IR 05000244/1999005
| ML17265A733 | |
| Person / Time | |
|---|---|
| Site: | Ginna  |
| Issue date: | 08/06/1999 |
| From: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML17265A732 | List: |
| References | |
| 50-244-99-05, NUDOCS 9908130141 | |
| Download: ML17265A733 (28) | |
Text
U. S. NUCLEAR REGULATORYCOMMISSION
REGION I
Docket No.
50-244 License No.
DPR-18 Report No.
50-244/99-05 Licensee:
Rochester Gas and Electric Corporation (RG8E)
Facility:
Ginna Nuclear Power Plant Dates:
May 10-14, and May 24-28, 1999 Inspectors:
Suresh Chaudhary, Team Leader, DRS John McFadden, Radiation Specialist, DRS Christopher Welch, Reactor Engineer, DRS Robert Schin, Senior Reactor Engineer, DRS Greg Cranston, Reactor Engineer, DRS Approved By:
Lawrence T. Doerflein, Chief Engineering Programs Branch Division of Reactor Safety 9908f30i4i 990806 PDR ADQCK 05000244
EXECUTIVESUMMARY During the weeks of May 10 and May 24, 1999, a team of inspectors conducted an onsite inspection of the licensee's corrective action program implementation using the guidance of NRC Inspection Procedure 40500, "Effectiveness of Licensee Controls in Identifying, Resolving, and Preventing Problems."
OPERATIONS Through the corrective action program, RG&E identified problems at a low threshold, and appropriately prioritized the resulting action requests.
Root cause evaluations and development of corrective actions were performed adequately.
Management awareness of and involvement in the corrective action process was well evident.
Feedback mechanisms used to assess corrective action effectiveness were adequate.
Corrective actions were developed and implemented in a timely manner.
Although a formal problem trending process did not exist, efforts for improvement were noted. The corrective action program was determined to be satisfactory overall. (Section 07.1)
The Ginna Action Report (AR) program for problem identification and documentation and the Work Request/Trouble Report system were used satisfactorily to document plant deficiencies/issues.
The established threshold for issuing an AR was found to be appropriate, and it was apparent that management had effectively communicated their expectations to the staff concerning the use of the AR system.
(Section 07.2)
RG8 E's root cause determinations were generally satisfactory.
Increased emphasis on improving the human performance evaluation portion of the root cause determination was noted.
However, the effectiveness of this effort was not yet apparent, as plant events directly attributed to personnel error continued to occur.
In addition, weaknesses in licensee evaluation of an excessive overtime issue were observed.
The team also noted several examples of problems, not specifically related to human performance issues, which were not fullyanalyzed or evaluated during the root cause determination process to fullyassess all contributing factors.
(Section 07.3)
ENGINEERING In general, the operability determinations reviewed were acceptable.
Afew of the operability determinations reached an appropriate conclusion, but were not thoroughly documented.
One operability determination, regarding the main steam non-return check valves was inadequate.
(Section E7.1)
The assumptions, analytical methods, and calculations used by the licensee to declare the main steam non-return check valves operable may not be conservative and may not be applicable in all cases.
The licensee did not show that the uncertainty in the calculation is less than the available margin of torque needed to close the valve. Therefore, operability of the main steam non-return check valves remains an open issue pending NRC review of additional information from RG8E.
(Section E7.1)
The team identified several inadequate safety evaluations related to changes made to the main steam non-return check valves.
Specifically, the valves were changed from free swinging gravity closing valves (as stated in the Updated Final Safety Analysis Report) to valves that required a substantial and increasing external force to close them, without addressing potential effects on steam generator integrity, containment integrity, steam generator tube integrity, reactor reactivity, or reactor vessel integrity. Other procedure changes failed to include safety evaluations.
The team believes that changing the main steam non-return check valves to require a significant breakaway closing torque represents an Unreviewed Safety Question.
This is an apparent violation of 10 CFR 50.59.
(EEI 50-244/99-05-01).
The licensee did not agree that these changes introduced an Unreviewed Safety Question.
(Section E7.1)
The inspectors identified multiple examples (April 1999) that appeared to exceed the overtime limits of the Technical Specifications.
Specifically, a maintenance technician substantially exceeded site overtime restrictions without the required written authorization.,
While exceeding overtime restrictions without the required authorization, the technician was involved in causing a reactor trip. This item remains unresolved pending further review by the NRC.
(URI 50-244/99-05-02)
(Section E7.2)
., The team concluded that the peer-assisted self-assessments were good; they included a strong independent perspective and numerous findings for improvement.
Corrective actions and program enhancements resulted from the assessments.
(Section E7.4)
TABLEOF CONTENTS PAGE EXECUTIVE SUMMARY I. OPERATIONS..
Quality Assurance in Operations..
07.1 Corrective Action Program.
07.2 Problem Identification and Resolution..
07.3 Root Cause Evaluations 1.
.1
.1
5 II. Engineering
.
E7 Quality Assurance in Engineering Activities E7.1 Operability Determinations...
E7.2 Onsite and Offsite Review Committees.
E7.3 Operating Experience Feedback E7.4 Self-Assessments III. Management Meetings
.
X1 Exit Meeting Summary.
PARTIALLIST OF PERSONS CONTACTED INSPECTION PROCEDURES USED.
.6
6
12
. 13
14
15 ITEMS OPENED, CLOSED, AND DISCUSSED ATTACHMENT1 ATTACHMENT2
. 15
Re ort Details I. OPERATIONS
Quality Assurance in Operations 07.1 Corrective Action Pro ram Ins ection Sco e 40500 92901 The team reviewed the scope and effectiveness of the licensee's corrective action program. A representative sample of recent licensee-identified problems was reviewed to identify how corrective actions were prioritized, tracked, implemented, and subsequently verified for effectiveness.
Also, the level of management involvement, the feedback mechanisms for assessing corrective action effectiveness, the timeliness of corrective actions, and the effectiveness of performance trending were reviewed.
b.
Observations and Findin s A sample of recent licensee-identified problems was reviewed to evaluate how the corrective actions were prioritized, tracked, implemented, and subsequently verified for effectiveness.
Licensee procedures, IP-CAP-1 (Rev. 9, "Abnormal Condition Tracking Initiation or Notification (ACTION) Report"), IP-CAP-2 (Rev. 3, "Root Cause Analysis"), and IP-CAP-3 (Rev. 0, "Investigation Team" ) were the basic procedures used for problem identification, prioritization, tracking, root cause evaluation, and implementation of corrective actions.
Corrective actions were tracked and implemented by supplemental procedures for work orders, technical service requests, engineering work requests, plant change requests, procedure change requests, commitment and action tracking, training work requests, Updated Final Safety Analysis Report change notices, and drawing change notices.
Newly initiated Action Reports (ARs) were reviewed each day at the Morning Priorities Action Required (MOPAR) meeting.
The Plant Operations Review Committee (PORC)
chairman or alternate was present at these meetings and, based on the assessment of the AR, had the opportunity to require PORC review of an AR. PORC review could be immediate, upon AR disposition, and/or before closeout.
The prioritization of ARs was performed by a Screening Committee (SC) which consisted of a senior operational reviewer, an operating experience analyst, a human performance analyst, a trending analyst, and an engineering representative.
ARs were prioritized "A" (highest priority)
through "D" (for trending only). The SC determined a preliminary cause code and assigned each AR to the responsible department manager (RM) for disposition. The RM determined the level of root cause to be performed.
Event Investigation was also an option ifhuman performance was a potential facto For the ARs in the inspection sample, a low threshold for AR initiation was demonstrated.
Problem identification was specifically addressed in Section 07.2 of this report. The ARs were prioritized appropriately, and the evaluation for cause and development of corrective actions was performed adequately.
The licensee made several recent changes to the program. The recently formed Screening Committee relieved the Production Superintendent of the screening and administrative duties associated with ARs. Procedure IP-CAP-1 was revised to include a five-page guide for use in prioritizing ARs. This procedure also provided the option of a Precursor Report to facilitate anonymous submittals and a low threshold for AR initiation.
Although no performance discrepancies were identified, several concerns in this area were noted.
First, due to allowance by procedure for closure of an AR after transferring the implementation and ownership of AR-generated corrective actions to a satellite procedural process, there was no single point of accountability for the evaluation and implementation of all corrective actions.
Second, the requirement had been added to the satellite procedures to obtain the concurrence from the same individuals who concurred on the original AR, with respect to the corrective action, but did not require concurrence for a change in a due date.
Third, while the AR procedure required that the dispositioner identify any generic implications, the AR form did not show that generic implications were being considered.
The licensee acknowledged these concerns and indicated that they would consider the need for changes to the applicable procedures.
The second concern identified above was related to an issue in a previous NRC Inspection Report (IR 50-244/98-01).
In that report, an inspection follow-up item (IFI 50-244/98-01-01) was opened based on the fact that the corrective action program allowed ARs to be closed without all of the corrective actions being completed, by transferring corrective actions to a satellite process.
RG8 E indicated that they would conduct a special QA audit to review this concern and determine ifthe practice had allowed corrective actions to be missed.
This IFI was updated in NRC IR 50-244/98-03.
The licensee performed the QA audit in February 1998 by selecting forty-seven closed ARs to determine ifthe corrective actions were completed or in progress, as stated in each AR. The audit identified four ARs with follow-up deficiencies and generated additional ARs (98-305,98-306, 98-307, and 98-308) to correct the deficiencies.
As a result, the licensee initiated a second audit for additional sampling and to assess ifthere was a trend in the problems identified in the first audit. The QA auditors also recommended a revision to procedure IP-CAP-1 to require referencing the AR in the satellite processes.
Ifthe licensee made a decision to deviate from the original AR disposition during a follow-up satellite process action, a reassessment of the original AR would be required.
During this current inspection, it was verified that additional sampling ofARs (April 1998) had been performed and that the changes to the administrative procedures in the satellite documents had been made.
The findings in the second QA audit stated that the ARs showed the same level of deficiencies as the in the previous QA surveillance, but that the deficiencies were relatively minor. Based on the results of the audits and the actions taken to date, IFI 50-244/98-01-01 is close The level of management involvement in the problem identification, evaluation, and resolution, and the feedback mechanisms used to assess corrective action effectiveness were evaluated.
Numerous mechanisms for, and examples of, management involvement were evident.
Newly initiated ARs were reviewed each day at the MOPAR meeting.
The MOPAR meeting minutes, including descriptions of the AR items, were made available via e-mail to all plant personnel.
The agenda for PORC Meeting No.99-065, May 13, 1999, included the review of twelve ARs. During the review of these ARs, there was wide participation by attendees, with detailed questions and considerations concerning the ARs. Based on the above, the inspectors considered the level of management awareness of and involvement in the corrective action process to be appropriate.
The feedback mechanisms used to assess corrective action effectiveness included external assessments, internal audits and surveillances, Nuclear Assessment Quarterly Reports, and self-assessments.
These audits and surveillances reviewed provided a fully'scoped and in-depth examination of the corrective action process and effectiveness.
Based on this, the feedback mechanisms used to assess corrective action effectiveness were considered adequate.
For the selected ARs, the disposition/evaluation due dates and the due dates for completion of corrective actions were met. A number of the ARs had been closed by transferring. some corrective actions to satellite processes.
Although no performance discrepancies were identified, two concerns were noted.
First, in a review of the Performance Monitoring Monthly Reports, a large percentage of work backlog performance indicator (Pl) goals were exceeded for the last six months, including technical service requests, plant change requests, training work requests, and action reports.
Second, a single performance measure for timely completion of all corrective actions for ARs was not available due to closure of ARs with some corrective actions being turned over to satellite processes for implementation.
The licensee acknowledged these concerns.
The effectiveness of performance trending was inspected.
The lack of a formal trending process and trending information was identified as an area ofweakness in past internal audits. At the time of this inspection, this weakness continued with no formal trending process; and the only formal trending information was that in the previously mentioned Nuclear Assessment Quarterly Analysis Reports.
Plans for improvement were noted.
First, the issue of a weakness in trending ofARs was an action item assigned to and being evaluated by the Quality Assurance/Quality Control (QA/QC) Subcommittee of the Nuclear Safety Audit and Review Board (NSARB). Second, plans had been documented to develop a formal trending procedure by June 30, 1999. Third, trending documents reflecting current AR experience and preliminary causal coding were provided to plant management on a weekly basis during the recent refueling outage, and provision of these documents on a monthly basis during non-outage time was planned.
The inspectors determined the problem-trending process to be weak and informal, but efforts towards improvements were note c.
Conclusion Through the corrective action program, RG8 E identified problems at a low threshold, and appropriately prioritized the resulting action requests.
Root cause evaluations and development of corrective actions were performed adequately.
Management awareness
..of and involvement in the corrective action process was well evident.
Feedback mechanisms used to assess corrective action effectiveness were adequate.
Corrective actions were developed and implemented in a timely manner.
Although a formal problem trending process did not exist, efforts for improvement were noted. The corrective action program was determined to be satisfactory overall.
07.2 Problem Identification and Resolution a
Ins ection Sco e 40500 To evaluate the effectiveness of the licensee's processes for identifying, resolving, and preventing further problems that affect the safety of plant operations, the team reviewed the station's corrective action procedures and a listing (with brief descriptions) of all ARs written in the past 16 months.
The review included selected staff interviews, and plant tours to assess how well the problems/issues. were identified and documented.
Insight into the Employee Concerns program also was obtained through interviews.
Observation and Findin s The AR system continued to be a low-threshold high-volume reporting system with approximately 2500 ARs generated during the past 16 months. A new initiative, a
'precursor report'hat can be anonymously submitted, was implemented in March 1999 to further encourage reporting of near misses and low level events.
Ofthe 49 precursor reports submitted, the screening committee had elevated 5 of these issues to ARs (4 to Priority "C" and 1 to priority "B") in response to the significance of the issue.
The team questioned whether three additional precursor reports (99-0007, 0027, and 0044) should have been elevated to AR status, but with additional informatiori and discussions, the inspectors accepted the initial screening.
Based on interviews, the team determined that management had effectively communicated its expectations to the staff concerning the use of the AR system.
All stated that management expected issues/deficiencies to be identified and appropriately documented on an AR, as required.
Many indicated that their supervisor often asked if an AR had been issued when discussing problems.
The interviews indicated that when the staff was not sure whether to issue an AR, they would conservatively issue one, as stated in IP-CAP-1. However, questions regarding initiation of an AR in conjunction with a Work Request/Trouble Report (WR/TR) indicated that confusion existed in this area.
In spite of the uncertainty, those interviewed indicated an AR would be issued along with the work order in most cases with the exception of normal or planned evolution e During plant tours, the inspectors noted that areas were well maintained and that deficiencies were identified with maintenance identification (MID)tags.
No unidentified deficiencies were noted by the team.
Six MIDtags, selected during the tour, were checked for a corresponding AR. Two had an AR, while the others were justified as not requiring an AR by the licensee and complied with a literal interpretation of IP-CAP-1.
However, based on the interviews with RGLE personnel, the team found the 4 items somewhat contradictory to the stated philosophy and expectations of the management, indicating the AR threshold was not as low as perceived by station personnel.
The team found the Employee Concerns program satisfactory.
No issues had been submitted in the past year. The team was informed of ongoing efforts to enhance awareness of the program and future training plans to further enhance open communications between supervisors and their staff.
C.
Conclusion The Ginna Action Report (AR) program for problem identification and documentation and the Work Request/Trouble Report system were used satisfactorily to document plant deficiencies/issues.
The established threshold for issuing an AR was found to be appropriate, and it was apparent that management had effectively communicated their expectations to the staff concerning the use of the AR system.
07.3 a.
Root Cause Evaluations Ins ection Sco e 40500
The inspection focused on root cause determinations that were. primarily associated with human performance issues, and RG&E's efforts to improve the Human Performance Event (HPE) evaluations.
Applicable station procedures and a sample of HPE evaluations/investigations were reviewed, and interviews were conducted to assess the technical adequacy and effectiveness of the root cause evaluation and the program in general.
b.
Observation and Findin s RGB E initiated an effort to improve root cause determinations in the area of human performance.
This was noted in interviews and in the establishment of a new HPE evaluation process, separate from the existing root cause procedure IP-CAP.2. Based on the particular event, either or both methods would be used to identify all contributing causes.
Normally, IP-CAP-2 was used for equipment issues and the HPE evaluation (IP-HPE-1 being developed) for personnel performance issues.
The procedure ND-HPE, "Human Performance Enhancement Program" was issued in May 1999 and established the expectations and requirements for the Human Performance Enhancement Program.
The document was well written and contained insights which should aid in improving personnel performance at the statio Although the root cause/human performance evaluations reviewed by the team were generally satisfactory, the team observed some weakness in the license's evaluation of human performance.
Specifically during a Plant Operations Review Committee (PORC)
meeting in which the written event evaluation report for the April 23, 1999 reactor trip was being reviewed, the team noted that PORC members discussed and challenged some of the potential root causes, but notably did not pursue an excessive overtime issue.
The team further reviewed this overtime issue and identified weaknesses in the licensee's control of overtime (See Section E7.3).
Also, examples were noted where problems/processes that were not specifically or solely a human performance issue did not get adequate analysis and evaluation in the root cause determination process.
For example:
Event Investigation Report 98-0383 had identified lessons learned from a previous event involving an inadequate HOLD (tagout) on the same system.
These lessons were not effectively captured to prevent recurrence.
However, no investigation was apparent to determine ifthe station had an adequate process to accomplish this task or to understand why the existing process failed.
Event Investigation Report 98-0383 identified that responsibilities and expectations of the HOLD process had become diluted in multiple layers of review and that weakness in the HOLD process was rooted in a lack of specific guidance for the expected level of detail at each established barrier. However, no critical evaluation of the process/procedures to determine if, in fact, the controlling documents had become cumbersome and lacked clearly defined responsibilities and expectations was documented.
Event Investigation Report 98-1183 documented the failure to identify and manage the most restrictive condition as a primary root cause for an unanticipated rod motion, and that this was a result of combining three procedures into one evolution. However, no investigation was conducted of how the operators control or maintain cognizance of ongoing evolutions, which may have a direct impact on the plant.
Event Evaluation 99-0336 identified the root cause for the improper weir gate rigging as failure to use the required procedure.
However, no investigation regarding the adequacy of the work control processes to ensure procedural requirements was identifie c.
Conclusion RGB E's root'cause determinations were generally satisfactory.
Increased emphasis on improving the human performance evaluation portion of the root cause determination was noted.
However, the effectiveness of this effort was not yet apparent, as plant events directly attributed to personnel error continued to occur.
In addition, weaknesses in licensee evaluation of an excessive overtime issue were observed.
The team also noted several examples of problems, not specifically related to human performance issues, which were not fullyanalyzed or evaluated during the root cause determination process to fullyassess all contributing factors.
III. En ineerin E7
-
Quality Assurance in Engineering Activities E7.1 0 erabilit Determinations a.
Ins ection Sco e 40500 The inspectors reviewed the licensee's guidance for performing operability determinations and reviewed 22 operability determinations that had been performed in 1999. The team identified several deficiencies regarding one operability determination for the main steam non return check valves.
Observations and Findin s Main Steam Line Non-Return Check Valves
~Back round On March 1, 1999, the plant was in hot shutdown, cooling down for a scheduled refueling outage.
During the performance of surveillance test PT-2.10.15, "Main Steam Non-Return Check Valve Closure Verification," the licensee identified that the torque required to initiate valve movement (breakaway torque) was in excess of the 600 ft-Ibs of torque acceptance criteria. The operators appropriately entered Technical Specification 3.0.3, initiated an Engineering Technical Evaluation to evaluate operability, and continued with the plant coo!down.
Prior to reaching cold shutdown, the Engineering Technical Evaluation determined that the main steam non-return check valves were operable provided the breakaway torque was less than 900 ft-lbs. The licensee reported this condition to the NRC in Licensee Event Report (LER)99-003.
The team reviewed the operablity determination associated with this conditio Valve Maintenance Histo The packing on the main steam check valves was tightened in 1992 to address problems with packing leakage and check valve flutter. The tighter packing changed the valves from free swinging gravity closing to valves that required approximately 600 ft-Ibs of torque to close.
During the 1997 refueling outage, the licensee repacked the main steam non-return check valves and left them with the required closing torque of less than 600 ft-lbs. During the plant operating cycle, the torque required to close the valves increased from 600 ft-Ibs to a maximum of 900 ft-lbs. During the 1999 refueling outage, the main steam non-return check valves were again repacked and left with a required closing torque of less than 600 ft-lbs. On April 23, 1999, approximately one month after restart, the closing torque had increased to approximately 775 ft-lbs. The plant was restarted with the valves left in this condition.
Licensin Basis The team reviewed the Updated Final Safety Evaluation Report (UFSAR) description of the main steam non-return check valves.
The UFSAR stated in Section 10.3.2.7:
"the main steam non-return check valves... are free swinging gravity closing type check valves. The check valves protect the main steam header against reverse flowfrom one generator to another in the event of a steam line rupture." The UFSAR, Section 15.1.5.1 states that: "Each steam line has a fast-closing MSIVand a non-return check valve.
These four valves prevent blowdown of more than one steam generator for any break location even ifone valve fails to close.
For example, for a break upstream of the main steam isolation valve in one line, closure of either the non-return check valve in that line or the MSIVin the other line willprevent blowdown of the other steam generator."
Testin Procedures The team reviewed PCN 93-4130, which processed PT-2.10.15, Rev. 2, dated March 12, 1993. This procedure revision incorporated the acceptance criterion of less than or equal to 600 ft-Ibs This procedure revision included no safety evaluation.
The stated basis for exclusion from a full safety evaluation was: "This change incorporates the change in test methodology recommended by DA-ME-92-024. This new method will provide a much greater degree of assurance that the subject valves are operable and willbe capable of closing during all conditions of operation., This change does not place equipment in a configuration that is adverse to plant safety. This new test method is in full compliance with the code commitments of the Ginna Pump &Valve and,ln-Service Testing (IST) Programs."
The team noted that the basis for exclusion from a full safety evaluation failed to recognize that this change represented a change to the plant as described in the UFSAR, and that a full safety evaluation was required.
The team additionally reviewed the current revision of PT-2.10.15 (Rev. 6). It included an acceptance criterion of 600 ft-lbs plus or minus 300 ft-Ibs The team noted that it allowed a higher torque than Revs.
1 and 2, and thus further increased the probability that a main steam non-return check valve would fail to close during a main steam line break.
Rev. 6, which had increased the maximum acceptable breakaway torque from
600 ft-Ibs to 900 ft-lbs, had been processed and approved under PCN 99-4171, dated April23, 1999.
PCN 99-4171 included no safety evaluation.
The stated reason for not including a 50.59 safety evaluation was that the change had been previously reviewed as design analysis DA-ME-92-147, Rev. 1, dated April 15, 1999. The team noted that the design analysis also had no safety evaluation.
Technical Evaluation On March 1, 1999 an Engineering Technical Evaluation concluded that, with a measured closing torque of 900 ft-lbs, the main steam non-return check valves remained operable.
This conclusion was based on "Design Analysis DA-ME-92-147, Rev. 0, dated November 10, 1992. The Design Analysis determined that at least 1567 ft-Ibs torque was available to close the main steam non return check valves assuming
/~ the design basis accident steam flow. Therefore, the licensee concluded that there was a significant margin above the maximum measured breakaway torque of 900 ft-lbs. The team identified a mathematical error in the Design Analysis that reduced the calculated closing torque to 963 ft-lbs. The Engineering Manager initiated action to correct the error in the calculation.
Based on this Engineering Technical Evaluation, the licensee concluded that the main steam non-return check valves were operable and were not in a nonconforming condition.
The NRC conducted a detailed review of calculation DA-ME-92-147. The NRC review observed that the steam flow past the check valve, with flow reversal occurring at the time of the incident, presents a very complicated flowgeometry and that a detailed flow field and pressure distribution on the valve disc is needed to properly analyze the effects on the check valve. Additionally, the licensee did not show that the uncertainty in the calculation was less than the available margin oftorque needed to close the valve and it was not clear that the worst case condition was used regarding steam line break size and associated flowpast the check valve. The team noted that, although the 963 ft-Ibs available torque exceeded the measured 900 ft-Ibs torque, it may not provide significant margin during a design basis double-ended main steam line break.
Factors that influence the applicability and conservatism of the calculation for which additional information is needed include: basis for analysis method used (a 3-dimensional computational fluid dynamics code may be needed to properly model this complex flow condition); how the analysis method is validated; basis for assuming that the steam flow is non compressible since flow in the line is changing in mass flow rate and reversing direction; affects of a steam flowfrom steam generator that is blowing down until the check valve closes; and, basis for the check valve disc being treated as a fiat circular disc when there is flow on both sides ofthe disc and when there are flow obstructions on the top ofthe dis Further, during a smaller main steam line break concurrent with the failure of a MSIV, the main steam non-return check valve may not close at all and may allow blowdown of both steam generators.
This would represent an unanalyzed condition for steam generator integrity (both steam generators faulted), containment integrity (blowdown of both steam generators),
steam generator tube integrity (emergency operating procedures [EOPs]
would require the use of a faulted steam generator), reactor reactivity (potentially increased coo!down), and reactor vessel integrity (increased cooldown could overstress the reactor vessel).
The risk associated with this issue represents a minimal reduction in the margin of safety.
For this event to be of concern, the main steam/feedwater system(s) within containment must be breached and a main steam isolation valve would need to fail to close. The probability of this sequence of events occurring is low. In addition, large dry containment buildings have been demonstrated to withstand internal pressure in excess of the design limits.
Corrective Actions The team concluded that the main steam non-return check valves were in a non-conforming condition and th'at the licensee had not fullydemonstrated operability.
In response to the team's concerns, the licensee initiated AR 99-0890. As compensatory actions for the nonconforming condition, the licensee:
1) Submitted a Work Order to lower the position of the counter weights on 3518 and 3519 to reduce required break-away torque, 2) Initiated an evaluation of removing the counterweight assembly and arms to further reduce the required break-away torque, 3) Initiated an evaluation of modifying the check valves to remove the packing glands and use a different type of sealing mechanism, and 4) Initiated an evaluation of a procedure change to provide for backup manual closure of the check valves.
In addition, the licensee initiated a computer calculation to determine the peak containment pressure resulting from a main steam line break inside containment, concurrent with a failure of one MSIVto close, and with less than 775 ft-Ibs of steam flowforce on the non-return check valve such that it would remain open. The licensee's calculation determined that containment pressure would peak at 55 lbs., which was less than the design pressure of 60 lbs. Based on that calculation, the licensee concluded that the main steam non-return check valves were currently operable and not in a nonconforming condition.
In response to the team concerns regarding not performing safety evaluations for test procedure changes, the licensee initiated AR 99-1000, "Potentially Inadequate 50.59s for Changes to PT-2.10.15." This AR addressed PCN 92T-0127, PCN 934130, and PCN 99R171 and noted that they had not appropriately addressed the fact that a smaller than design basis steam line break could result in the blowdown of more that one steam generator.
The AR also noted that a required 50.59 safety evaluation was not always included and that the UFSAR had not been update In addition, the licensee also initiated AR 99-0959, "Action Report 99-0890 on Main Steam Line Check Valve Should Have Classified Condition as Nonconforming." This AR was to address potential weaknesses in the areas of Action Reporting, Operability Determination process, and related training deficiencies.
Conclusion In general, the operability determinations reviewed were acceptable.
A few of the operability determinations reached an appropriate conclusion, but were not thoroughly documented.
One operability determination, regarding the main steam non-return check valves was inadequate.
The assumptions, analytical methods, and calculations used by the licensee to declare the main steam non-return check valves operable may not be conservative and may not be applicable in all cases.
The licensee did not show that the uncertainty in the calculation is less than the available margin of torque needed to close the valve.
Therefore, operability of the main steam non-return check valves remains an open issue pending NRC review of additional information from RG8 E (See Attachment 2 of this report for additional questions).
The team identified several inadequate safety evaluations related to changes made to the main steam non-return check valves.
Specifically, the valves were 'changed from free swinging gravity closing valves (as stated in the Updated Final Safety Analysis Report) to valves that required a substantial and increasing external force to close them, without addressing potential effects on steam generator integrity, containment integrity, steam generator tube integrity, reactor reactivity, or reactor vessel integrity. Other procedure changes failed to include safety evaluations.
The team believes that changing the main steam non-return check valves to require a significant breakaway closing torque represents an Unreviewed Safety Question.
This is an apparent violation of 10 CFR 50.59.
(EEI 50-244/99%5-01).
At the exit meeting on May 28, 1999, the licensee did not agree that these changes introduced an Unreviewed Safety Question.
Onsite and Offsite Review Committees Ins ection Sco e 40500 The inspectors reviewed meeting minutes, attended onsite and offsite review committee meetings, interviewed committee members, and reviewed action tracking systems to determine the extent of committee involvement, oversight, and independence.'bservations and Findin s The team noted that members of both the onsite Plant Operations Review Committee (PORC) and the offsite Nuclear Safety Audit and Review Board (NSARB) asked good questions and initiated action items which were adequately tracked.
During a PORC meeting, every member contributed substantially, indicating that they were both knowledgeable and prepare The'team noted that the NSARB was lacking in members with a current or broad nuclear industry experience outside of Ginna. The NSARB had 10 members, including the chairman and vice-chairman, and five of those participated in daily site activities and had offices onsite..Those five included the Site Vice President (who was the NSARB Vice-Chairman), the Plant Manager, the Manager of Nuclear Assessment, the Manager of Nuclear Engineering Services, and the Manager of Nuclear Training. The other five members included Senior Vice President (who was the NSARB Chairman), the corporate Manager of Marketing and Sales (who had formerly worked at the Ginna plant), a retired former Ginna engineering manager, and two current managers from the nearby Nine Mile Point nuclear plant.
The team noted that the current NSARB procedure included the applicable quality assurance program requirements on membership.
Those requirements stated that no more that a minority of the NSARB members could be "members of the onsite operating organization." The purpose ofthat requirement was to provide independence from the activities that the NSARB was overseeing and to add current and broad industry experience to the committee.
The team found that NSARB members and the site Quality Assurance Manager did not have a clear understanding of which members could be considered to not be "members ofthe onsite operating organization." The NSARB Chairman stated that he had recognized a need for another offsite member and was already pursuing that goal. The team noted that the NSARB had an open action item, assigned to the Chairman, to obtain another offsite member.
While attending a PORC meeting, the team observed a presentation on the sequence of events and causes of a reactor trip that occurred on April23, 1999.'he written event evaluation report included several potential root causes.
One was the fact that the lead technician who was involved in causing the trip was tired and had worked 11 consecutive 12-hour days.
PORC members discussed and challenged some of the potential root causes, but notably did not pursue the excessive overtime. The overtime was apparently not considered to be a significant contributor to the trip.
'Since information on the use of overtime was notably lacking in the PORC meeting, the team followed up on the issue.
The team found that the officialwork record for the lead technician, who had been involved in causing the trip, showed that the day of the trip was his 11th consecutive work day without a day off. During the 10 days prior to the trip, he had worked 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> on seven days, 13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br /> on two days, and 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> on one day.
He had substantially exceeded the limitof 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> in a seven-day period.
During the seven day period from April 14 to April20, he had worked 86 hours9.953704e-4 days <br />0.0239 hours <br />1.421958e-4 weeks <br />3.2723e-5 months <br />.
He also apparently exceeded the limitof 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> in a 48-hour period on two occasions during that 10 days.
Further, he had worked 84 hours9.722222e-4 days <br />0.0233 hours <br />1.388889e-4 weeks <br />3.1962e-5 months <br /> in an earlier seven-day period from April 5 through April 11. The licensee had no written authorization for this technician to exceed the limits on maximum work hours on any of these occasions, which violated the requirements on overtime.
In addition, his assignment as a lead technician while working in excess of the overtime restrictions violated the licensee's requirements.
Further, the team noted that the licensee's event investigation report stated that the lead technician had asked to have the day of April23 off, but his request had been denied by his supervisor.
In response to the NRC-identified issue, the licensee initiated AR 99-097 When asked for the written overtime authorization, the manager of the lead technician showed the team an E-Mail memo from the Plant Manager to the timekeeping clerks dated April 9, 1999. The memo stated that it was authorization for overtime exceedances, that may have occurred as a result of a decision to curtail work activities on Easter Sunday (April4), and that may have exceeded the 72-hour rule. The manager of the lead technician stated that the memo was the written authorization for the overtime in question.
The Plant Manager stated that his memo did not authorize anyone, including the lead technician in question, to exceed the site overtime restrictions after April 9.
In addition, they did not use the form prescribed by procedure A-52.10,
"Overtime Policy for Key Maintenance Personnel," Rev. 9. They did not list the individuals who were authorized to exceed the overtime restrictions and did not state when individuals were authorized to exceed the overtime restrictions. The dates during which the blanket overtime authorization was in effect were not clearly stated.
Also, there was no stated limiton how long the overtime restrictions could be exceeded without having a day off. The team concluded that the Plant Manager's authorization to exceed overtime restrictions around the time of Easter Sunday, without including clear limits or controls, indicates a licensee weakness in control of overtime. This item remains unresolved pending further review by the NRC. (URI 50-244/99-05-02)
C.
Conclusion
~ The inspectors identified multiple examples (April 1999) that appeared to exceed the overtime limits of the Technical Specifications.
Specifically., a maintenance technician substantially exceeded site overtime restrictions without the required written authorization.
While exceeding overtime restrictions without the required authorization, the technician was involved in causing a reactor trip. This item remains unresolved pending further review by the NRC. (URI 50-244/99-05-02)
E7.3 0 eratin Ex erience Feedback Ins ection Sco e 40500 The team selected several NRC Information Notices (INs), INPO Significant Operating Experience Reports (SOERs), and INPO Significant Event Reports (SERs) from 1998 and 1999 to assess the effectiveness ofthe licensee's applicability reviews, dissemination of information to licensee personnel, and initiation and tracking of any corrective actions.
b.
Observations and Findin s The team verified that the licensee had received and evaluated each of the selected INs, SOERs, and SERs, and had entered each into a computerized tracking system called the Commitment and Action Tracking System (CATS). The CATS included assignments and due dates for reviews and corrective actions along with a description of each issue and the related corrective actions to be complete '
The team found that the licensee's operating experience feedback reviews and corrective actions, for the selected issues, were reasonably comprehensive and timely.
The team concluded that the licensee's program was adequately addressing industry operating experiences.
Conclusion The team concluded that the licensee was adequately addressing industry operating experiences.
E7.4 Self-Assessments a.
Ins ection Sco e 40500 The team performed a review of the use and effectiveness of self-assessments.
Information was gathered by a review of several peer-assisted self-assessments and the licensee's responses to the associated findings. The team interviewed Ginna personnel, including department managers and a self-assessment team leader, and reviewed responses to self-assessment findings.
Observations and Findin s To enhance overall performance, RG8 E senior management committed the organizations involved with the operation and support of the Ginna station to perform a series of peer-assisted self-assessments using criteria selected from an industry standard {Institute of Nuclear Power Operations [INPO] Publication 96-006,
"Performance Objectives and Criteria for Operating Nuclear Electric Generating Stations" ). In early 1997, teams were assembled and self-assessments were conducted in the functional areas of operations, maintenance, engineering, radiation protection, and chemistry. Assessments were also conducted in the "softer" areas of safety culture, self-evaluation, human performance, and organizational effectiveness.
The assessment results identified programmatic strengths and weaknesses, and were presented to senior management in a self-evaluation report. Overall, the team found the quality of the assessments to be very good.
The self-assessments included industry peers, were generally well written and critical, and contained recommendations for improvement.
Line organizations responded to the self-assessments by evaluating the findings and recommendations, implementing corrective actions or program enhancements, and, ifnecessary, initiating ARs.
Examples of program enhancements initiated as a result of the self-assessments during 1997/98 included:
the operations group lowered the threshold for managing reactivity changes; the shift supervisor became more involved in MOPAR and PORC meetings,
'i the radiation protection and chemistry organizations increased the use and publication of station goals, and
the maintenance department initiated various plant cleanups and painting projects.
Furthermore, the team noted that the operations group maintained a self-monitoring program that appeared to gain performance insights and improve performance.
However, some managers acknowledged that their staff perceived the peer-assist as an
"outside-audit" rather than as a true self-assessment, and that the assessments were conducted only to identify weakness.
C.
Conclusion The team concluded that the peer-assisted self-assessments were good; they included a strong independent perspective and numerous findings for improvement.
Corrective actions and program enhancements resulted from the assessments.
V. Mana ementMeetin s
X1 Exit Meeting Summary Meetings were held periodically with licensee management during this inspection to discuss e
inspection observations and findings. A summary of preliminary findings wa's discussed at the conclusion of the on-site inspection on May 28, 1999, and a telephone call on June 24,1999, from the NRC Region I office.
PARTIALLIST OF PERSONS CONTACTED.
P. Wilkins J. Widay T. White G. Wrobel W Thomson T. Alexander M. Lilley R. Popp G. Hermes F. Puddu T. Laursen N. Goodenough Sr Vice President Plant Manager Manager Operations Manager NS&L Manager RP8 Chemistry Manager Operations Review Manager Quality Assurance Production Superintendent Acting Mgr. Primary System Engineering Senior Analyst Operations Experience Operating Experience Analyst Maintenance
INSPECTION PROCEDURES USED IP 40500 Effectiveness of Licensee Controls for Identifying, Resolving, and Preventing Problems IP 92901 Follow-up Operations Opened ITEMS OPENED, CLOSED, AND DISCUSSED EEI 50-244/99-05-01 Unreviewed Safety Question/Inadequate Safety Evaluation for Main Steam Non-Return Check Valve.
Opened URI 50-244/99-05-02 Breach of Overtime Requirements for the On-site Personnel Closed IFI 50-244/98-01-01 Tracking of Corrective Action Closures
uestlons Re ardin Ginna's Main Steam Non-Return Check Valves A review by the NRC of Calculation DA-ME-92-147, Rev. 2, dated 5/27/1999, "Main Steam Non-Return Check Valve Closure Analysis," for the Ginna Station of Rochester Gas and Electric Corporation, indicates that the assumptions, analytical methods, and calculations are not conservative and may'not be applicable in all cases.
The NRC review observed that the steam flow past the check valve, with flow reversal occurring at the time of the incident, presents a very complicated flow geometry and the detail flow field and pressure distribution on the valve is needed to properly analyze the effects on the check valve. The licensee must show that the uncertainty in the calculation is less than the available margin of torque needed to close the valve. This needs to be demonstrated for breaks less than full double-ended guillotine breaks such that it represents the worst conditions regarding steam line break size and associated flow past the check valve attempting to close it. Lower flow rates from a less than full break would put even less closing torque on the valve.
Because of the NRC concern, the following questions are provided:
2.
3.
5.
6.
7.
8.
What analysis method willbe used'?
It is felt that a 3-dimensional computational fluid dynamics code is needed to properly model this complex flowcondition. How willthe analysis performed be validated for this type of application?
What is the basis for assuming that the steam flow is non-compressible?
Since flow in the line is changing in mass flow rate and reversing direction, what is the basis for assuming constant pressure (during normal operation the flow past the check valve is about 914 Ib@sec.; then, subsequent to the line break the flowat the check valve reverses and decreases to 603.3 Ib@sec.)?
Is the mass flow rate of 603.3 Ib@sec in the calculation based on choked flowat the exit?
How was the mass flow rate coming from the 'line break'G considered in the calculation of the 603.3 Ib@sec coming from the 'operational'team Generator?
What is the basis for assuming the check valve closes in one (1) second'
What is the basis for the check valve disc being treated as a flat circular disc? Won'
there be flow on both sides of the disc since the disc is round with gaps between the disc and the valve body?
What are the area and dimensions of clearance between the open disk circumference and the valve body? This information is needed to determine the area that is available for steam fiowto exit the space above the open disk. And please provide, ifreadily available, in conjunction with your analysis,
~
the cross-sectional area:
~
for steam flowto enter the area above the open disk,
~
inside the inlet pipe to the valve,
~
at the most flow restrictive point inside the open check valve (e.g., the minimum throat area),
~
inside the outlet pipe from the valve, and, the volume:
~
above the disk,
~
in the valve body upstream ofthe minimum throat area, in the valve body downstream ofthe minimum throat area.