for comparatively low uranium levels, er 2) the Daves-Gray titration method for comparatively high uranium levels. In preparing the sample for the titration method, the inspectors were informed that it was important to add sufficient chemicals to the sample such that all of the uranium would react, so as to allow the procedure to not result in an erroneously low uranium concentration. The results of the wet chemistry analyses were then provided by the laboratory technician to two separate plant operators who independently calculated the uranium quantity contained in the liquid batch stored in the safe geometry columns.

l_

The inspectors questioned the independence of the sample analysis process, in that, a

! single laboratory technician prepared the two samples to be analyzed and a single l spectrophotometer instrument was used to measure both samples. The inspectors also raised the issue to the NCS staff for review and disposition concerning other similar NCS-significant sample ans'yses conducted in other facilities on site using mass spectrometers. The inspectors were informed by the NCS staff that a very rigorous

} quality assurance (QA) program was applied to laborLiory pocedures dealing with 16 3

4 uranium sample analysis at the plant such that a criticality event caused by an error in tne e.aytical process was not credible. However, the inspectors could not identify any NO'S which covered the operations or bases supporting the NCS staff position.

Technical Safety Requirement 3.11.2 requires, in part, that all operations involving uranium enriched to 1.0 weight percent or higher uranium-235 and 15 p, rams or more of uranium-235 shall be based upon a documented NCSE and shall be porformed in accordance with a documented NCSA. The failure to develop and base the conduct j laboratory analyses of NCS-significant samples, that relied upon a single laboratory 1 technician to perform the wet chemistry process, with and without the use o' the spectrophotometer instrument and with the mass spectrometer instrument, on a documented NCSE is a Violation (070-7002/99004-02a).

The inspectors discussed the details of the QA process and the specifics of the laboratory operations with the Buildings X-705 and X-710 laboratory staff. Control samples were typically run with each sample analysis when using the wet chemistry process. However, for mass spectrometers, a control was run at the beginning of each shift or after ten measurements. For the spectrophotometer, a control was run at the beginning of each shift. The Buildings X-705 and X-710 laboratory staff informed the

, inspectors that typically, less than ten measurements were made in one shift by a single spectrophotometer. ,

The laboratory staff provided the inspectors with control charts for a Building X-705 spectrophotometer and two Building X-710 mass spectrometers. For one of the mass spectrometers, the inspectors noted that five measured values, out of 764 control measurements made over the last three years, were reported between two and three standard deviations or less than 0.1 percent of the standard value. Two measured values were reported to differ from the standard value by slightly more than three standard deviations. For the Building X-705 spectrophotometer, the inspectors noted that seven measured values, out of 3,412 control measurements since 1990, were reported as differing from the standard value by more than three standard deviations.

Three standard deviations for the spectrophotometer amounts to less than 10 percent of the standard value.

The inspectors noted that for the Building X-705 microfiltration sampling process, an NCS significant result would only be associated with the control measurement that was significantly below negative three standard deviations since the NCS staff was controlling the filter press mass to 350 grams U-235 which was less than half of the minimum critical mass for fully enriched uranium.

The laboratory staff reviewed all control charts generated since 1997 for the mass spectrometers and for the Building X-705 spectrophotometers. The laboratory staff reported that for the mass spectrometers,46 of 11,963 measurements were reported as being out of control, i.e., either two consecutive measurements between two and three standard deviations or one measurement outside three standard deviations.

The acceptable accuracy for the mass spectrometers was reported as being above 99.5 percent for this time period.

The inspectors asked the laboratory staff if the rare out-of-control measurements for systems used to analyze NCS significant samples were reported to the NCS Organization. The inspectors were informed that the NCSAs contained the requirements i 17 l

L.

f 1

l I

dealing with sampling procedures that were relied upon to satisfy the double contingency principle from the standpoint of NCS, such as the sampling procedure discussed above for the microfiltration filter press operation, and that all out-of-control measurements were reported only to the Material Control and Accounting Organization and not the NCS l Organization.

The inspectors noted that the plant laboratory QA program failed to adequately address NCS for mass spectrometer and spectrophotometer sample analyses in that rare out-of-control measurements on the low side (control measurements that were less than three standard deviations of the actual standard value) may not be detected in a timely manner, thus potentially causing a loss of double contingency. Also, the inspectors did not identify any NCSA or approval that was developed for analysis of NCS significant samples which included controls for the handling of out-of-control low measurements.

The inspectors noted that in addition to the Material Control and Accounting Organization, it would be appropriate to include specific procedural requirements for NCS review of such rare out-of-control measurements, i.e., control measurements that are 1 outside negative three standard deviations, for instruments used to measure samples I required by Part B of the NCSAs, before any decision was made to release uranium bearing material to uncontrolled situations.

Technical Safety Requirement 3.11.2 requires, in part, that all operations involving uranium enriched to 1.0 weight percent or higher uranium-235 and 15 grams or more of uranium-235 shall be based upon a documented NCSE and shall be performed in accordance with a documented NCSA. The failure to develop a NCSE for the processing of NCS-significant samples using the mass spectrometer and spectrophotometer instruments, to ensure that any out-of-control measurements on the low side would be  ;

detected and evaluated in a timely manner to avoid a potential loss of double contingency is a Violation (070-7002/99004-02b)..

The inspectors reviewed the February 10,1999, event that occurred in the Building X-705. The event involved a plant operator who failed to correctly determine the uranium-235 concentration (Control A) and operator's supervisor failed to find the error by verifying the data and calculations (Control B) prior to transfer of a liquid batch into an unsafe geometry container. The Building X-705 management indicated that the error was caused by misreading the placement of the decimal point (in a non-conservative direction)in the sample analysis results sheet by the two plant operators relied upon for maintaining Controls A and B. The building management further indicated that the ,

erroneous values reflected typical uranium concentrations for the microfiltration storage l columns. The building management also described the CAs which included the ,

modification of the sample analysis sheet which would eliminate potential errors by the  !

two independent operators relied upon for maintaining Controls A and B.

The inspectors reviewed the recent changes made to Procedure XP3-EG-EG1037,

" Establishing & Controlling Quality Boundaries," Revision 1 and the plant staff's application of the procedure to NCSA-0705-015.A09, " Waste Water Treatment (Microfiltration System)." On June 2,1998, the NRC issued a Notice of Violation (VIO 70-7002/98206-02) for a failure by the plant staff to identify several structures, systems and components (SSCs), relied upon to meet the double contingency principle, as Augmented Quality-NCS (AQ-NCS). The violation involved a review of Part B (NCS controls) of upgraded NCSA 0326-013.AO, " Cascade Operations in the X-326 Building,"

and upgraded NCSA 0326-024.A04," Feeding of 5-inch,8-inch and 12-inch Cylinders in 18 a

the X-326 Product Withdrawal Area." The plant staff's response to the Notice indicated that a reason for the violation was that the procedure failed to provide adequate guidance for classifying SSCs required to meet the double contingency principle as AQ-NCS. On August 27,1998, the plant staff revised the procedure. - Some significant changes included in the revised procedure were requirements for the Ouality Boundary Evaluator to initiate a PR if an SSC needed to be re-classified as AQ-NCS from NS, and for the NCS staff to identify in the NCSE/A, any assumptions and equipment used as the basis for double contingency. A new section, ? Establishment of Quality Boundary Criteria," was also added to the procedure. The new section described a methodology for identifying active systems, passive barriers, alarms, instrumentation, and support systems as AQ-NCS SSCs. A 10 CFR 21 applicability determination for newly identified AQ-NCS SSCs was also addressed in the revised procedure. The inspectors also noted that a flowchart, included in the procedure for use in determining AQ-NCS SSCs, was simplified and that the flowchart would likely result in the identification of several new AQ-NCS SSCs.

The inspectors performed a check of the plant staff's application of the revised Procedure XP3-EG-EG1037 for upgraded NCSE/A 0705-015.A09, " Waste Water Treatment"(Microfiltration System). The inspecto,s noted that the NCSA identified the filter press blank plate as a physical control relied upon for double contingency.

However, the plant staff had not identified the filter press blank plate as an AQ-NCS SSC and the filter press blank plate was not indicated as an AQ-NCS component in the Boundary Definition Manual. The inspectors identified the filter press blank plate as an AQ-NCS SSC when following the process outlined by the procedure. The inspectors

- discussed the finding with the NCS staff and were informed that the filter press blank plate would be re-classified as an AQ-NCS SSC.

Technical Safety Requirement 3.11.1 requires, in part, that the NCS program shall be '

established, implemented, and maintained as described in the SAR and shall address identification of safety system components and support systems necessary to meet the double contingency principle. The Safety Analysis Report, Section 5.2.2.8 stated, in part, that functional and physical characteristics of operations controlled for NCS were described in NCSE/As. Components and features which were identified in the NCSE/As were analyzed to determine the " boundary" of the system, encompassing those items that were essential to ensure operability. Structures, systems, and components which require configuration control were identified as quality or AQ-NCS. The failure

' of the plant staff to identify in the Boundary Definition Manuals and to classify the filter press blank plate, a physical control relied upon for double contingency in NCSA-0705-015.A09, as an AQ-NCS component is a Violation (VIO 70-7002/99004-03).

Task 4: Enhance Nuclear Criticality Safety Aooroval Trainina The following subtask numbers were reviewed during the inspection:

4.1 Compile list of administrative controls document changes; 4.2 Identify target training audience by organization or job function; 4.3 Develop training module for each target group; and 4.4 Develop training effectivwess evaluation test.

The inspectors reviewed documentation for the task and subtasks and noted that the senior plant managers had satisfactorily conducted meetings to communicate 19

management's expectations regarding the proper implementation of NCS controls to all plant staff.

In addition, reviewed training records indicated that appropriate information was developed for each training module. Training modules communicated such topics as

" lessons leamed" from the NCSA walkdowns that were performed. The exact nature of this training varied from required reading of memos and e-mail messages to formal classroom training depending upon the nature and complexity of the subject matter.

Verification of training was conducted through appropriate test methods. The inspectors determined that the task was complete.

Iff 6: Comolete Comorehensive Root Cause Analysis The following subtask numbers were reviewed during the inspection:

6.1 For each step of the NCSA process review a representative sample of the prs associated with that step and identify the specific root cause(s);

6.2 Verify that the root causes identified above were included in the summary level root causes which were identified by the root cause team. For those that were not, revise the root cause documentation, increase the sample size, and submit a lessons leamed or enhancement in accordance with the procedure described in Task 9;

6. 3 If the detailed review root causes were included in the summary level root causes, document the review; and 6.4 For each root cause associated with each step of the NCSA process, or associated Management Systems, develop programmatic changes to correct root cause as part of this CAP for input to Task 12.

In reviewing the evidence files that provided documentation for the task and subtasks, the inspectors noted that the senior plant managers had satisfactorily. reviewed a representative sample of the approximately 1,000 NCS-related prs written in 1997.

Taproot analyses were satisfactorily conducted. The Taproot analyses generally determined that lack of accountability and inadequate administrative controls were the root causes of the issues identified in the associated prs (see Section 4). The inspectors determined that appropriate CAPS were developed to address the associated I root causes. The inspectors determined that Task 6 was adequately completed. ]

1 TASK 8: Vertical Slice Review l In May 1998, the senior plant management had an independent review of the NCS ,

program performed. The vertical slice review was conducted by a team of independent  ;

NCS industry personnel and included a review of a select number of NCSE/As and the i associated supporting activities and documentation to determine the current status of the i NCS program and its implementation.  ;

The report documenting the review findings (Vertical Slice Report) identified at least 200 recommendations and observations of flawed NCSE/As and circular double

' contingency logic. In general, the report found that each of the selected NCSE/As reviewed contained the following deficiencies: 1) one or more unjustified 20

assumptions; 2) inadequate accident sequences; 3) faulty double contingency arguments; 4) inadequate process descriptions; and 5) inadequate basis for controls utilized to fulfill the double contingency principle. The report also concluded that a more rigorous peer review would have eliminated a substantial number of these problems. '

l The inspectors found that many of the repo t recommendations encompassed the current inspection findings. For instance, the report found that the basis for judging an event as unlikely was not always documented in the NCSEs and that the summary of controls resulting from the NCSEs were not always adequate for analyzing conditions upon loss of a control or occurrence of a contingency. The report also recommended that the contingency analyses should discuss the mitigating measures sustaining subcriticality upon loss of a control.

The inspectors discussed the vertical slice report findings with senior plant managers and the NCS staff to determine what actions were being taken to address the findings. )

i According to the NCS staff, ten prs were written to address the vertical slice report findings and were placed into the BPS database in approximately October 1998.

According to an intemal memorandum (POEF-832-99-013), dated February 8,1999, additional staff reviews identified approximately 190 remaining findings that were .

imbedded in the report and these were entered into a database in January 1999 in order l to group similar topics, track resolution, and supplement the closure documentation. i Intemal memorandum (POEF-832-99-013) also indicated that CAs for the original ten issues was initiated in May and July 1998, but because a final version of the report was not available, most of these efforts were not tracked to completion.

Discussions with the NCS Manager indicated that the manager was not aware that any of the NCS staff had reviewed the Vertical Slice Report or were specifically directed to do so. Discussions with the Nuclear Safety Manager also revealed that an incorporation of the findings into planned and upgraded NCSE/As had not occurred and that no formal

! action was currently planned. The inspectors noted that the senior plant management, in I correspondence numbered GDP-97-0216," Transmittal of the Revised Corrective Action Plan for the Portsmouth Nuclear Criticality Safety Program," dated December 22,1997, committed to prioritizing NCSE/A upgrades as programmatic improvements were identified through the vertical slice review, and providing a feedback mechanism to adjust the interim NCSA process as necessary (Task 9). In addition, lessons leamed from implementation of tne interim NCSA review process was to be fed back to the programmatic review and improvement effort as appropriate.

Review of the authorization page of the Vertical Slice Report revealed that the document has dates of May 29 and July 15,1998, and a later signed version date of January 1999. Cursory information from the NCS staff revealed that quality issues were raised conceming the original report (May 29,1998) in that the original report had numerous revisions to correct technical inaccuracies and as a result, more than one version of the report existed. Plant Procedure XP2-BM-C11031," Corrective Action Process," Revision 0, dated June 15,1998, Section 5.0 required, in part, the development and implementation of CAPS and responses to prs in an effective and .

I timely manner. As of the end of the inspection period, the inspectors had not determined if the plant staff's actions to document in problem reports, evaluate, and develop CAs -l '

actions for the issues identified in the independent assessment of the NCS Program (Vertical Slice Report) in an effective and timely manner; therefore, the plant staff's handling of the issues raised in the report will be tracked as an Unresolved item (URI 70-7002/99004 04).

21

)

Task 11: Outside and Independent Assessments The assessments completed for Task 11 were conducted by the certificatee prior to the Task 8 review. The following subtask numbers were reviewed during the inspection:

11.1 Identify reports that fall within the scope of this task;  !

11.2 Review each report and document potential problem areas; 11.3 Compare the list of problem areas with the list of known deficiencies identified through other means (e.g., comprehensive root cause, ANSI reviews, etc.); and 11.4 For those problem areas that were not documented under other tasks, identify potential programmatic improvements using the process established for feedback of lessons leamed and enhancements.

in reviewing the documentation for the task and subtasks, the inspectors noted that i appropriate NCS staff had satisfactorily conducted a review of previous NRC inspection reports and reports submitted by independent groups. The documentation indicated the NCS staff reviewed the following reports: 1) NRC inspection reports (1997 and 1998);

2) contractor reports (reports of April 1997 and June 1997 and report of April 1998);

3) Plant Performance Review Committee report of September 1997; 4) Independent group audit of March 1997 and intemal plant self assessment of June 1993; and

5) Safety, Safeguards and Quality audit of April 1996. The inspectors determined that plant staff appropriately listed the potential findings identified in these reports and developed and processed the " lessons leamed"into applicable CAPS. Each finding was satisfactorily resolved by the plant staff and associated CAs were implemented by the NCS staff. However, the inspectors noted that two findings dealing with administrative procedures and the locations of Criticality Accident Alarm System monitors in Building XT-847 and the Building X-705 South Annex had still not received United States Enrichment Corporation (USEC) Headquarters (HQ) approval. Therefore, the inspectors considered Task 11 as open until these two items were resolved by USEC-HQ.

Task 13: Revise Trainina Proaram for Site Personnel  !

The following subtask numbers were reviewed during the inspection:

13.1 Evaluate results of prior training; I 13.2 Develop a list of programmatic problems; I 13.3 Perform a root cause evaluation on each identified problem; i 13.4 Identify actions to address root causes; I 13.5 incorporated improvements into the training administrative control process; i 13.6 Upgrade training modules to reflect improvements; 13.7 Determine retraining required to reflect upgrades; and ,

13.8 Administer required retraining using the revised program. '

Based upon a documentation review, the inspectors noted that training was reviewed by i NCS staff and subsequent root cause evaluations were conducted as appropriate to .

identify programmatic issues in the overall training for NCSE/As (see Section 4). The .

l CAP identified improvement steps to ensure that the training process was properly developed and implemented. Subtasks 13.6 through 13.8 were ongoing evolutions as

' the subtasks pertained to the development of appropriate training modules and administration of required training to communicate the revisions of NCSE/As and associated procedures. The NCS staff resolved to track the incomplete Subtasks 13.6 22

through 13.8 under Task 3, NCSE/A Upgrade Project. The inspectors determined that Task 13, up to Subtask 13.5, was complete.

Task 15: Confiauration Manaaement Proaram Enhancements The following subtask numbers were reviewed during the inspection:

15.1 Develop program changes to ensure that engineering documents were maintained consistent with NCSAs; and

-15.2 Implement the new program requirements.

Upon documentation review, the inspectors noted the plant staff developed and implemented Procedure XP2-EG-CF1031," Implementation Action Work Sheets," on February 28,1998. The procedure addressed configuration control for NCSE/As. The work sheets were designed to ensure that NCSE/A revisions were appropriately incorporated into applicable plant procedures and other engineering documents and that adequate staff training was provided for completing the work sheets correctly. However, the inspectors noted that the worksheets did not include a supervisory signature block as an independent check on the NCS revisions that were to be incorporated into the applicable plant procedures. The inspectors considered Task 15 as complete.

Task 19: Continued Use of Nuclear Criticality Safety Evaluation and Acoroval Review Board Review of the documentation indicated that the NCSE/A Review Board was discontinued.

The inspectors noted that the NSC staff resolved to enhance the training given to NCS subcommittee members, thereby eliminating the need for the NCS Review Board. The inspectors determined that the enhanced training was sufficient to ensure that appropriate and effective reviews of NCSE/As were conducted by the NCS subcommittee members. The inspectors determined that the task was complete.

Tasks 20 and D.8: Nuclear Criticality Safety Field Operational Assistants and Nuclear Criticality Safety Oversiaht on the Ooeratina Floors The inspectors reviewed documentation which indicated that the Field Operational Assistants (FOA) had been assigned to provide on-shift and field support to operations staff regarding NCS concems. The task closure documentation further indicated that enhanced NCS training (see Task 4) was conducted consistent with the goals of Tasks 20 and D.8. As a result, the plant staff determined that the FOA position was no

' longer needed. The inspectors determined that the tasks were complete.

c. Conclusions The inspectors identified violations involving incomplete or inappropriately completed  !

activities associated with the implementation Nuclear Criticality Safety Corrective Action Plan, Task 3. The violations indicated inadequate rigor in the performance of some CAs ]

and inadequate oversight or review of other corrective actions. The inspectors also identified indications that prompt actions may not have been taken for issues developed as a part of an independent review of implementation of the NCS Program. The inspectors determined the corrective actions required under Tasks 4,6,13,15,19,20, and D.8 were completed.

23

l N1.2 Chances to the Nuclear Criticality Safety Corrective Action Plan

a. Inspection Scope The inspectors resiewed a reduction-b-scope of CAs developed to resolve and prevent the recurrence of problems associated with the NCS Program. The problems were identified, collectively, as a significant condition adverse to quality. The reduction-in-scope involved a decrease in the number of reviews and upgrades to NCSE/As categorized as priority 2 and 3 activities.

b. Observations and Findinas in November 1998, the plant staff initiated a reduction in the number and scope of reviews and upgrades to existing NCSE/As compared to the number originally planned as a part of the CAs to PR. The redbetion deleted 125 of 147 of the Priority 2 and 3 listed NCSE/As. Originally, the CAP for Problem Report PR-PTS-97-08987 identified that the plant staff would review and upgrade all of the existing NCSE/As. The purpose of the reviews and upgrades was to correct the results of known deficiencies in the plant staff's implementation of the NCS Program. The known deficiencies were Delieved to have resulted, generically, in the development of NCSE/As that were confusing, incomplete, may contain technical errors, and which did not adequately identify all of the controls relied upon for NCS. In addition, the CAs were intended to assure plant management that the documents were developed in accordance with the existing plant policies for the NCS Program.

The inspectors discussed with engineering management the process used to identify the changes made to the CA plan and the management and staff reviews that were conducted to validate the appropriateness of the changes. The inspectors were informed that the changes were made following discussions between the engineering and operations staff which were focused on identifying those NCSE/As that the operations staff were not having a problem implementing. Based upon these discussions the engineering management implemented the reduction-in-scope change to the PR CA plan as tracked by the plant staff under the numbers PR-PTS-97-08987 and NCSA-PRI-01.

This change to the CAP was also communicated to the NRC in a quarterly status letter dated November 13,1998. The inspectors were informed that the plant staff did not document the discussions or logic developed to support the CA plan change. In addition, the engineering management and staff indicated that the appropriateness of the changes were not assessed against the original CAs nor were the changes reviewed and approved by the CARB.

During discussions with the NCS staff, the inspectors were informed that the plant staff had begun an effort to re-review the changes previously made to the scope of the CAs for the NCS program. The re-review was to be accomplished using a set of safety criteria. The inspectors noted that the process was more formal that the review completed in the fall of 1998; however, the process still appeared to be inconsistent with the plant procedural requirements for changes to CA plans associated with issues categorized as "significant conditions adverse to quality." Specifically, the proposed process did not assess whether the changes were warranted and did not determine if the revised CAs would still address the root causes identified for the original problems.

10 CFR 76.93, Quality Assurance, requires, in part,' that the plant staff shall establish, maintain, and execute a quality assurance program satisfying each of the applicable 24

requirements of American Society of Mechanical Engineers (ASME) NOA-1-1989,

" Quality Assurance Program Requirements for Nuclear Facilities." Section 2.16 of the Quality Assurance Program, " Corrective Action," required, in part, that conditions adverse to quality shall be identified promptly and corrected as soon as practical. In the case of significant conditions adverse to quality, the cause of the condition shall be determined and CAs taken to preclude recurrence. Procedure XP2-BM-Cl1031," Corrective Action Process," Revision 0, Change A, dated June 15,1998, Section 6.0 required, in part, that plant staff shall verify that changes to CAs for significant conditions adverse to quality:

1) were warranted; 2) would still fix the root or contributing causes, as originally determined; and 3) were reviewed and approved by the CARB. The failure to develop, assess, and implement a change to the CA plan for a significant condition adverse to quality, which involved a decreased in the scope and depth of reviews for some Priority 2 and 3 NCSE/As, is a Violation (VIO 70-7002/99004-05).

c. Conclusions The inspectors identified a violation in that the plant staff failed to ensure that changes to a CA plan, originally developed to resolve and prevent the recurrence of a significant condition adverse to quality, were properly developed, assessed, and implemented in accordance with the QA Program and plant puedures. As a result, the plant staff deleted scheduled reviews of 125 potentially incomplete or inaccurate NCSE/As.

N1.3 NCSA W'alkdowns

a. Insoection Scope The inspectors assessed the implementation of selected NCSAs through discussions with operations staff, field walkdowns, and a review of associated procedures. The NCSAs reviewed included:

• NCSA-0326-024.A04," Feeding of 5-inch,8-Inch and 12-inch HEU Cylinders at the X-326 Product Withdrawal and Product Purification," dated March 19,1998; e NCSA-0326-013.A06, " Cascade Operations in the X-326 Building," dated i May 18,1998; and e NCSA-0330-004.A03, " Cascade Operations in the X-330 Building," dated May 18,1998.

b. Observations and Findinas Based upon discussions with the various building operators and FLMs, the inspectors determined that the operations staff were knowledgeable of the NCSAs in effect for their areas and the associated administrative controls that were documented in the operations procedures. The inspectors verified with the FLMs that the systems and operation descriptions, included in the NCSAs, were accurate. The inspectors also verified that applicable NCS field postings were consistent with the NCSA requirements.

25

1

c. Conclusions The inspectors determined the operations staff were knowledgeable of the NCSAs applicable to their areas and that the associated administrative controls were adequately incorporated into applicable building procedures. In addition, applicable NCS field postings were found to be in accordance with the requirements specified in the applicable NCSAs.

V. Manaaement Meetinas X Exit Meeting Summary l l

The inspectors met with facility management periodically during the inspection. The inspectors presented the inspection scope and findings to members of the plant staff at the conclusion of the inspection on March 12, and during a re-exit meeting conducted on April 28,1999. No classified or proprietary information was discussed. At the exit and re-exit meetings, Portsmouth management and staff acknowledged findings.

26

e PARTIAL LIST OF PERSONS CONTACTED Lockheed Martin Utility Services

• J. Anzelmo, Commitment Manager.

• • T. Brooks, Nuclear. Safety Manager -

• S. Casto, Work Control Manager

• D. Couser, Training Manager

• • L.Cutlip, Shift Supervisor

• R. Dively, Commitment Management Staff

• D. Faust, Corrective Action Manager

• C. Forshey, Commitment Management Staff.

• S. Fout, Operations Manager

• • R. Helme, Acting Engineering Manager L J. Hobensack, Process Services Manager

• • M. Hone, Nuclear Safety Staff

S. Martin, Nuclear Safety Staff

• • P. Musser, Enrichment Plant Manager

• J. Oppy, Operations Staff

• K. Sherwood, Nuclear Regulatory Affairs Staff

• • J. Shewbrooks, independent Assessment Group Manager

-"R. Smith, X-705 Manager-

• R. Smith, Production Manager

• E. Wagner, Nuclear Safety Staff

• M. Wayland, Maintenance Manager J. Woodard, X-705 Building Manager G. Workman, Mass Spectrometry Manager United States Enrichment Corporation

• • M. Brown, General Manager "L. Fink, Safety, Safeguards & Quality Manager

• P. Miner, Regulatory Affairs Manager

• • T. Sensue, Nuclear Regulatory Affairs Engineer

• Denotes those present at the exit meeting on March 12,1999.

" Denotes those present at the re-exit meeting on April 28,1999.

27

6 INSPECTION PROCEDURES USED IP 40500: Effectiveness of Licensee Controls ITEMS OPENED, CLOSED, AND DISCUSSED Ooened 70-7002/99004-01 VIO Failure to determine all credible process upsets, to identify appropriate controls necessary to meet the double contingency principle, and to identify and validate assumptions and equipment needed to ensure criticality safety.

70-7002/99004-02 VIO Failure to develop an nuclear criticality safety evaluation to control the performance of laboratory analyses for samples taken in support of NCSE 0705-015.

70-7002/99004-03 VIO Failure to identify a physical control (filter press blank plate) as an item relied or, for criticality safety in the Boundary Definition Manuals and a failure to classify the filter press blank plate as an AQ-NCS item.

70-7002/99004-04 URI Handling of observations and recommendations included in the Vertical Slice Report.

70-7002/99004-05 VIO Failure to ensure that a change (reduction in the scope and number of NCSE/A reviews) to a corrective action plan for a significant condition adverse to quality was developed, assessed, and properly implement in accordance with the Quality Assurance i Plan and plant procedures. '

Closed None Discusted ,

None Certification issues - Closed None 28

LIST OF ACRONYMS USED AQ-NCS Augmented Quality Nuclear Criticality Safety BPS Business Prioritization System CA Corrective Action CAP Corrective Action Plan CARB Corrective Action Review Board CF Causal Factor CFR Code of Federal Regulations CM Commitment Manager CAQ Conditions Adverse to Quality E&CF Event & Causal Factor FLM First Line Manager FOA Field Operational Assistants HPE Human Performance Error IAG Independent Assessment Group JCO Justification For Continued Operation NCS Nuclear Criticality Safety NCSA -

Nuclear Criticality Safety Approval NCSE Nuclear Criticality Safety Evaluation NCSE/A Nuclear Criticality Safety Evaluation and Approval NRC Nuclear Regulatory Commission NOR Not Quality Related PDR Public Document Room PR Problem Report PRSA Problem Report Screening Administrative Committee PRSC Problem Report Screening Committee PSS Plant Shift Superintendent QOOP Quality Of Operations Plan OA Quality Assurance SAR Safety Analysis Report SCAQ Significant Condition Adverse to Quality SSC Structures Systems and Components

-SS&Q Safety Safeguards & Quality URI Unresolved item USEC United States Enrichment Corporation VIO Violation I

i 29 l

2