ML20137M416
| ML20137M416 | |
| Person / Time | |
|---|---|
| Site: | Saint Lucie  |
| Issue date: | 12/21/1995 |
| From: | Johnson P FLORIDA POWER & LIGHT CO. |
| To: | |
| Shared Package | |
| ML20137M095 | List: |
| References | |
| FOIA-96-485 NUDOCS 9704080082 | |
| Download: ML20137M416 (36) | |
Text
_-
.i r
INDEPENDENT ASSESSMENT 4
OF MAINTENANCE RULE IMPLEMENTATION at the l
ST. LUCIE PLANT Florida Power & Light Company i
l December 11 -15,1995 l
Conducted by Stuart A. Scow, Senior Engineer, Nuclear Assurance Division Palo Verde Nuclear' Generating Station i
and Philip H. Johnson, Maintenance Rule Technical Specialist issued for the Team:
December 21,1995 v
PHILIP H. JOHNSON MAINTENANCE Rule TECHNICAL SPECIALIST 815 BROOKsIDE DRIVE, DANVILLE CALIFORNIA 94526 (5101 d10-5744 1
97040E0082 970401 I
-rs~
l-INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 Il e
TABLE OF CONTENTS Executive S umma ry......................................
i ii Recommendations Assessment Report 1.
Ba c kg ro u n d....... '................................
1 2.
Scope of the Assessment..............................
1 2
3.
Overal: Program Status 3
4.
Program Definition 5.
Responsibilities and Organization.........................
4 6.
Sco pe of the Rule...................................
5 7.
Determination of Risk Significance........................
6 8.
Performance Criteria.................................
7 9.
Performance Monitoring..............................
12
- 10. Dispositioning SSCs into (a)(1) or (a)(2); Establishing Goals 16
- 11. Considering Plant Safety When Removing SSCs from Service....
19
- 12. Balancing Unavailability and Reliability................. '...
22
- 13. Considering industry Operating Experience.................
22
- 14. Periodic Assessments of Meintenance Effectiveness 23
- 15. Documentation of Maintenance Rule Activities..............
25
- 16. Tra i n i n g.........................................
26
- 17. Individuals Contacted During the Assessment...............
27 Attachment A -- Specific Procedure Comments 28
- Attachment B -- Specific Technical Comments..................
32 Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL sPECIAUST
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST LUCIE PLANT - DECEMBER 1995 EXECUTIVE
SUMMARY
h 10 'CFR 50.65, Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants becomes effective on July' 10,1996. This regulation, commonly known as The Maintenance Rule, requires that structures, systems, and components (SSCs) within its scope be monitored against established performance l
criteria, and that goals be defined to provide improved performance of SSCs which have not satisfied their performance criteria. Other program e'lements specified in i
j the Mrintenance Rule include periodic evaluations of the maintenance program and assessenent of safety impact when removing SSCs from service for maintenance.
This report provides the results of a one-week independent assessment of actions being taken to implement the Maintenance A!e at the St. Lucie Plant. The assess-7 ment was conducted to evaluate the overall approach and progress to date toward initial implementation of the rule. The assessment was based upon the requirements of 10 CFR 50.65 and the guidance provided in related NRC and NEl documents.
1 I
' Although substantial work has been done on implementation, the assessment team concludes that St. Lucie's Maintenance Rule program does not currently meet the expectations of the rule and related guidance. Overall, little progress toward implementation of the Maintenance Rule has been made since July 1995, when a j
realignment was made in the plant organization and Unit outages began to demand heightened management attention. Significant dedicated effort and management attention will be needed to ensure timely implementation.
- Maintenance Rule scoping and risk significance determinations appear to have been completed effectively, and are considered a program strength. Observations of principal concern to the team, however, are the following: (1) many performance criteria appear inconsistent with industry norms and NEl guidance; (2) historical SSC l
performance will need to be reexamined after performance criteria are reestablished; (3) monitoring is not well defined; (4) the review and goal setting process has not been completed for systems.with. observed poor performance; and (5) limited progress has been made in developing other required program elements.
Full compliance with the Maintenance Rule is required by July 10,1996, and a Unit 1 refueling outage will require significant attention before that date. The assessment team recommends that (1) significant management emphasis and dedicated resources be allocated to prograrn completion, (2) in view of the Unit 1 refueling outage, a completion deadline no later than March 31,1996 be estab-lished, and (3) a comprehensive reassessment be conducted by QA or another appropriate group before the end of June 1996.
The principal recommendations resulting from this assessment appear on the next page. Other recommendations and suggestions are included within this report.
Compiled by PHILIP H. JOHNSON -
MAINTENANCE RULE TECHNICAL SPECIALIST i
o.-
g 4- -
e
,m gavn v
g
,w
,,we,-.
4 o -ew
, - e
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 ASSESSMENT REPORT b
1.
. BACKGROUND 10 CFR 50.65, Requirements for Monitoring the Effectiveness of Maintenance at Nucicar Power Plants, becomes effective on July 10,1996. This' regulation, commonly known as The Maintenance Rule, requires that structures, systems, i
and components (SSCs) within its scope be monitored against established performance criteria. SSCs which meet their performance criteria are considered to be performing effectively pursuant to paragraph (a)(2) of the Maintenance Rule. SSCs which exceed their performance criteria must be considered for goal-setting, pursuant to paragraph (a)(1), to provide for improved performance. The Maintenance Rule also requires that (1) the program be periodically evaluated, (2) industry operating experience be considered in certain aspects of the program, (3) SSC unavailability for monitoring or maintenance be balanced against the resulting improvement in reliability, and (4) the resulting safety impact be considered when removing SSCs from service for monitoring or maintenance.
In issuing the Maintenance Rule, the NRC characterized it as a "results-oriented rule" -- that is, that it principally addresses the results required of the utility's program without specifying the methods of achieving them, in response, the Nuclear Management and Resources Council (NUMARC, now the Nuclear Energy Institute, or NEI) issued NUMARC 93-01, Industry Guide /ine for Moni-toring the Effectiveness of Maintenance at Nuclear Power Plants (Revision 1 of NUMARC 93 01 has been circulated for industry review and comment). The NRC subsequently issued Regulatory Guide 1.160, Monitoring the Effective-ness of Ma/ntenance at Nuclear Power Plants (reissued as Revision 1 in January 1995), which states that NUMARC 93-01 provides methods accept-able to the NRC staff for complying with 10 CFR 50.65. RG 1.160 also states that licensees may use other methods to meet the intent of the Mainter:ance Rule, with the NRC to determine the acceptability of such other methods on a case-by-case basis.
2.
SCOPE OF THE ASSESSMENT This was a one-week independent assessment, emphasizing industry and regulatory perspectives, of actions taken or in progress to implement the Maintenance Rule at The St. Lucie Plant. The assessment was based upon the direction or guidance provided in 10 CFR 50.65, NUMARC 93-01, NRC RG 1.160, and the NRC's inspection Procedure 6270E, " Maintenance Rule."
The results of nine pilot inspections conducted by the NRC in 1994 -1995 (using a draft of Inspection Procedure 62706) were also considered.
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST 1
l l*
i i
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 Some elements of St. Lucie's Maintenance Rule program are still being developed, and were considered insofar as practicable. In some cases (e.g.,
periodic assessment), this was limited to review of draft procedures or instructions.
3.
OVERALL PROGRAM STATUS
\\
a.
Substantial work has been done on implementation of the Maintenance Rule, but significant additional effort will be needed before the rule becomes effective on July 10,1996.
Strenath.. Scoping and risk significance determinations have been completed in an effective manner.
Performance criteria have been established, although many of these appear inconsistent with industry norms and with the guidance i
provided in NUMARC 93-01 and endorsed by the NRC. The assess-ment team recommends that these be re-reviewed (discussed further in Section 8 below).
The plant's approach to the monitoring of functional failures or i
maintenance preventable functional failures is not consistently applied.
Historical SSC performance should be reexamined after the performance criteria have been reviewed for consistency with NUMARC guidance.
A number of systems have been placed in paragraph (a)(1) of the rule, but the review process and the establishment of corrective actions and goals do not appear to have been completed for any of these. In the absence of measurable goals, little meaningful monitoring of (a)(1) SSCs is being conducted.
Little progress toward implementation of the Maintenance Rule has been made since July 1995 because of a realignment in the plant organization and the attention required by planned and unplanned outages.
Some progress has been made in establishing program elements 6
required by paragraph (a)(3) of the rule (periodic assessments, balancing unavailability and reliability, and assessing the safety impact of removing equipment from service), but additional work.is needed in these areas.
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST i
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 1
b.
Overall, the assessment team concludes that the Maintenance Rule program does not currently meet the expectations of the rule and related guidance, and that significant dedicated effort and management attention will be needed to ensure timely implementation.
4.
PROGRAM DEFINITION The DRAIT Program Compliance Plan (PCP) defines initial implementation of the Maintenance Rule, and needs to be approved and issued. ADM-17.08 (Revision 1), " Implementation of 10 CFR 50.65, The Maintenance Rule,"
i addresses ongoing functioning of the program.
a.
The assessment teara recommends that these documents be reviewed for consistency with one another and with the guidance in NUMARC 93-01, while considering the observations arid comments in the various sections of this assessment report. Although not specifically mentioned elsewhere i
in this report, the process for changing scoping and risk significance after i
initialimplementation should also be defined. Specific comments regarding these documents are provided in Attachment A.
i b.
Although a specific review was not performed, the team did not see any indication that existing procedures which may interface with the Maintenance Rule program had been revised to reflect this. For example:
Administrative procedures for the Nuclear Plant Work Order (NPWO) and St. Lucie Action Report (STAR) programs should be. revised to better interface with the Maintenance Rule program. Specifically, these programs must ensure that all potential failures are captured for Maintenance Rule program review to determine whether a functional failure or maintenance preventable functional failure (MPFF) occurred.
Design change procedures should specify that any change to plant design which could affect scoping or risk significance should be communicated to the Maintenance Rule Coordinator.
' Procedures for the processing of industry operating experience should be revised to reflect the need to communicate certain types of industry information to the Maintenance Rule Coordinator and/or the System / Component Engineers.
Significant changes to emergency operating procedures (EOPs) which could affect Maintenance Rule scoping should be brought to the attention of the Maintenance Rule Coordinator.
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST -
1 INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 i
5.
RESPONSIBILITIES AND ORGANIZATION a..
Significant> effort has been expended to date in initial program implementation. However, implementation of the rule appears to have made little progress since about July 1995, when an organizational realignment was made and unit outages began to require significant attention. The assessment team recommends that increased, dedicated resources be promptly allocated to the Maintenance' Rule to provide for timely implementation of the program.
b.
Although ADM-17.08 assigns responsibilities to both the Working Group and the Expert Panel, discussions during the assessment indicated that I
the Working Group will cease (or has ceased) to exist as a functioning body. The team recommends that the Expert Panel's responsibilities be reviewed to ensure that functions previously performed by the Working i
Group will be performed by the Expert Panel. The Expert Panel's responsibility to review dispositioning to (a)(1) and (a)(2) should also be included.
i c.
The assessment team observed very little involvement in Maintenance Rule activities by individuals other than those charged with initial implementation. Excessive reliance appears to have been placed on the Maintenance Rule Coordinator for development and ongoing functie..ing of the program (e.g., identification of all functional failures).
Additional management emphasis should be provided to promote ownership of certain Maintenance Rule activities by appropriate organizational groups at St. Lucie. For example, the assessment team recommends that System and Component Engineering have lead responsibility for identifying functional failures and for monitoring all (a)(1) SSCs and those (a)(2) SSCs which have specific performance criteria.
d.
The assessment team also observed minimal understanding of the Maintenance Rule among other than those directly involved in its implementation. Emphasis is recommended to more fully integrate the rule into existing plant processes. Note that this should not involve significantly increased effort (after initial implementation) on the part of most organizations. Rather, the rule should be integrated as a thought process to be applied to day-to-day activities (e.g.,
sensitivity to potential functional failures and SSC availability criteria in the conduct of scheduled maintenance and the review of plant problems),
j Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST i -
~
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 l
6.
SCOPE OF THE RULE The scope of the Maintenance Rule, as defined in 10 CFR 50.65(b), includes safety-related SSCs and those non safety-related SSCs (1) that are relied upon to mitigate accidents or transients, or are used in emergency operating procedures (EOPs); (2) whose failure could prevent safety-related SSCs from fulfilling their safety functions; and (3) whose failure could cause a reactor scram or safety system actuation. Guidance on the conduct of scoping activities is provided in Section 8.2 of NUMARC 93-01.
Strenath. Overall, the selection of structures, systems, and components (SSCs) within the scope of the rule appears to have been done well. Initial scoping was done by Juno Engineering, with refinements by the Working Group. Specific comments related to the scoping effort are as follows:
The Juno Engineering document identified itself as Revision A, and a.
indicated that other updates (Revisions B, C, and 0) were intended before it would be considered final. In addition, some refinements to scoping (e.g., including containment isolation valves within the Containment system) were made by the Working Group. The documentation of scoping activities needs to be completed. This documentation should include the basis for.= coping decisions and should show, without a need for significant explanation by those involved, how scoping was.
conducted, b.
While not specifically required by the Maintenance Rule itself,.18.2.1.5 and 8.2.1.6 of NUMARC 93-01 specify that industry operating experi-ence for the past two refueling cycles should be reviewed to determine SSCs which have caused scrams, safety system actuations, and impact by non-safety related SSCs on safety related SSCs. Scoping documents should include a listing or reference to the industry operating experience which was reviewed in the scoping process.
NOTE: The Maintenance Rule and NUMARC 93-01 apply safety system actuations in two different ways. Paragraph (b)(1) of the Maintenance Rule states that SSCs whose failure could cause "actuaticn of a safety-related system" should be within scope.
In the conduct of scoping, this should be applied to actuation of any safety-related system. On the other hand, an Unplanned Safety System Actuation (as used by some plants in plant level performance criteria) is defined in NUMARC 93-01 to include
" unplanned emergency core cooling system actuations or i
emergency AC power system actuations due to loss of power to a safeguards bus."
Compiled by' PHILIP H. JOHNSON MAINTENANCE Rule TECHNICAL SPECIALIST -
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995
~
c.
Review of a listing of plant systems indicated the results of the scoping to be generally appropriate; that is, the listing of systems within scope appeared consistent with that generally seen in the industry. The follow-ing exceptions were noted, however. These should be re-reviewed in the light of current guidance. If it appears that these should not be in scope at St. Lucie, the basis for excluding them should be clearly documented:
The grounding and cathodb protection systems (Systems 60 and
~
75, respectively) -- The NRC indicated during its pilot inspections that these could be in scope. Has the grounding system resulted in a scram or safety system actuation at St. Lucie or a similarly configured plant? In view of reported corrosion experience at St.
Lucia, could failure of the cathodic protection system lead to failure of a safety related system?
4 it was noted that the fuel handling building ventilation system is not j
currently in scope. Does this system at St. Lucie have a safety-related function (e.g., maintain a negative pressure or otherwise provide a boundary in the event of airborne activity)?
1 7.
DETERMINATION OF RISK SIGNIFICANCE-Paragraph (a)(1) of the Maintenance Rule states that goals "... shall be established commensurate with safety...." The guidance o.f NUMARC 93-01 provides for first assessing the risk significance of SSCs as a basis for the establishment of reisted performance criteria.
Section 9.3.1 of NUMARC 93-01 recommends the use of an Expert Panelin determining risk significance, to compensate for the limitations of probabilistic safety assessment (PSA) approaches (e.g., commonly available only for Mode 1). It further discusses three specific risk importance calculational methods, and recommends review of system performance against critical safety functions as an alternative method.
Strenoth. The determination of SSCs which are risk significant appears a.
to have been done well, closely following the' guidance defined in NUMARC 93-01. The three risk methods (and associated cutoff values) recommended in the NUMARC guideline were used, with the results provided to the Working Group (analogous to the expert panel described in NUMARC 93-01).
b.
The approach to risk determination used at St. Lucie was conservative, actually using both alternative forms of the risk reduction worth method for a total of four inputs. Any one input above the recommended cutoff value generally resulted in the SSC being categorized as risk significant.
Compiled by PHILIP H. JOHNSON MAWTENANCE RULE TECHNICAL SPECIALIST 4
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995
]
j i
As a consequence, noout 55% of the systems within Maintenance Rule l
l scope are risk significant, compared with an average in the industry of 35 to 40%. It is possible that the higher percentage of risk significant systems at St. Lucie may have resulted in part from system boundaries which result in a smaller total number of in-scope systems. The higher percentage of risk significant systems at St. Lucie is not a regulatory concern, but offers two potential disadvantages:
lt spreads the additional attention which is intended for risk significant systems over a larger number of systems. Since the same risk significant determinations may later be used for risk-based regulation or graded QA activities, it could also result in the realization of less benefit from those programs.
lt requires specific performance. criteria and individual system l
rnonitoring for more systems, as opposed to the possibility of monitoring more systems based on overall Unit performance (against plant level criteria).
NOTE: Plant systems which are risk significant should be so categorized, and no indication to the contrary is intended.
However, it is recommended that systems which appear to be marginal with respect to risk significance be reevaluated to determine whether undue conservatism was applied. The plant may wish to evaluate in particular those systems which were,
below the cutoff on only one of the three principal risk deter-mination methods (risk reduction worth, risk achievement worth, and total core damage frequency).
8.
PERFORMANCE CRITERIA Section 9.3.2 of NUMARC 93-01 recommends that performance criteria be established as a basis for evaluating SSCs. SSC-specific performance criteria are usually established for risk significant SSCs and for non-risk significant SSCs which are normally in a standby condition. Plant level performance criteria are used to monitor all other SSCs, acept those which cannot be effectively monitored by plant-level criteria; in such cases, SSC-specific criteria are used. SSCs which are satisfying their performance criteria are considered to be within the scope of Paragraph (a)(2) of the Maintenance Rule -- i.e., they are being effectively controlled by the preventive maintenance program.
St. Lucie's timely establishment of performance criteria for Maintenance Rule SSCs is recognized. However, many of the performance ctiteria (both SSC-specific and plant level) do not follow the guidance in NUMARC 93-01, and are Compiled by PHILIP H. JOHNSON MAINTENANCE RutE TECHNICAL SPECIALIST INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 therefore different from those being used by most utilities. The NRC has endorsed NUMARC 93-01 as an acceptable method of implementing the Main-tenance Rule. While other methods may also be acceptable, the NRC will judge the appropriateness of these on a case-by-case basis. Since this Introduces additional regulatory uncertainty, it is recommended that the NUMARC guidance be used unless there is a good reason to do otherwise.
Based on review of the PCP, ADM-17.08, and existing performance criteria, the following comments are presented:
Soecific Performance Criteria. Many of the established SSC-specific per-a.
formance criteria do not appropriately address the guidance in NUMARC 93-01 (19.3.2) and from the NRC, which states that risk significant and standby SSCs should have specific train level performance criteria which assure that the reliability and availability assumptions in the PSA are maintained.
(1)
The performance criteria for unavailability should be related to that assumed in the Unit's PSA. Generally,5% was used as a perform-ance criterion for unavailability,,although a lower unavailability was in most cases assumed in the PSA. For example, the PSA baseline used 0.8% (73 hours8.449074e-4 days <br />0.0203 hours <br />1.207011e-4 weeks <br />2.77765e-5 months <br /> per train per year) unavailability for high i
pressure safety injection (HPSI - pump and flow path). Recalcula-tion using 5% unavailability yields a 32% increase in core darnage frequency. The nexus between unavailability criteria and the Unit's PSA will be important to the balancing of reliability and availability, discussed later.
NOTE: It is recommended that PSA-assumed unavailability be used as a starting point for performance criteria. If this does not allow sufficient out-of-service time in some cases for necessary maintenance, use a higher value on a case-by-case basis, and assess the totalimpact on core damage frequency which would result. Criteria expressed as hours unavailable during the most recent 12 months facilitate monitoring activities. This " rolling" 12-month period is suggested for monitoring, since it gives a good indication of recent performance, but allows a period of poor per-formance to be dropped from the indicator after one year.
(2)
The existing unavailability performance criteria are applied on a quarterly basis - that is, the system would be considered for (a)(1) status if the criterion were exceeded for one quarter, even if average performance over the past year was acceptable. It appears that this approach could result in SSCs being placed in (a)(1)
Compiled by PHILIP H. JOHNSON MAINTENANCE Rule TECHNICAL SPECIALIST i
INDEPENDENT MAINTEOsNCE RULE ASSESSMENT ST. LUCIE PLANT - DECEMBER 1995 prematurely. Unavailable hours over at least the last 12 months
.would appear to provide a more stable yet meaningfulindication of SSC. performance.
(3)
Many of the present performance criteria do not adequately address SSC reliability. Most utilities monitor reliability using performance criteria for maintenance preventable functional failures (MPFFs) or functional failures in the most recent 12 (or 24) months. Please see the discussion of functional failures and MPFFs in 18.b below.
(4)
Some performance criteria (e.g., those for the chemical and volume control system, CVCS) use Technical Specifications (TS) limiting conditions for operation (LCOs) or licensee event reports (LERs) as a performance criterion. These are also not a good indicator of reliability, since many component failures will not exceed an LCO (e.g.,if the condition is corrected within the TS allowed outage time) or require submittal of an LER. The CVCS system summary also indicates that failing one surveillance requirement would exceed performance criteria. This appears inappropriate (overly sensitive?),
since some surveillance test failures may not involve a functional failure. Please see 18.b below.
b.
Functional Failures. Inconsistency was observed in the monitoring of functional failures. The definitions in the PCP and AD.M-17.08 indicate that functional failures, not ma/ntenance preventable functional failures (MPFFs), will be used. Counting all functional failures is a more conservative approach. However, if used, this approach should be consistently applied. Page 20 of ADM-17.08 states that personnel errors by other than maintenance personnel will not be considered, which appears inconsistent.
Most utilities monitor SSC reliability using performance criteria based on MPFFs (see the definition in NUMARC 93-01). Some, including Palo
-Verde, count all functional failures regardless of cause. A functional failure, as discussed in NUMARC 93-01, is the failure of an SSC to perform the function which required its inclusion within the scope of the rule. A functional failure is considered maintenance preventable if its cause is attributable to a maintenance-related activity (broadly applied --
see the definition of Maintenance in NUMARC 93 01).
The assessment team recommends that St. Lucie determine whether MPFFs or all functional failures will be used for performance criteria, and that the appropriate definition be consistently applied. Other comments related to this topic include the following:
Compiled by PHILIP H. JOHNSON MAINTENANCE Rule TECHNICAL SPECIALIST
.g.
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 (1)- If all functional failures will be counted against performance criteria, it should be recognized that this could result in a system being dispositioned to (a)(1) for non-maintenance related causes.
(2)
If MPFFs will be used, it is recommended that procedures and the
'PCP clearly define those functional failures which are not maintenance preventible; for example, as follows:
Functional failures which result from personnel error by other than maintenance personnel and which are not related to corrective or preventive maintenance or monitoring (including surveillance) activities are not counted as MPFFs (refer to the definition of Maintenance in NUMARC 93-01, Appendix B).
A failure which resulted from poor or questionable design need not be considered maintenance preventable the first time the failure occurs if the failure should not have been expected.
However, if the same failure has occurred previously, or is known to have occurred on similar equipment at another site, and appropriate maintenance activity should have prevented the failure, then the failure should be considered maintenance preventable.
If it is not clear whether a functional failure was maintenance preventable, it should be counted as an MPFF.
(3)
The PCP and ADM-17.08 give slightly different definitions of functional failure (the latter refers to safety function, although some Maintenance Rule functions may not be safety-related). Note that the determination of functional failure relies heavily on having identified the SSC's functions (i.e., the functions which caused it to be included within the rule's scope). For a safety related system, this could be any of its safety functions, plus not impacting plant level performance criteria. On the other hand, the Maintenance Rule function of the extraction steam system could be defined as not causing a scram or an unplanned loss of plant capacity.
(4)
Many of the specific performance criteria currently use < 1 (i.e.,
zero) repetitive functional failures. This criterion may monitor the effectiveness of corrective actions, but does not effectively monitor SSC reliability. It also does not add meaning to the criteria, since any SSC which experiences a repetitive functional failure must be dispositioned to (a)(1) regardless of performance criteria. In i
addition, the NRC has expressed concern with the use of zero values in performance criteria, since the number zero cannot be Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST INOEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 trended (i.e., no warning is provided before the criterion is exceeded).
For these reasons, the assessment team recommends that repetit/ve functional failures not be used as a performance criterion, although ADM-17.08 should clearly require dispositioning to (a)(1) for corrective actions and goat-setting when one occurs. This is discussed further in Section 9, Monitor;ng.
Plant Level Performance Criteria. NUMARC 93-01 recommends that non-c.
risk significant and non-standby SSCs be monitored by plant level performance criteria (PLPC). NUMARC recommends criteria for plant scrams, unplanned caoacity loss, unplanned safety system actuations, and other parameters.
(1)
St. Lucie is presently using similar criteria, but is applying them on a system level. While this appears workable in some regards, it presents concerns in others:
A criterion of 95% plant availability (i.e., no more than 5%
capability loss) per quarter for each monitored system is presently > sed. By this approach, three systems could each cause 4% of capability loss (total Unit capability loss of 12%)
without any system exceeding its performance criteria.
A criterion of <1 (i.e., zero) reactor trips per cycle. per system is used for a number of systems. As noted previously, the number zero is not trendable. However, use of $1 or <2 per cycle could allow one trip each from several systems without any performance criteria being exceeded.
Plant level performance criteria which measure overa// Unit performance appear to be more easily managed. These should be consistent with corporate goals or other performance indicators monitored by management. As long as plant level performance meets the criteria, SSCs monitored by the PLPC can be considered to be performing properly and remain under paragraph (a)(2) of the rule. Whenever one of the plant level criteria is exceed, the SSC(s) which principally caused it to be exceeded should be considered for dispositioning to paragraph (a)(1) of the rule.
(2)
The team recommends that plant level performance criteria be applied to all systems, including those which also have specific performance criteria. In other words, the reactor coolant system Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST - -
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 (which has specific performance criteria) should be considered for dispositioning to paragraph (a)(1) if it caused PLPC to be exceeded.
The following plant level performance criteria, which are consistent with NUMARC 93-01 and with those used by many other utilities, are suggested for consideration (on a per-unit basis).
s 1 unplanned automatic trip during the last 12 months (consistent with company goals)
~
s 5% unplanned capability loss during the last 12 months (or
[
use days of unplanned capability loss, if desired, to be consistent with the monitoring of company goals) s 1 unplanned safety system actuation during the last 12 months (defined as an unplanned diesel generator (DG) start i
due to an actual low voltage condition, or a safety system actuation) l s 1 TS-reportable. radioactivity release during the past 12 i
months.
Any other plant level criterion which may be needed to monitor the function of SSCs which do not have specific criteria (e.g.,
Unusual Events, shutdown safety events).,
if any plant level performance criterion is exceeded, then the SSC(s) which caused or principally contributed to it should be considered for dispositioning to paragraph (a)(1) of the rule (see 110.e below).
9.
PERFORMANCE MONITORING Monitoring, as discussed in 19.4.2 of NUMARC 93-01, refers to the evaluation of equipment performance against established performance criteria and goals.
During initial implementation of the rule, monitoring also includes a review of historical SSC performance which serves as a basis for dispositioning SSCs to paragraph (a)(1) or (a)(2) (see 19.3.3 of NUMARC 93-01). Monitoring should be conducted in a manner which permits the recognition of performance trends, and should use existing monitoring programs to the extent possible, a.
The review of historical SSC oerformance during the baseline period was conducted for all SSCs to permit their initial dispositioning to paragraph (a)(1) or (a)(2) of the role. However, since many of the performance criteria appear inappropriate, as discussed in Section 8 above, it will be necessary to repeat the baseline review (for overall plant level performance and for SSCs with specific performance criteria) after the Compiled by PHILIP H. JOHNSON MAWTENANCE Rule TECHNICAL SPECIAll5T INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 i
performance criteria are reestablished. Note that the PCP should also be revised to document the methods by which the historical review was conducted as part of initial implementation.
Paragraph 9.3.3 of NUMARC 93-10 states that the initial dispositioning of SSCs to paragraph (a)(1) or (a)(2) of the rule should be based on review of historical SSC performance for a minimum of 36 months. This would apply to the determination of repetitive functional failures or MPFFs. Performance with regard to reliability, unavailability, and/or other established criteria should be reviewed for the most recent 12 or 24 months, consistent with the period specified in the performance criteria.
In conducting this initial baseline review, an 18-month period from July 1993 to December 1994 was used, with the understanding that 36 months would have passed by the time the Maintenance Rule became effective. However, this would appear to require updating the baseline period to the end of June 1996 before the July 10,1996 effective date of the rule. When conducting the re-review of historical SSC performance (after performance criteria are reest'ablished) it is recommended that a baseline period of 36 months (January 1993 through December 1995) be used. No updating of the historical review would then be necessary after initial dispositioning, and monitoring of SSC performance beginning in 1996 would be conducted pursuant to the defined program for ongoing functioning of the Maintenance Rule.
I b.
Onaoina monitorina of SSC performance, under paragraph (a)(1) or (a)(2) of the rule, was not effectively defined in ADM-17.08. Overall, it appears that very little monitoring of SSC performance has been done since Juiy 1995. Discussions and record reviews indicated that an i
orgatational realignment shortly thereafter, combined with planned and unplanned unit outages, resulted in a significant decrease in. Maintenance Rule activity.
The assessment team recommends that ADM-17.08 (and interfacing procedures, if appropriate)' be revised to provide additional direction on the conduct of ongoing monitoring after initial program implementation, for both (a)(1) and (a)(2) SSCs (refer to NUMARC 93-01 19.3.2 and i
9.4.2).
Activities which should be addressed include the following:
(1)
Oroanizational resoorisibilities for monitorina. The tearn suggests assigning lead responsibility to System and Component Engineering (or another appropriate organizational group) for monitoring SSC performance against specific performance criteria. An appropriate group with timely access to plant performance data should have lead responsibility for monitoring Unit performance against plant level performance criteria (PLPC).
l Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST )
. _ - = - -. -. - -
- INDEPENDENT MAINTENANCE RULE ASSESSMENT ST. LUCIE PLANTyDECEMBER 1995 i
4 (2)
The freauency and methods of reouired monitorina. The team suggests that monitoring against SSC unavailability and unplanned capability loss (if adopted as a PLPC) be done monthly, and that the number of MPFFs (or functional failures), scrams, safety system-t actuations, or other discrete event-type indicators be compared to specific or plant level criteria when a monitored event is determined to have occurred.
(3)
Reportino of monitorina results. The team noted that monitoring l.
results (e.g., availability) are reported graphically. However, a bar chart which compares SSC availability of 98 or 99% to a criterion of 2: 95% is less informative than a chart which shows the same' data in the form of 20 availability -- that is, comparing values of 1 or 2% to a criterion of s 5%. The latter provides a much more I
sensitive graphicalindication of the data, and makes trends more j
discernible.
(4)
Evaluation of notential functional failures. The program presently appears to rely on the Maintenance Rule Coordinator to personally identify all functional failures by reviewing STARS, surveillance test records, and other documents..More comprehensive methods i '
should be established to ensure that all potential functional failures are capt'ured in the STAR, NPWO, or another program, and reviewed to determine (1) whether a functional failure occurred and (2) if MPFFs are to be used at St. Lucie, whether any identified functional failure was maintenance preventable. Figure 5, " Determination of l
Functional Failure," of ADM-17.08 aids in the documentation of j
~;
functional failures, but should be reviewed to address " maintenance preventable"if MPFFs are to be used. The team suggests that principal reliance for screening possible functional failures be assigned to an organizational unit (e.g., System and Component i
Engineering) other than the Maintenance Rule Coordinator.
(5)
Determinina renetitive functional failures (or MPFFs. if used).
One person or group should be assigned lead responsibility for maintainin0 a record of functional failures or MPFFs and monitoring
)
\\ for repetitive failures (failure of the same or a similar component for the same cause). This responsibility is presently assigned to the j
1 Maintenance Rule Coordinator, which appears appropriate. If another group is assigned this responsibility, the procedure (s) should
)
require that the Maintenance Rule Coordinator be informed promptly if a repetitive functional failure occurs.
1 NOTE: If review following a functional failure (or MPFF) determines that design changes which might prevent a recurrence will Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST '
1 lNOEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995
\\
not be'made because they are not cost effective, based 'on consideration of estimated costs and the expected safety and/or plant reliability benefits, then a subsequent functional failure (or MPFF) for the same cause need not be i
considered repetitive (reference: NUMARC 93-01, paragraph 9.4.5). ' Any such determination that design i
changes will not be made should be documented and approved by a specified authority (e.g., Expert Panel). The l
documentation should include the reasons why a design l
change to prevent recurrence is not cost effective, including an evaluation of the consequences of future failures and the basis for concluding that "run-to-failure" is an acceptable consequence. In such cases, the SSC's design condition should remain unchanged. Such determinations should be i
retained in appropriate Maintenance Rule files.
(6)
Additional monitorina needed for (a)(1) SSCs. It was not clear to the assessment team how trending is accomplished while monitor-ing SSCs placed under paragraph (a)(1) (see additional comments related to goal-setting in Section 10 below). The Statements of Consideration for 10 CFR 50.65 indicate that, where failures are likely to cause loss of an intended function, monitoring against
)
established goals under (a)(1) should be predictive in nature, providing early warning of degradation. Additional guidance is found in Section 9.4.2 of NUMARC 93-01. Note that SSCs in (a)(1) i should still be monitored against their performance criteria while being monitored against established gocis, (7)
Actions reautred for unacceotable nerformance. Procedures should specify the process and criteria for considering SSC dispositioning to paragraph (a)(1) of the rule (refer to 19.4.4 of NUMARC 93-01).
Section 7.3 of ADM-17.08 did not specify a cause determination for a risk significant SSC failure if the goal or performance criteria were
- met, c.
Equipment unavailability data are presently compiled by System and Component Engineers by review of the equipment out-of-service logs in the vault. A computer-based method would appear to be feasible at minimal cost, and is suggested for consideration. This could function and provide potential benefits as follows:
(1)
Prepare a data base file, using existing hardware and software (e.g.,
the LAN and Microsoft Access), which could receive "out of ser-vice" and "back in service" times. These could.be entered by Operations personnel. Note that Operations entries could be faci-Compiled}by' PHluP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST,
INDEPENDENT MAINTENANCE RULE ASS::SSMENT - ST. LUCIE PLANT - DECEMBER 1995 litated by providing an ability to " click" on a desired system, then on a desired component, etc. Time and date (and possibly LCO return-to-service time) could be added automatically by the computer.
(2). The data base file could be configured to display for Operations use (when considering plant safety before removing SSCs from service, as discussed below) a current listing of Maintenance Rule SSCs which are out of service. It is possible that this display could replace or reduce the scope of the equipment out-of-service log presently being kept.
(3)
The data base could also generate a listing of SSC unavailability periods which, after validation by the System / Component Engineer, would be used in monitoring SSC unavailability.
- 10. DISPOSITIONING SSCs INTO (a)(1) OR (a)(2); ESTABLISHING GOALS Sections 9.3.3 and 9.3.4 of NUMARC 93-01 discuss dispositioning of SSCs to (a)(1) category (i.e., requiring goal setting) during initial implementation.
Sections 9.4.4 and 9.4.3, respectively, discuss (1) dispositioning of SSCs to (a)(1) when unacceptable SSC performance is observed following program implementation and (2) subsequent dispositioning of (a)(1) SSCs back to (a)(2) after acceptable performance has been demonstrated. Dispositioning to (a)(1) must be considered when performance criteria are exceeded, if a repetitive MPFF occurs, or if a failure of a risk significant SSC occurs, even if performance criteria are met.
Sections 9.4 and 9.4.1 of NUMARC 93-01 provide guidance for the estab-lishment of goals. Goals should monitor the cause of the observed poor performance and the effectiveness of corrective actions which were 3
established to address the cause. Goals should be more specific and more trendable than performance criteria, and should facilitate monitoring which can predict whether SSC performance may degrade unacceptably before the next monitoring interval. The Maintenance Rule and NUMARC 93-01 also state that
' risk and industry operating experience should be considered when setting goals.
a.
The initial dispositioning of SSCs to either paragraph (a)(1) or (a)(2) of the Maintenance rule has been completed. However, in view of the need to revise many of the performarice criteria to be consistent with industry norms and NEl guidance, the historical review of SSC performance should be reexamined (please see 19.a). The initial dispositioning of SSCs to (a)(1) or (a)(2) status should then be reexamined based on the revised performance criteria and historical SSC performance against them during the baseline period.
Compiled by PHILIP H.- JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT DECEMBER 1995
~
b.
Procedure ADM 17.08 provided generally appropriate guidance for dispo -
I sitioning SSCs to paragraph (a)(1) after initial implementation of the rule, along withian attached figure which is completed for documentation.: It j
did not appear to address, however, the need for cause determination and review for two of the issues listed in NUMARC 93-01 19.4.4 --
a goal not being met, and failure of a risk-significant SSC when l
performance criteria were not exceeded.
l c.-
The' assessment team noted that the process for placing SSCs into (a)(1) f does not appear to be working effectively. Discussions during the.
i assessment indicated that, although several systems have been dispo-sitioned to paragraph (a)(1) of the rule, the specified documentation has been completed for very few, if any,'and no goals have yet been established except to apply the performance criteria as the goal. For l
example, of three (a)(1) systems reviewed in detail (Unit 1 reactor coolant a
system (RCS), Unit 2 engineered safeguards system, and Unit 2 condensate system), no measurable goals had been identified. One i
example noted was the RCS:
t February 1995 --
experienced code safety valve leakage which prompted shutdown July 1995 --
STAR written to document the concern. RCS j
assigned to (a)(1) for goal-setting.
December 1995 -- Definition of corrective actions and goals still pending In addition, the Unit 1 and Unit 2 diese! generators have been placed into (a)(1) because of mi atitive goverrior failures (May 1995 for Unit 1, following a similar failure in December 1994 for Unit 2). At the time of.
this assessment, the (a)(1) disposition was not included in the system files, and no evidence could be found that a STAR had been initiated as directed by Section 7.3.3 of ADM-17.08. Interviews indicated that no coals have been established, other than to apply the performance criteria as goals. The assessment team notes that this should be reviewed as a j
potential compliance issue, since the Maintenance Rule became effective for the diesel generator system in September 1995.
j d.
In the absence of effective goals, no meaningful monitoring or trending is being performed. Since the related documentation had not been completed, it was not clear whether corrective actions had been defined i
for the systems discussed in the previous paragraph to address the causes of the observed poor performance. Measurable goals should be established for the systems presently in paragraph (a)(1) of the rule.
i Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST 17
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 e.
The assessment team recommends that Section 7.3 of ADM-17.08 be revised to include additional guidance on (a)(1) dispositioning and goal-setting. Topics which should be addressed include:
(1)
The reasons for initiating the (a)(1) review process (see NUMARC 93 01, 19.4.4).
(2)
Additional guidance on the' establishment of goals and corrective actions (should address the cause of the observed poor performance).
(3)
Guidance regarding the consideration of safety (risk) and industry operating experience when setting goals.
(4)
Guidance on specifying the type and frequency of monitoring which should be applied to established goals.
(5)
Additional definition of responsibilities and the process for reviewing SSCs for potential dispositioning to paragraph (a)(1). System /Com-ponent Engineers should be involved if the smaller size of this group l
st St. Lucie permits. The following general process sequence is suggested:
System / Component Engineer, supervisor, and Maintenance Rule Coordinator confer when any of the conditions in 19.4.4 of NUMARC 93-01 is observed.
System / Component Engineer conducts the cause determina-tion specified in 19.4.4 of NUMARC 93-01.
Maintenance Rule Coordinator makes preliminary disposition.
If left in (a)(2), document basis and refer to Expert Panel for approval. If dispositioned to (a)(1), assign to System /Com-ponent Engineer for establishment of corrective actions and goals (after considering safety / risk and industry operating experience).
Maintenance Rule Coordinator reviews proposed corrective actions and goals.
Expert Panel approves (a)(1) disposition, along with established corrective actions and goals, or recommends appropriate adjustments.
Monitoring against established goals proceeds.
Compiled by PHILIP H. JOHNSON MAINTENANCE HULE TECHNICAL SPECIALIST _ _ _ _ _.
J INDEPENDENT MAINTENE.;E RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 m _.
f.'
The dispositu s og of SSCs back to paragraph (a)(2), after acceptable performance har: been demonstrated, is discussed in 19.4.3 of NUMARC 93 01. This prccess was addressed by Section 7.4 of ADM-17.08. The assessment teara recommends that additional guidance be added to this section of the procedure, as follows:
(1)
The guidance of NUMARC 93-01 19.4.3 should be included.
)
'(2)
The procedure indicates that the periodic assessment should be used as a method of dispositioning SSCs back to paragraph (a)(2) of the rule. While this should not be ruled out as a consequence of the periodic assessment (please see Section 14 below), the process should provide for returning an SSC to (a)(2) when acceptable a
performance has been demonstrated. Keeping an SSC in (a)(1) beyond that time, until the next periodic assessment, appears to
.I unnecessarily burden the program.
]
(3)
The role of System / Component Engineering or other groups should be discussed.
- 11. CONSIDERING PLANT SAFETY WHEN REMOVING SSCs FROM SERVICE
' Section (a)(3) of the Maintenance Rule requires that the overall effect on safety functions be considered when removing equipment from service j
for preventive maintenance or monitoring activities. Section 11.0 of NUMARC 93-01 also addresses this element of the Maintenance Rule. The 1
NRC's draft inspection procedure (page 11,103.02.c.3) provides further guidance on this subject. In addition, NRC representatives have stated that Temporary instruction 2515/126, " Evaluation of On-Line Maintenance," will be used to evaluate performance against Maintenance Rule requirements.
l The essence of this Maintenance Rule requirement is that:
I A current status of Maintenance Rule SSCs must be maintained, and The safety impact of removing an SSC or SSCs from service, taking into account the existing plant configuration and expected changes to it, must be assessed before the additional SSC or SSCs are taken out of service.
Note that this applies regardless of plant mode.
Although Section 7.8 of ADM-17.08 discussed the impact of removing SSCs from service, the coverage was general and did not address the issue in a safety impact of out-of-service equipment led to the following observations:
substantive manner. Overall, the team's review of provisions for assessing the Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST i '
i
- _- - =. ;
l INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST LUCIE PLANT - DECEMBER 1995 4
Strength.- AP 0010460, " Critical Maintenance Management," appears to i
. a.
be providing good control of on-line maintenance activities which fall within its scope. This effort has been well supported by the PSA staff.
i b.
-Though not specifically. reviewed during this assessment, it is. expected i
that the existing shutdown safety program provides appropriate coverage
]
of this issue in Modes 5 and 6.
In discussions during the assessment, a scheduling supervisor stated that c.
he had been tasked with, developing a procedure to apply a safety assessment approach to the existing short notice outage (SNO) process.
This procedure should be defined in a manner which addresses the 4
' assessment of safety impact when SSCs are removed from service in i
Modes 2,3, and 4.
d.
Review of AP 0010460 and discussion with an Operations representative indicated that SSCs are often removed from service in Mode 1 outside i
the scope of the Critical Maintenance Management (CMM) procedure.
]
This includes SSCs made unavailable because of routine surveillance tests or for preventive maintenance which is required more frequently than once per 18 months. The Maintenance Rule would also require that overallimpact on plant safety be evaluated before these SSCs are removed from service.
The equipment out of service (EOOS) process at St. Lucie is applied to all e.
out-of-service SSCs, not just those in a TS action statement. This appears to provide an EOOS status of all equipment In one /o' cat /on, which is an expectation indicated by 103.03.c of the NRC's inspection Procedure 62706, " Maintenance Rule." Operations might also benefit from the utilization of a LAN-based database to track EOOS, as discussed in 19.c above.
f.
The team recommends that the CMM procedure be broadened to include items c. and d. above, or that a separate procedure (s) be provided for them. The team's suggestions regarding expansion of the program are as follows:
(1)
The 12-week schedule which is being implemented provides a good foundation upon which to build the program. Plans were discussed to conduct a PSA review of the 12-week schedule to minimize the significant safety impact which could result from removing certain systems from service concurrently. The team encourages St. Lucie to continue with these plans, and recommends an ultimate objective of having each weekly work schedule reviewed by PSA before it is issued for implementation.
MAINTENANCE RULE TECHNICAL SPECIALIST Compiled by PHILIP H. JOHNSON
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 (2)
The approach taken at Palo Verde, and most commonly within the industry, is to develop a PSA-based matrix which shows the safety impact of removing any two risk-significant SSCs (in the same train) from service at the same time. The matrix could be color coded (green / yellow / red or green / yellow / orange / red) to show the estimated risk achievement worth or increase in core damage frequency which would result.
(3)
The matrix could be configured, with appropriate guidance, such than any SSC not listed on it or indicated otherwise could be considered to be below the threshold of safety concern, and could be removed from service. Procedures could also indicate that equipment removed from service pursuant to the CMM program or a PSA-reviewed schedule need not be further addressed. Combi-nations of SSCs beyond the scope of the matrix (e.g., three or more SSCs out of service concurrently) which are not indicated on a PSA-reviewed schedule should be referred to PSA for review before they are removed from service.
(4)
If a matrix approach is used, the governing procedure should define the thresholds of increased risk.for each of the colored areas (e.g., green for increase in core damage frequency of :s 1 E-6).
The procedure should also specify the level of management control applied to each color or threshold (e.g., Assistant Nuclear Plant Supervisor authorized to approve clearance of S$Cs in green squares on the matrix), with increasing levels of manage. ment 4
approval required for more safety significant configurations.
(5)
Although the CMM procedure states that it does not apply when a Unit enters an LCO because of equipment' failure or a need. to repair degrading equipment. It should be noted, however, that SSCs which are removed from service under such emergent conditions must be considered when evaluating the safety impact of removing additional SSCs from service. Procedures should state that, if such emergent LCO entries result in significant safety impact because redundant or diverse SSCs are already out of service, action should be initiated to restore an acceptable configuration as soon as practicable.
(5)
The plant may find it easiest to address the assessment of safety impact for all Modes 1 - 4 work in the same procedure, using a matrix, with relaxations in the matrix specified as appropriate in Mode 3 and (particularly) Mode 4.
Compiled by PHILIP H. JOHNSON MAWTENANCE RULE TECHNICAL SPECIALIST.
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995
- 12. BALANCING UNAVAILABILITY AND RELIABILITY Section (a)(3) of the Maintenance Rule states that the unavailability of SSCs for monitoring and preventive maintenance should be balanced against the resulting improvement in SSC reliability. Additional guidance is provided in 112.2.4 of NUMARC 93-01. The Rule and NUMARC 93-01 indicate that the balance between unavailability and reliability should be considered as part of the periodic assessment, in addition, it should be considered at other times if unavailability or SSC reliability appears to be challenging established performance criteria or goals.
Section 7.7.4 of ADM-17.08 discussed the balancing of unavailability and reliability in general terms, but was principally a reiteration of methods NUMARC 93-01 states could be used. it does not appear that the balance has yet been addressed in a substantive manner. Of the options recommended by NUMARC 93-01 ('112.2.4) the following are recommended:
(1)
As discussed above under Performance Criteria, ensure that the unavailability criteria for risk-significant systems have a clear PSA basis.
Then, as long as the unavailability criteria are not exceeded, an appropriate balance is being maintained.-
(2)
Whenever an unavailability criterion is exceeded, the balance should be further assessed for that SSC by reviewing SSC reliability and considering other approaches recommended in NUMARC 93-01. '
(3)
As part of the periodic assessment, reassess the overall basis of the unavailability criteria. Reexamine the unavailability values usad, and compare them with the values assumed in the baseline PSA. Also, consider performing a retrospective PSA using the actual unavailability values since the previous assessment, to show the actual overall impact on core damage frequency which resulted from of out-of-service equipment.
- 13. CONSIDERING INDUSTRY OPERATING EXPERIENCE The Maintenance Rule requires that industry operating experience (IOE) be taken into account, where practical, (1) when setting goals and (2) during periodic evaluations of the program's effectiveness. Paragraphs 8.2.1.4 and 8.2.1.5 of NUMARC 93-01 also recommend that IOE be considered during the scoping process, and indicates in 19.3.1 that IDE may be used as an input into risk significance determinations.
The assessment team's observations related to IOE are as follows:
Compiled by PHILIP H. JOHNSON MAINTENANCE Rule TECHNICAL SPECIAUST b
INDEPENDENT. MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995-IOE was used during the scoping process to identify SSC failures which a..
have caused a scram or actuation of a safety-related system at other
- facilities or.have caused a non-safety-related SSC to interfere with the performance of a safety-related SSC! As stated in Section 6 above, the IOE information which was considered during the scoping process'should i
be retained as part of the documentation of initial Maintenance Rule implementation.
~
1 b.
'St. Lucie has an existing program for providing IOE to cognizant members j
of.the staff as it is received. This process should be reviewed to ensure that it interfaces appropriately with the Maintenance Rule, and revised if -
appropriate. The process should provide for the forwarding (and i
consideration) of information which could relate to the performance of I
Maintenance Rule SSCs to appropriate individuals (e.g., Maintenance Rule 1
Coordinator, System / Component Engineers, or others).
l 4
Paragraph (a)(1) of the Maintenance Rule states.that lOE should be taken c.
into account, "where practical," when setting goals. While this is discussed briefly in paragraph 7.3.4.G of ADM-17.08, specific guidance on how IOE should be considered was not provided. Interviews also indicated that it has not been considered to date in the goal-setting process. The team suggests procedural guidance to require, after an SSC 4
is dispositioned to (a)(1), that a specific review of related lOE be 1
conducted as a preliminary to establishing corrective, actions and goals.
As presently indicated in ADM-17.08, this should be documented.
d.
Paragraph (a)(3) of the Maintenance Rule also requires consideration, "where practical," of IOE when conducting periodic assessments. This is discussed further in Section 14 below.
]
- 14. ' PERIODIC ASSESSMENTS OF MAINTENANCE EFFECTIVENESS The Maintenance Rule requires that periodic assessments of the maintenance program's effectiveness be conducted at least every refueling cycle, not to exceed an interval of 24 months. Section 12.0 of NUMARC 93-01 provides 1
[
guidance on the conduct of periodic assessments.
I The periodic assessment i.s discussed in section 7.7 of ADM-17.08.
j a.
However, this section partly reiterates requirements, and does not j
i provide adequate direction on how (or by whom) the periodic assessment 1
is to be conducted. This section should be rewritten to provide more comprehensive direction on the conduct of the assessment (refer to Section 12.2 of NUMARC 93-01 and paragraph 03.03.a of the NRC's Maintenance Rule inspection procedure). The team recommends consideration of the following.
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEM Pursuant to paragraph (a)(3) of the rule, the assessment should (1) review industry operating experience during the cycle to (1) confirm that it was properly considered as it was received and (2) determine whether it indicates changes which should be made to the Maintenance Rule program (see 112.2.2 of NUMARC 93-01).
As discussed in Section 12.2 of NUMARC 93-01, the assessment
.(2) should review the continued appropriateness of performance criteria i
I and goals, the overall performance of SSCs during the period, and the effectiveness of corrective actions.
ADM-17.08 (e.g.,17.3.9 and 7.4.2.E) and discussions during the (3) assessment indicated that the periodic assessment is intended to serve as a method of documenting activities which have occurred since the previous assessment. It is recommended that other methods be provided for ongoing documentation of monitoring and other Maintenance Rule activities (such as disposition of SSCs back to (a)(2)), and that the periodic assessment focus on assessing the e//ectiveness of maintenance and the program during the period.
a The periodic assessment may also consider (per 112.2.4 of (4)
NUMARC 93-01, at the utility's option) the effectiveness of the' program for assessing the safety impact of removing equipment from service for maintenance. Paragraph 03.03.b of the NRC's Maintenance Rule inspection procedure indicates an expectation that the unavailability / reliability balance be considered.
ADM-17.08 presently calls for performing a separate pe'riodic (5) assessment for each unit, shortly after completion of the refueling outage. This may offer disadvantages, such as competing with outage assessments, documentation, and other activities which usually follow a refueling outage. The plant may wish to consider performing the periodic assessment for both units at the same time, scheduled at a time when both units will be on line, The assessment team notes that some activities reviewed by the periodic b.
assessment, such as goals, performance criteria, and SSC performance, are technicalin nature and would appear to be best examined by those (e.g., System / Component Engineering, Maintenance Rule Coordinator) involved in their establishment. On the other hand, an independent assessment team may be better positioned to assess overall program effectiveness and elements such as the consideration of safety impact when removing equipment from service. The team suggests considera-tion of a three step assessment approach to conducting the periodic assessment (approximately every 18 months, for both units concurrently:
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST.
~._-
INDEPENDENT MAINTENANCE RULE ASSESSMENT o ST. LUCIE PLANT - DECEMBER 1995 (1)
System / Component Engineers provide an input to the Maintenance Rule Coordinator (using a one-page form for each system) on the goals, performance criteria, functiona: failures, corrective actions, performance, etc. associated with the, system since the previous 1
assessment.
(2)
Maintenance Rule Coordinator consolidate inputs from System /
Component Engineers, review corrective actions taken, assess SSC and overall Unit performance, assess the continued appropriateness of goals and performance criteria, discuss any necessary changes to scoping or risk significance, and (with assistance from PSA) address the balance between unavailability and reliability. Consider lOE received since the previous assessment. Prepare an assessment input report which discusses these topics, the effectiveness of corrective actions, and any areas where improvement is needed.
(3)
An independent assessment team (three to five members, including an SRO) sponsored by QA or another organization conduct a one-week independent assessment of maintenance effectiveness.
Review the Maintenance Rule Coordinator's input report, and independently assess the principal elements of the Maintenance Rule program, including IOE and the EOOS program.
The third step above (independent assessment team) would not be specifically required by the Maintenance Rule or NUMARC 93-01, but J
would be perceived by regulators as a strength. The first two steps appear to provide the minimum needed to meet requirements.
- 15. DOCUMENTATION OF MAINTENANCE RULE ACTIVITIES Section 13.0 of NUMARC 93-01 provides guidance on documentation of Maintenance Rule activities.
The assessment team concluded that Maintenance Rule documentation needs additional attention, as discussed in earlier sections of this assessment report.
Notable examples are as follows.
The Program Compliance Plan should be issued as an approved j
a.
document. Since the PCP defines the methods used to initially implement the Maintenance Rule, it should be reviewed to ensure that it is appropriately describing the processes used, and that it is consistent with procedure ADM-17.08, implementation of 10 CFR 50.65, the Maintenance Rule."
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST - -
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 b,
Procedure ADM-17.08 should be revised to better define many of the processes which are key to implementation of the rule, as discussed in earlier sections of this report.
c.
Scoping documentation should be completed, including a final version of the Juno Engineering scoping reports and documentation of changes made by tne Working Group. A summary of IOE considered during the j
scoping process should be included. The documentation should be as self-explanatory as possible, requiring little explanation by those who were involved in the scoping process, since they may be unavailable at 4
some future time when it is examined.
j d.
As noted previously, the documentation associated with dispositioning of SSCs to (a)(1) status, corrective actions, and goal-setting is incomplete.
- 16. TRAINING Informal discussions were held with a number of plant staff members to determine the general understanding of the Maintenance Rule. The assess-ment team was informed that training on the Maintenance Rule has been provided, although some time had elapsed.since the training. Based on its discussions, the team concluded that very few individuals know basic I
Information about the Maintenance Rule.
The team recognizes that the Maintenance Rule will be "trinsparent" to many members of the plant staff. However, significant portions of the pl. ant organization (e.g., management, System / Component Engineering, SROs, outage planning individuals, maintenance supervision and planners, and others) should understand the principal elements of the rule. The Maintenance Rule should not be regarded as a discrete program which is owned by the Maintenance Rule Coordinator. Rather, management emphasis is recommended to make it a part of the thought process (e.g., (a)(1)/(a)(2)?
functional failure /MPFF7 unavailability criteria?) which is applied to day-to-day maintenance, engineering, and other plant activities, j
Additional training is recommended, particularly for System / Component Engineers and others who will be working with the program during and after its implementation.
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST i
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995
- 17. INDIVIDUALS CONTACTED DURING THE ASSESSMENT Ji.m Scarola, Plant General Manager j
Wes Bladow, Site Quality Manager i
Joe Marchese, Maintenance Manager i
Jim Voorhes, Quality Assurance Supervisor Mike Snyder, Technical Programs Lead Engineer and Maintenance Rule Coordinator Tony Marvin, Unit 2 Schedule Supervisor Bill Parks, Reactor Engineering Supervisor Catherine Swiatek, System Engineer Brien Vincent, PSA Engineer R. J. Davis, Operations Maintenance Coordinator Elliot Sumner, Unit 1 Watch Engineer Bill Green, Reactor Coolant System Engineer Dave Lowens, QA Engineer l
-Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST l
' INDEPENDENT' MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995
- p. :
ATTACHMENT' A
. SPECIFIC PROCEDURE COMMENTS t
1
'A.-
Program Compliance Plan (PCP).
1 '.
The PCP was reissued on December 11,1995, but-is still considered a i
draft. This document should be issued in approved form and retained as 2
a record of how initial implementation of the Maintenance Rule was accomplished.
2.
. Page 2, " Functions" paragraph: Suggest adding the following after the second sentence: "The Maintenance Rule function for some non-safety 1
related systems may be to not cause a scram or ' unplanned capability
}
loss."
3.
Page 2, " Key Components" paragraph: This paragraph presents good 4
l guidance on the definition of key components and their functions.
4.
Page 3, " RISK DETERMINATION" section: As discussed in Section 7 of j
the assessment report, the plant may wish to reevaluate the risk
[
significance of some systems which are considered marginal. As part of this, it may be appropriate to use only one of the two risk reduction worth (RRW) methods listed (NUMARC 93-01 19.3.1'.1 identifies them as alternative methods), and change the description to indicate that the
[
Working Group should further evaluate the risk significance of systems which are indicated as risk significant by only one of the three risk
[
determination methods.
l 5.
Page 6, third pare. graph: The first sentence states that plant management will determine if an SSC should go into (a)(1). It is recommended that the statement be more specific (e.g., Expert Panel or specified management position).
6.
Page 6, "Industrv Wide Ooeratina Experience in Goal Settina and Monitorina" section: This section appears to be explaining why IOE will not be considered when se,tting goals. Paragraph (a)(1) of the Maintenance Rule specifically requires that IOE be considered, "when practical," in setting goals. The team recommends that this section be rewritten to address how IDE will be considered in goal-setting.
7.
Page 7, last paragraph:- This paragraph explains why performance criteria j
will not be established at the train level. However, discussions during the
- assessment indicated that performance criteria for risk-significant
. Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER systems are applied at the train level. Also, Step 7.5.3 of ADM-17.08 states that parformance criteria should be at the system, train or -
' component level, in addition,;12.4.5 of NUREG-1526, " Lessons Learned
- from Early implementation of The Maintenance Rule at Nine Nuclear Power. Plants," indicates an NRC expectation that monitoring of risk significant and standby SSCs be performed at the train level. The team recommends that this section be revised to support train level performance criteria.
- 8..
Pages 9 and.10, "fa)(3) PER! ODIC EVALUATION" section: The team recommends that these sections be dropped from the PCP (since they are j
not part of initial implementation) end that reference be made to discussion of these topics in ADM-17.08. If these sections remain'in the 1
PCP, they should be rewritten to more completely define the processes
[
i involved (see 111,12, and.14 of this assessment report)'
B.
ADM-17.08, Revision 1, " Implementation of 10 CFR 50.65. The Maintenance f
Rule" in general, it is recommended that this procedure be re--reviewed against the guidance in NUMARC 93-01 and in the light of comments in this assessment l
report. Other specific comments are as follows:
i 1.
Sections 4.3 and 4.4: These sections specify responsibilities of the Working Group and the Expert Panel. Discussions during the assessment indicated that the Working Group will cease to exist, but that the Expert l
Panel will remain as an oversight body for the Maintenance Rule program.
l The use of such a group for program oversight has been considered a strength by the NRC. It is recommended that these sections be revised at an appropriate time to consolidate all necessary responsibilities under j
the Expert Panel. Other Expert Panel responsibilities (e.g., dispositioning l
SSCs to (a)(1) and (a)(2)) should also be included.
r 2.
Section 4.5, f Mt Engineer /Gupervisor responsibilities:
Item 3 assigns responsibility for monitoring SSCs in (a)(1) when j
requested. The team recommends that principal responsibility for l
monitoring all SSCs in (a)(1) and all risk significant and standby i
SSCs in (a)(2) be assigned to System / Component Engineers.
item 4 assigns responsibility for evaluating whether the increase in l
risk from removing equipment from service for preventive mainte-nance is balanced by a gain in reliability. This is difficult, if not impossible, to do quantitatively for an individual out-of-service period, and is principally a matter of judgment. Unavailability can be
' Compiled by PHILIP H.' JOHNSON MAINTENANCE RULE T.ECHNICAL SPECIALIST l.
- e-m
. a.
-.4
INDEPENDENT MAINTENANCE RULE ASSESSMENT ST. LUCIE PLANT - DECEMBER 1995 better balanced against reliability over a longer term, such as be' verifying that risk-based unavailability criteria are being satisfied
-(see 18.a of the assessment report). The team suggests that the wording of item 4 be revised to make it less demanding.
3.
Section 4.6, PSA Engineer responsibilities, item 1: Suggest adding.
" data" after " risk significance," since the Working Group (not the PSA Engineer) determined risk significance for SSCs. The PSA Engineer provided input data which served as a basis for the determinations.
4.
Section 5.1, Definitions: In general,it is recommended th'at these definitions be reviewed for consistency with Appendix B of NUMARC 93-01. Specifically:
Item 3, Functional Failure: Suggest changing "its safety function" to "a Maintenance Rule function." Some functions for Maintenance Rule SSCs (i.e., the reason (s) which required its inclusion in scope) may not be safety related. Also, some SSCs may have more than one Maintenance Rule function.
Item 11, Periodic Evaluation: The team recommends that the last two sentences be dropped; the periodic evaluation should not be the principal means of documenting such activities, which should be appropriately documented as they occur.
5.
Page 12, item 7.3.E: It appears that this step should be discussing (a)(1) and Figure 2, not (a)(2) and Figure 3 (dispositioning to (a)(2) is discussed on page 15).
6.
Page 13, NOTE: This NOTE should be deleted, since it gives an incorrect understanding of adequate monitoring. Monitoring for (a)(1) SSCs should be specifie:1 during the goal-setting process.
7.
Page 14, Step 7.3.9: This step is unclear, and should be revised.
Requiring the periodic assessment to document all monitoring activities would be very cumbersome. Also, monitoring is a continuing activity
+
which is performed for both (a)(1) and (a)(2) SSCs.
i
-8.
Page 14, Step 7.3.10: This step states that goals must be met before an SSC can be returned to (a)(2),'but does not define the period..Please refer to 110.f of the assessment report.
9.
Page 17, Step 7.5.4.B: The team recommends revision of this step to i
require that unavailability begin when the failure occurred, if this can be Compileil bV PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST,
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 t
determined (e.g., the time when a component was incorrectly assem-bled). Otherwise, unavailability should begin at the time of discovery.
- 10. Page 18, IdOTE at bottom of page: The team recommends that this note be deleted, since it is misleading. Functional Failure is a broader term than Maintenance Preventable Functional Failure; Repetitive Functional l
Failure is not.
-11.
The NOTE before Step 7.6.2 is misleading and should be deleted. While l
filling the SIT is may not be a safety function of the HPSI pump, any
~
failure to start upon demand must be considered an inability of the pump.
to perform its design basis safety function -- and therefore a functional failure -- unless the circuit or component which prevented the HPSI pump from being started manually plays no role in the pump's fulfilling its safety function.
- 12. Page 20, Step 7.6.4: This statement is inconsistent with earlier statements that functional failures, not MPFFs will be monitored. Please refer to 18.b of the assessment report.
- 13. Page 21, Step 7.6.7: Recommend adding "for the same cause" at the end of the definition. Also, a period of 36 months is recommended (as more easily monitored) rather than two refueling cycles.
- 14. Page 30, under "Pumos": Suggest adding " Failure to' meet performance specifications (e.g., head / flow)."
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 i
o ATTACHMENT B
(
r SPECIFIC TECHNICAL COMMENTS A sample of Maintenance Rule systems was salected for review to assess effective implementation of St. Lucie's Maintenance. Rule Program. The following systems were selected based on risk significance and system performance:
System 11011 Maintenance Rule Cateaorv Reactor Coolant System (RCS) 1 (a)(1) - Risk Significant Engineered Safeguards 2
(a)(1) - Risk Significant Condensate
.2 (a)(1) - Non-Risk Significant High Pressure Safety injection 2
(a)(2) - Risk Significant.
Condensate 1
(a)(2) - Non-Risk Significant Paragraph (a)(1) of the rule requires licensees to establish goals commensurate 1.
with safety and, where practical, to take into consideration industry operating experience. For the systems reviewed, it appears that the initial (a)(2).
category performan:e criteria are being used as goals for systems that have been placed into (alt?). No objective evidence could be found that specific goals had been set for the (a)(1) systems. Therefore, in the absence of specific goals, consideration of safety and incorporation of industry operating
~
experience into goal setting could not be verified.
The Maintenance Rule requires that the performance or condition of (a)(1)
SSCs be monitored against established goals and that appropriate corrective action be taken when the goals are not met. The Statements of Consideration for 10 CFR 50.65 also indicate that corrective action must be taken when a clearly declining trend in SSC performance or condition indicates that goals could be exceeded before the next monitoring interval. Again, without specific, measurable goals for (a)(1) systems, monitoring and corrective actions could not be assessed for the systems selected.
2.
Specific criteria accounting for functional failures were not found in the Program Compliance Plan (PCP) System Summaries. NUMARC 93-01 states that performance criteria for risk significant SSCs should include SSC unavailability and reliability. As discussed in Section 8 of this assessment i
report, reliability or failure rate cannot be determined without specific performance criteria for functional failures (or MPFFs).
3.
Ol-16-PR/PSL-2, "St. Lucie Action Report (STAR) Program," describes the current program for the identification and resolution of equipment failures.
After reviewing several STAR documents, a review of the STAR program was 1
Compiled by PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST
INDEPENDENT MAINTENANCE RULE ASSESSMENT - ST. LUCIE PLANT - DECEMBER 1995 e
i 1
1' deemed necessary. As a result, it was determined that the STAR process is~
not fully integrated with the Maintenance Rule program. Review of the i
procedure and discussions with the former Maintenance Rule Coordinator confirmed that the STAR process and the Nuclear Plant Work Order system i
2 j
l
- have not yet been revised to interface with the Maintenance Rule; iie., to ensure that all potential failures are captured for subsequent review to determine whether they are functional failures or MPFFs.
i 4.
STAR 950784 was generated on July 18,1995 to document the need to place the RCS into category (a)(1) after it exceeded it performan.ce criteria of 95% plant capacity factor. The STAR indicates that Unit 1 was shut down on February 27,1995 due to pressurizer code safety valve leakage. The five
' ~
months between the shutdown of Unit 1 and the initiation of the STAR was not timely with respect to requirements for goal setting and monitoring. Under i
the " Actions" section of the STAR, the Maintenance Rule Coordinator requested the responsible department to complete an attached form for goal.
setting and monitoring. At the time of this assessment, the form had not been completed; appropriate goal-setting and monitoring has therefore not been accomplished since the system exceeded its performance criteria in February of this year.
5.
Paragraph (a)(1) of 10 CFR 50.65 states that appropriate corrective action i
4 shall be taken when the performance or condition of an SSC does not meet established goals. Corrective actions were adequately addressed in the initial (a)(1)/(a)(2) determinations for the Unit 2 condensate and engineered safeguards systems. However, no documentation of the Expert Pa.nel's i-i determination of corrective actions was found for the subsequent dispositioning of the RCS to paragraph (a)(2). This may indicate a weakness in the overall process for dispositioning systems from (a)(2) to (a)(1).
1 6.
Paragraph (a)(2) of the rule indicates that monitoring as specified in paragraph (a)(1) of the rule is not required where it has been demonstrated that the performance or condition of the SSC is being effectively controlled through the performance of appropriate preventive maintenance such that it remains capable of performing its intended function. The (a)(2) systems reviewed, Unit 2 HPSI and Unit 1 condensate, were performing within their established performance criteria. No repetitive functional failures were identified from the file records reviewed. However, functional failures do not appear to have been reviewed by the Maintenance Rule Coordinator since July 1995. Since the primary objective of reviewing functional failures is to identify repetitive failures, there may potentially be some repetitive failures which have not yet b'een' identified.
Compiled'by' PHILIP H. JOHNSON MAINTENANCE RULE TECHNICAL SPECIALIST.
-