ML20207J248
| ML20207J248 | |
| Person / Time | |
|---|---|
| Issue date: | 06/30/1986 |
| From: | Cwalina G, Bernard Grenier, Jankovich J, Koontz J, Le N, Mclaughlin P, Persinko D Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| NUREG-1212, NUREG-1212-V01, NUREG-1212-V1, SECY-85-129, NUDOCS 8607280194 | |
| Download: ML20207J248 (54) | |
Text
{{#Wiki_filter:l
\
NUREG-1212 l Vol.1 I Status of Maintenance in the U. S. Nuclear Power Industry 1985 Volume 1: Findings and Conclusions U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation ( "scoq,
, ,2 = ::.
1212 R
NOTICE ) Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be available from one of the following sources:
- 1. The NRC Public Document Room,1717 H Street, N.W.
Washington, DC 20555
- 2. The Superintendent of Documents, U.S. Government Printing Office, Post Office Box 37082, Washington, DC 20013-7082
- 3. The National Technical Information Service, Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publications, it is not intended to be exhaustive.
Referenced documents available for inspection and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; NRC Office of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; Licensee Event Reports; vendor reports and correspondence; Commission papers; and applicant and licensee documents and correspondence. The following documents in the NUREG series are available for purchase from the GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of Federal Regulations, and Nuclear Regulatory Commission Issuances. Documents available from the National Technical Information Service include NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission. Documents available from public and special technical libraries include all open literature items, such as books, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, and congressional reports can usually be obtained from these libraries. Documents such as theses, dissertations, foreign reports and translations,and non-NRC conference proceedings are available for purchase from the organization sponsoring the publication cited. Single copies of NRC draf t reports are available free, to the extent of supply, upon written request to the Division of Technical information and Document Control, U.S. Nuclear Regulatory Com-mission, Washington, DC 20555. Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available there for reference use by the public. Codes and standards are usually copyrighted and may be purchased from the originating organization or, if they are American National Standards, from the American National Standards Institute,1430 Broadway, New York, NY 10018. i i I i i
NUREG-1212 Vol.1 Status of Maintenance in the U. S. Nuclear Power Industry 1985 l Volume 1: Findings and Conclusions l Manuscript Completed: May 1986 Date Published: June 1986 l i Division of Human Factors Technology Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555 s~" %,, fe F
I ABSTRACT This report presents the findings, conclusions and recommendations derived from activities performed under Phase I of the NRC Maintenance and Surveillance Program (MSP). Findings are based on trends and patterns derived from operational data compiled by the NRC for the period 1980 through 1985, site surveys conducted at eight plants, and questionnaires administered to NRC Resident Inspectors to characterize nuclear power plant maintenance programs and practices. These activities have shown that plant maintenance programs and practices are highly variable from plant-to-plant and are currently undergoing major changes. While measured plant performance has improved overall since 1980, the maintenance-related centribution to reportable events and challenges to safety systems remains high and is increasing by some measures. The findings of Phase I of the MSP confirmed a number of problems in nuclear power plant maintenance which warrant further NRC and industry attention. iii
EXECUTIVE
SUMMARY
This report presents the major findings, conclusions, and recommendations resulting from Phase I of the NRC Maintenance and Surveillance Program (MSP). The purpose of Phase I, as described in the Maintenance and Surveillance Program Plan (MSPP) (SECY-85-129), was to survey the current status of maintenance in the U.S. commercial nuclear power industry. To accomplish this objective, a program of activities was designed to collect and analyze maintenance data and information and to coordinate NRC and industry initiatives related to maintenance. A summary of major findings, conclusions i and recommendations arising from these activities follows. Trends and Patterns in Maintenance Actual plant performance data, based upon information reported to NRC, have been found useful to monitor overall industry maintenance performance. Trends and patterns in plant maintenance performance between 1980 and 1985
; were studied using 31 measures of plant performance. Although measures of I plant performance from 1980 to 1985 show some improvement in overall plant performance, there are still a significant number of reliability problems i
that warrant attention by the NRC. Overall plant system / component reliability has improved since 1980 as shown by a decline in average numbers of forced outages experienced by the nuclear industry. However, component failure forced outages continue to comprise over 60% of forced outages, indicating component reliability problems which could be attributable in part to maintenance. Safety system reliability for all the plants has not significantly changed since 1981, as determined from annual average numbers of LERs. A significant increase in the percentage of maintenance-related LERs from 1984 to 1985 was observed. In fact, 48% of the total number of 1985 LERs are maintenance related. Challenges to safety systems, as measured by average total numbers of scrams, have declined between 1980 and 1985. However, scrams due to maintenance The percentage and of testing did not engineered safety significantly(decrease between features ESF) actuations due 1980 to and 1985. maintenance, surveillance, and component failure was about 75% of total ESF actuations in 1984 and 1985, indicating that a significant challenge to plant reliability and safety can be attributed to maintenance-related activities. t ^ Industry plant average radiological exposure data show no statistically significant improvement trend since 1980 according to NUREG-0713, Occupational Radiation Exposure at Commercial Nuclear Power Reactors. Radiological personnel exposures due to maintenance represented 46% of the total exposures in 1984. Quality programs and surveillance SALP ratings improved between 1980 and 1985 while operations SALP and maintenance SALP ratings have not changed significantly. i
- y
Taken together, these measured trends in plant performance show overall improvement since 1980. However, it is important to note that a significant portion of measurable reliability problems in terms of forced outages, LERs, and ESF actuations are attributable to maintenance, surveillance or component failures and have not shown significant improvement. While overall trends are positive, there is a need to improve performance of maintenance and surveillance in order to reduce challenges to safety systems and improve safety systems availability and reliability. Current Status of Maintenance As determined from 8 plant site visits and questionnaire responses from 66 Resident Inspectors, there is wide variability in plant maintenance programs and practices. Industry maintenance programs are undergoing many changes; reorganizations are taking place and major new programs are being initiated. Maintenance programs and practices at some plants appear to differ considerably from the tendencies of the rest of the industry. Maintenance is now a separate organization at most plants, responsible for planning and scheduling their own electrical, mechanical, and I&C maintenance work. While most plants are found to have some form of a preventive maintenance program, one-fourth of the programs appear to be minimal. The maintenance staff at the majority (70%) of the plants spend one-fourth or less of their time on preventive maintenance tasks. Scope and coverage of preventive maintenance programs in the U.S. nuclear industry vary widely but appear to be less extensive than in some other countries (e.g., Japan) and other industries (e.g., aviation industry). Most of the industry is collecting maintenance performance data and reporting it to INP0. However, many plants also report having difficulty correlating the performance measures with specific maintenance practices. Most plants are not well-equipped to do extensive formal equipment trending and analysis. Use of automated systems for data management, trending, and analysis is a changing aspect of most plant maintenance programs. According to the Resident Inspector questionnaire responses, adequacy of workshop space and equipment laydown areas is a problem at some plants and spare parts availability appears to be a problem at all but a few plants. The majority of plants have not yet received INP0 accreditation of their electrical, mechanical and I&C training programs. Only about one third of the plants appear to have formal maintenance personnel training covering plant modifications or requalification training programs. Work order preparation and spare parts inventory control systems are completely manual systems in some plants while other plants are moving toward computerized systems. There is a great variability in the quantities of work orders backlogged among plants. Site surveys indicated that the work order backlog ranged from 2-4 weeks up to 6 months. Most plants have a decreasing or constant trend in work order backlog. vi
Quality control of maintenance work activities varies widely. While most plants limit quality control activities to only safety-related work, a few plants do use comprehensive quality control of both safety and nonsafety-related work. Implications of NRC and Industry Maintenance Initiatives NRC and industry maintenance initiatives were not well coordinated before Phase I of the MSPP. Both NRC and industry have made significant contributions to a better understanding of maintenance issues during the last year. NRC and industry initiatives related to equipment aging and performance are producing equipment-specific data and techniques. These programs provide important data for effecting improvements in equipment reliability. Compatibility and usability of the results of component performance studies could be improved by focusing on common goals for reliability.
- Most of the NRC and industry studies of factors affecting human error and personnel performance in maintenance focus on individual causative factors or events resulting from errors. The complex relationships among factors influencing human performance are still not well known or extensively investigated. The well-documented high contribution of human error to reportable events and to risk warrants identification of means to improve human reliability. Improvement strategies currently focus on personnel training as a means to improve human reliability. A study of maintenance procedures documentation, use, and control revealed wide variability in the quality of maintenance procedures in the industry with the potential to impact safe plant operations.
The role of management and organization in nuclear plant maintenance is less well defined in the U.S. nuclear power industry than in some other countries (e.g., Japan). Although the role of management is repeatedly emphasized in event follow-up investigations, the current regulation of maintenance does not provide a sufficient basis for defining maintenance and reliability goals for use by management. The NRC has been activefy participating in and encouraging development and use of industry standards for maintenance. It is still unclear what integration problems may arise from implementation of diverse standards addressing various equipment and activities. There is no widely implemented programmatic guidance for maintenance. The diverse industry-developed standards and guidance need integration to ensure reliable maintenance. The effect of INP0 guidelines (INP0 85-038) for maintenance programs published in October 1985 is not yet visible. The maintenance and operations interface has not been extensively addressed by NRC except in the context of human errors during maintenance. Identification of improved surveillance and test methods may allow for better vii
i coordination of maintenance and surveillance testing activities. NRC activities related to quality of operations and maintenance are currently focused on distinctions between classifications of equipment and activities rather than on performance goals. Conclusions The trends and patterns analysis of industry maintenance performance clearly indicates that although overall industry reliability appears to be improving (as shown by declining forced outage and scram rates) maintenance continues to be a significant contributor to industry reliability problems (as much as 75% by some measures). This contribution of maintenance to reliability problems indicates that some maintenance programs and practices are not effective. Results of Phase I of the maintenance and surveillance program confirm that problems identified in the MSPP are indeed evident throughout the nuclear industry. That is, that: needed maintenance is not being accomplished or is not performed effectively; a high percentage of failures result from improper performance of maintenance; the maintenance and operations interface is inadequate; maintenance-related activities are responsible for a significant number of challenges to safety systems; and a major portion of occupational radiation exposure to personnel results from maintenance. Phase I findings also confirm that there are wide variations in maintenance practices among utilities and that many licensees are implementing changes in their maintenance programs to address identified problems. Further, industry has established a variety of programs aimed at self-improvement. However, NUMARC and INP0 initiatives which address industry maintenance problems are still in their infancy. While many changes are taking place in industry maintenance programs and practices, improvement efforts do not appear to be 4 well integrated or effectively implemented in some cases. While industry improv!ments in overall performance are evidenced by measurable trends, maintenance performance is not measurably improved. The contribution of any specific causal factor (e.g., INP0 or NRC initiatives) to performance trends observed cannot be inferred at this time. Therefore, improved objective and reliable measures of plant performance are needed. viii _ - . - . ..- - - . _ . . ~ . -.
To solve these problems, several NRC and industry actions should be taken: Improve development and use of measures of plant performance and safety; Determine the contribution of human performance in maintenance to plant reliability and identify the safety impact of alternative strategies for improving human reliability; Define maintenance needs, including the role of preventive maintenance, for achieving acceptable levels of reliabilite and safety; Define goals for nuclear power plant reliability and safety that foster
; management involvement in ensuring effective operations and maintenance, including the use of performance measures and incentives; Develop and implement performance-oriented maintenance criteria and standards; Identify approaches to maintenance which enhance the interface between maintenance and operations and ensure reliability and safety; and Coordinate NRC and industry initiatives related to maintenance.
Resolution of the issues identified by the MSPP and Phase I initiatives will be addressed during Phase II of the MSPP. i ix
TABLE OF CONTENTS Page ABSTRACT iii EXECUTIVE
SUMMARY
v
1.0 INTRODUCTION
1 1.1 Definition of Maintenance 1 1.2 MSPP Phase I Objectives 1 1.3 MSPP Phase I Activities 1 1.4 NRC and Industry Maintenance Activities 2 2.0 FINDINGS 3 2.1 Trends and Patterns in Plant and Maintenance Performance 4 in the U.S. Nuclear Power Industry: 1980-1985 2.2 Survey of NPP Maintenance Practices 10 2.3 Other NRC Programs and Industry Activities 13 Related to NPP Maintenance
3.0 CONCLUSION
S 16 4.0 RECOMMENDATIONS 17 APPENDIX A Selected NRC Programs and Industry Activities Related to Nuclear Power Plant Maintenance APPENDIX B Current NRC Regulation of Maintenance in Commercial Nuclear Power Plants xi
1.0 INTRODUCTION
On January 11, 1985, Phase I of the NRC Maintenance and Surveillance Program Plan (MSPP) (SECY-85-129) was approved for implementation by the Executive Director for Operations (ED0). The purpose of this report is to describe the status of maintenance programs in the nuclear industry as identified by Phase I activities and to identify goals for improving commercial nuclear power plant (NPP) maintenance consistent with the NRC Policy and Planning Guidance (NUREG-0885, Issue 5). The scope of this report includes a summary of the Phase I MSP activities, findings, issues warranting further attention, and recommendations for improving NPP maintenance among the nuclear utilities. 1.1 Definition of Maintenance Maintenance is defined as a process with the objective of preserving the reliability and safety of nuclear power plant structures, systems, and components or restering that reliability when it is degraded. As defined in the MSPP and for the purposes of this report, NPP maintenance and surveillance includes: (a) diagnostic or periodic testing, surveillance and inspection to determine the condition of structures, systems, and components, (b) preventive and corrective actions such as repair, replacement, lubrication, adjustments, or overhaul; and (c) proper equipment isolation, restoration to service, and post-maintenance testing to assure' adequacy of corrective action. Maintenance and surveillance are performed during all modes of NPP operation by plant staff, vendors, or contractors. The scope of this definition of maintenance and surveillance is consistent with the International Atomic Energy Agency's Safety Guide on Maintenance of Nuclear Power Plants (No. 50-SG-07, IAEA, 1982). 1.2 MSPP Phase I Objectives Phase I of the MSPP was designed to survey current maintenance practices in the U.S. nuclear utility industry and evaluate their effectiveness. The projects performed as part of Phase I of the MSPP were also intended to address the technical and regulatory issues of NPP maintenance. A number of activities were initiated by NRC and the industry to achieve these objectives. 1.3 MSPP Survey Activities In partial fulfillment of these objectives, survey activities performed as part of Phase I of the MSPP included: t Collection, categorization and assessment of available data to describe the status of maintenance in the U.S. nuclear power industry: I l 1 I information pertinent to current NPP maintenance programs, practices and performance was collected and assessed to serve as a baseline reference for evaluating industry initiatives in maintenance. Review of the Salem preventive maintenance program: the Salem maintenance programs and practices prior to the 1983 ATWS event were compared with post-1983 practices to determine the differences and the perceived benefits of changes in Salem's maintenance program. Development and administration of a questionnaire to obtain plant-specific maintenance program data from NRC Resident Inspectors: maintenance program descriptive data obtained from Resident Inspectors
. who completed the questionnaire were compiled and analyzed.
Development and administration of a maintenance review protocol to obtain plant-specific maintenance program data during on-site NRC surveys: a protocol was developed and used for collecting comprehensive descriptive data on plant maintenance organization and administration, facilities and equipment, procedures, personnel, and work control. Collection and assessment of descriptive data from selected nuclear power plants: Eight site surveys were conducted by NRC teams to collect information on plant-specific maintenance programs and practices. The sites surveyed included Kewaunee, Millstone, Davis-Besse, Rancho Seco, Arkansas Nuclear One, Salem, Turkey Point, and Brunswick. The results of these activities are more fully described in Volume 2 of NUREG-1212. Findings, conclusions and recommendations derived from MSPP Phase I activities and their implications are summarized in Sections 2.0, 3.0 and 4.0 of this report. 1.4 NRC and Industry Maintenance Activities In addition to the survey projects performed under Phase I, the MSPP monitored and coordinated activities with the following NRC programs and industry initiatives related to NPP maintenance: NRC MSPP projects: Analysis of Japanese and U.S. maintenance programs;
-- Human factors in in-service inspection; and Human error in events involving wrong unit or wrong train.
Related NRC projects and activities:
-- Quality Assurance (QA) program plan;
-- Systems important to safety; j -- Comprehensive re-evaluation of standard technical specifications; i
-- Nuclear plant aging research;
-- Environmental equipment qualification;
-- Reliability research;
-- Training and qualification of personnel;
-- Safety implications of control systems (Unresolved Safety Issue (USI) A-49);
-- Interfacing systems loss of coolant accident (LOCA) at BWRs (Generic Issue (GI) 105); and
-- Radiation protection plans (GI III.D.3.1).
Related industry activities sponsored by:
-- Standards groups;
-- Institute for Nuclear Power Operations;
-- Nuclear Utilities Management and Resources Committee (NUMARC);
-- Electric Power Research Institute, and
-- Vendor groups.
The most significant maintenance-related results and implications of these initiatives curing Phase I of the MSPP are summarized in Section 2.3. Further discussion of these activities is provided in Appendix A. 2.0 FINDINGS This section describes the current status of maintenance in the industry based on MSPP research activities. MSPP Phase I activities have identified maintenance program characteristics through in-depth site surveys of maintenance activities at selected plants and through the administration of a questionnaire to NRC Resident Inspectors to obtain plant-specific maintenance information. Measures of maintenance performance derived from nuclear power plant operational data have also been collected and analyzed. Identification of important maintenance program characteristics and functions and their relationship to operational safety has not previously been achieved by the NRC or the industry. This approach is useful for establishing a baseline of industry data against which to measure improvements and for providing information needed to focus NRC resources and action where problems are identified.
2.1 Trends and Patterns in Plant and Maintenance Performance in the U.S. Nuclear Power Industry: 1980-1985 Thirty-one measures of maintenance and nuclear plant performance were obtained ' rom published NRC data for operating plants for the years 1980 J through 1985 for an analysis of trends in the nuclear power industry : (NUREG/CR-4611, Trends and Patterns in Maintenance Performance in the U.S. I Nuclear Power Industry: 1980-85, June 1986). The 31 measures developed for this analysis were organized into the following 5 categories: (1) overall system / component reliaoility, (2) overall safety system reliability; (3) challenges to safety systems; (4) radiological exposure; and (5) regulatory assessment. A summary listing of the measures associated with each category is presented in Table 2.1, along with a breakdown of the statistically significant trends and patterns that occurred for a minimum of 3 years out of the 5 years examined. For the age group analysis, it is frequently the case that there are too few plants in the youngest age group for this group to be included in the analysis of 1980 and 1981 data. A description of the information presented in Table 2.1 and a review of the i major important trends and patterns identified in the detailed analysis report are provided in the following sections. Overall System / Component Reliability. Between 1980 and 1985, several distinct trends and patterns were identified. First, the nuclear industry experienced significant declines in the number of forced outages and forced outages due to component failures, while at the same time decreasing the average gross heat rate (i.e., increasing the average thermal efficiency). No significant changes were noted for the industry for availability and forced outage rate. Second, the older nuclear power plants had fewer forced outages than the newer plants, but took longer to return to service following outages. Finally, smaller plants had significantly longer periods between component failure forced outages (i.e. a greater mean-time-between failure). The percentage of forced outage time due to component failures, a measure of overall component reliability related to maintenance, has been greater than 59% since 1980 and shows a fairly stable trend (see Figure 2.1). This percentage may be used as an index of the overall quality of plant operation and maintenance and meets INP0's criteria for a good performance indicator. Since maintenance is responsible for assuring the reliability of systems and components, this measure reflects the quality of maintenance being performed. Overall Safety System Reliability. With respect to licensee event reports (LERs), the major findings were that older plants (plants that went critical before 1975) had the fewest number of LERs and maintenance-related LERs. Also, smaller plants (less than 600 MWe) were also found to have significantly fewer LERs and maintenance-related LERs than any other plant size category. The TAiLE .'. Lur of Trean ! Patterns in Maleteccce Ferforsance
.. T :e
~,nc. i r ri p.c s..
- r;Ag5Eg i..._..._....;... _..__.............__ ...___. _____.........___.....__..'
' C; FACI!Y
! 1 8 -1795 ' ELP vs FWF i idE !
CHALLE GEi TO ~3 ET< iiiiEM5 : U:e tf E:rai: t. m y , l Cre 1C 0 cebEst
- 600 G e-Fewest :
hnter cf icrus 10 krs. :: y ' --- Pee I W -Fe.est O.).Se-Fenest l Cr:ticsi :: l l 8 i i
, , I, fiuster of Ecrus Le to
! .606 Ne-Fesest i s y ---
! Fre 1970-Femest :
Cce m ent FS11.re .' l l i , , ,
- o M :sr of Cte:: ent rai kre 'l y ; ---
! Dre 1970 resest G00 NedeWest
; Screr6100 Hrs Criti;al ; l I i 9 00:er of 5:rass [ee ta !! ---(t) l
, Mainter,er:e aM T stir; E 'I , l ! i
, ,, , , i
; Nrter :f Scre.s he tt '; --- W !
l --- --- I
- Mainterance vd isstirq l'00 Hrs .; '
l i Critita! ,
. , , e , ,, . e. .
t f;rt e r cf EST A: tat 15ns !! kW ; ---
' Fost !?S0-t'est 11000 M'Je-Mest
, , tketer :f E5F Atteati;a.s he tt l da :' l ---
l F0st 1C30-M05t ! }100) Ne-Most
! Malct./5crv..'Ces;9 rent Fat bre :: ;
;__....____......._ ... ..............______..____....____ .....__......__.._____.___............__... ...___ ___.l FE%LlICP A53E55?'ENT i e Cteretitr.s 5'LP Gating : ---
l l ---
! GO) Ne-Lowes: ; l- ! (4 brs.)
I I , ti i, , , Deality Frogrees 5:.LP Fating l Ytd) --- ; --- 0 Eurveillance 5;;F F atia: ; ', ' l YMi , l e I 1 l c Mainteur:e 5;LF Fatir; --- --- --- --- iC' 5e:2erce Cc n c LER data available tegian n; lost (th Trend ar.alysis fu 19clC25 due te ieptrtir; change ir IC39 it): 'k coastster.t trend ne' tre5 fear perid (d:: Fe'lects 1*provin; 5AS ratirg
---: la sigrif: car:t difference reted ft- ure than 5 of 5 yers d: 11 creasing trer.d Y: Decreasin: t re r.d TABLE 2.1 Summary of Trends & Patterrs in Maintenar.ce Performance
! ll TFENES i FATTEF!!S :
- MEASUFES 1;-----------l-----------------------------------------------------------l 1 ll 1950-1935 ! EWR vs FWR
- AGE ! CAFACITt !
OVEPALL SYSTEM!COMPO!!ENT PELIABILIT) ll l l ! l l o % Unit Availability ll --- l --- l l ll ! ! ! l o Gross Heat Pate ll y 1 ---
! Newest Plants have !
l ll l l hignest efficienty l l 1 !! ! ! l l l o Natter of Ferced Outa;es t: y 1 l Newest-Mest l (600 MWe-Fewest l l ll l l Oldest-Fewest ! l l ll l l l ! ! o Adjusted % Forced Outage Tise : --- l --- --- l l l !! l l l l ! o Nuater of Component Failure ll y EWR-Fewer : --- l l Ferced Detages !! l i l l l ll l l ! ! l o % Forced Outage Time :: --- 1 l --- l --- t Cosponent Failures ll l i l ll l l l l o Mean Tise Between Corpotent l! g l l --- l (600 MWe-Higbest l l Failure Forced Outages !: ! l l l l l; l ; l o Mean Tise to Feturn te Service ll --- l --- l Oldest-Longest : --- l ! After Ceactnent Forced Outages : : : M of 5 yrs.) l l l OVEFALL EAFETY SYSTEM PEllA91LTV(a) !! ! l l :
- o Total f;tster of LERs !! ---(t) l PWR-Fewer ; Oldest-Fewer 11000 F.We-Highest !
l :: l : 00 MWe-Femest ! l 1: : : : (3 yrs! l l c !!ur.ber ef Component Failure LEPs ll y(b) FWR-Fewer : --- l l l ll l l l ;
; o % Cetponent Failure LEPs to ll y l l
l l Total LEPs ll l l l : l lt ! ! ! l o Nester of Maintenance- ll gt) l PWR-Fener l Pre 1975-Fewer l (600 N e-Fewer ; l Pelated LERs l! l l l >1000 MWe-h st l l ll l l l l l e % Main'.enance Felated LEFs ll g l --- 1 --- l l ta Total LEPs !! l l l 1 FADIOLOGICAL EIF0SURE !! l l ! l l o Total ManPea Exposure : --- 1 BWR-Higter l --- l 600-800 MWe-Highest l l ll l l l l c ManRee Exposure Due te ll --- l BWR-Higher l --- l 600-800 MWe-Highest l l Maintenance ll l l l l percentage of maintenance related LERs since 1980 has been greater than 21% with recent increases to 39% in 1984 and 48% in 1985 (see Figure 2.2). This measure represents the contribution of maintenance to plant safety performance. It should be noted that even though the total number of LERs showed a decrease in 1984 due to changes in reporting requirements, the percentage of LERs attributed to maintenance increased. Both the number of maintenance-related LERs and the percentage of maintenance-related LERs increased significantly from 1984 to 1985, while the number of component failure LERs declined by 10% in 1985. Challenges to Safety Systems. There has been a significant decline since 1980 in the average number of challenges to safety systems as represented by the decrease in four of the scram-based measures. Plants that went critical before 1970 and plants with 600 MWe or less capacity were found to have significantly fewer challenges to safety systems than either the newer or larger plants. The percentage of ESF actuations due to maintenance, surveillance, and component failure represents challenges to engineered safety features of the plant. Almost 75% of all ESF actuations in 1984 and 1985 were attributable to maintenance, surveillance and component failures, representing errors in maintenance and surveillance and the lack of maintenance effectiveness ensuring component reliability. Radiological Exposure. Personnel exposure levels (total and maintenance-related) remained stable between 1980 and 1984 according to NUREG-0713,
" Occupational Radiation Exposure at Commercial Nuclear Power Reactors." The only significant and consistent pattern noted was that BWRs have approximately double the mean exposure levels as those experienced by PWRs.
Maintenance-related radiological exposures represent about 46% of total exposures. Regulatory Assessment. Between 1980 and 1985 the nuclear industry experienced significantly decreasing trends on several regulatory assessment measures. Specifically, the quality programs and surveillance SALP ratings improved. Overall Trends and Patterns. Across the five categories of maintenance measures, several important trends and patterns emerged. One of the most important concerns the overall industry trends. Over the 6 years examined, the industry experienced significant decreases in all five categories. For 10 of 25 measures on which trend data were available, plant performance changed in the direction of what conventional judgment would consider improved performance. Some measures of safety system reliability (i.e., number of maintenance-related LERs), and challenges to safety systems (i.e., ESF actuations) showed significant increases. Reactor type differences are not generally pronounced nor consistent, except for PWRs having lower exposure rates and somewhat better performance on three of the regulatory assessment measures than BWRs. There are significant age patterns for three nf the five maintenance performance categories. Within these three categories, the most pronounced tendency is for older plants (plants critical before 1975) to have better scores on the measures. Of the 11 measures where significant age patterns were identified, these older plants performed better on 9. A visual inspection of the trends for a number of the measures, I FIGURE 2.1 PERCENT OF TOTAL FORCED OUTAGE TIME DUE TO COMPONENT FAILURE BY PLANT ELECTRICAL CAPACITY
- ANNUAL MEAN, 1980 - 1985 Plant Capacity 80- .
(Net MWe): 70- .5 Y ,k (j[ ! ej 60- ;[ G 599 or Less fI shg IIB 600-799 50-- . g , {. O 800-999 Pe ent 40- .{ ,,
? 2 E 1000er+
30- i I ; -{Ij ? l j g e- Industry Mean 4* 20- E $ 10- i g
; c; ,
4 1980 1981 1982 1983 1984 1985 Year
- Plant Electrical Capacity - Maximum Dependable Capacity (Net MWe):
Mean Percent of Total Forced Outage Time Due to Component Failure by Plant Electrical Capacity (Net MWe) 1980 1981 1982 1983 1984 1985 MWe 599 or Less 68.2 58.6 61.6 44.7 52.2 77.4 600 - 799 68.6 73.8 63.0 52.2 63.4 61.8 800 - 999 74.2 79.4 75.5 78.4 73.3 73.3 1000 or + 62.7 63.1 58.8 59.4 64.9 64.6 Industry Mean 69.2 70.1 65.6 59.9 64.4 68.4 Ind. Std. Dev. 34.3 32.0 32.6 37.5 34.6 31.6 l
FIGURE 2.2 PERCENT OF MAINTENANCE RELATED LERs TO TOTAL LERs BY PLANT ELECTRICAL CAPACITY
- ANNUAL MEAN, 1981 - 1985 60- Plant Capacity (Net MWe):
50-EE 599 or Less 0- e?y E 600799 e" p cent 30 /
- O 800-999 7 f*
"o- T
- s E 1000 or+
20- ?i 4 i a ? ss: : No !
~' * **"
10- LER S A i i 3 $ 5 - I is Data g sj j gf !. g 0 1980 1981 1982 1983 1984 1985 Year
- Plant Electical Capacity - Maximum Dependable Capacity (Net NWe)
Note: Definition of reportable events (LERs) changed in 1984. Mean Percent of Maintenance Related LERs to Total LERs by Plant Electrical Capacity (Net MWe) 1981 1982 1983 1984 1985 MWe 599 or Less 25.9 23.9 24.7 36.6 50.9 600-799 20.9 25.2 20.3 38.5 46.6 800-999 24.0 21.0 26.4 42.6 48.8 1000 or + 15.4 22.3 23.1 37.3 48.9 Industry Mean 9.1.8 23.0 23.6 39.0 48.7 Ind. Std. Dev. 8.8 8.1 10.8 15.2 17.3 however, indicates that between 1980 and 1985 the average scores for the newer plants have been converging toward the older plant scores. The only category where no significant trend was observed was in radiological exposure. Significant plant capacity differences were noted across all five of the performance categories. Significant plant capacity differences were identified on fourteen of the measures, with the general pattern of 600 MWe or smaller plants performing better than some or all of the other plant size categories. 2.2 Survey of NPP Maintenance Practices The following findings, derived from NRC site surveys and questionnaires administered to all NRC Resident Inspectors, characterize the current status of maintenance at operating plants (see Volume 2 of NUREG-1212). The findings presented are based primarily on the responses to the Resident Inspector ; questionnaire and supplemented, as indicated, by data obtained during the ' on-site surveys. The majority of U.S. plants service equipment _in accordance with vendor technical recommendations, procure appropriate spare parts and tools for required equipment repairs, hire qualified personnel to perform maintenance, control work-in-progress, and perform quality checks on completed jobs. There are some plants in the industry whose maintenance programs are especially well-structured and administered. In addition, there are aspects'of some plant maintenance programs that are excellent and good examples for the entire industry, and other aspects that are weak. Finally, there appears to be a small residual of plants with serious deficiencies throughout their maintenance programs. Examples of variability in maintenance programs include: Maintenance is a separate department at most of the plants, but reports through operations at 20% of the plants. I&C is typically under the maintenance department, but at one-third of the plants, I&C is located in the operations or technical services department. Planning and scheduling is typically done under the maintenance department, but at some plants planning and scheduling is done outside the maintenance department. In the opinion of the Resident Inspectors, about one-quarter of the plants were judged to have an extensive preventive maintenance program, about one-half have an adequate preventive maintenance program, and about one-quarter have a minimal preventive maintenance program. A majority (f3%) of the plants systematically monitor some type of maintenance performance measures, while the rest do not. Most plants, however, have trouble correlating their performance measures with specific aspects of their maintenance program. While 70% of the plants reportedly use a system for trending and analysis of failures, only two of the sites visited evidenced a workable system for trending and analysis of root cause. In general, a majority of the plants have adequate workshop space even though some work is done outside the workshop. However, some plants (9%) have inade,uate work space as evidenced by the necessity to often perform work cutside of shops due to lack of space. Laydown space and equipment accessibility generally present some problems at all plants, but not enough to affect maintenance work adversely. However, one-fourth of the plants have equipment laydown or worker accessibility problems. Spare parts availability is, to some extent, a problem at all but a few plants. Approximately 60% of the plants reportedly operated in a degraded mode in 1985 due to unavailability of spare parts. There is a wide range (from several dozen to several hundred) in the numbers of maintenance personnel across U.S. nuclear power plants. This variability occurs not only across different unit sizes (e.g. one-unit versus two-unit plants) but also exists among plants of the same unit size. Shift coverage, or the numbers of maintenance staff working on each of the three shifts (day, swing, night) also varies across plants. Approximately half of U.S. nuclear power plants have no scheduled maintenance coverage - on the swing and night shifts during normal operations. While most plants have yearly turnover rates in the 2%-10% range, one plant indicated an 18 month turnover rate of 90% for I&C technicians. There is also some variation in the existence of requalification and plant modifications training for maintenance personnel in U.S. nuclear power plants. Only about one-third of the plants have a formal program in these areas. The site surveys indicated that some plants use career development programs such as career plan interviews, succession planning, systems to match people with available jobs, and assessment centers to identify and < develop maintenance department managerial potential. Other site survey plants indicated that they have no career development programs or formal avenues of advancement. Work order generation, accumulation of required work documentation, and , maintenance follow-up documentation practices ranged from completely ; manual systems with no computer interface to completely automated systems where even the work order was computer generated.
Warehouse inventory control ranged from weekly manual inventory checks to fully computerized inventory control. The time spent on the actual maintenance work by craft workers depended greatly on the type of work control system. In the case where work packaging was done before work was started, the craft workers were able to spend approximately three-fourths of their time doing actual maintenance work at the job site. In the case where craft workers had to do some or all of the work packaging, the craft workers were able to spend only about one-half of their time doing actual maintenance work at the job site. QC interaction during preparation, performance, and follow-up of maintenance tasks ranged from review of only safety-related job packages and use of hold points only on safety-related work, to a review of all work orders, hold points and random checks on safety-and nonsafety-related work, and a systematic review of all safety-related work packages and spot checks on nonsafety-related work packages. Other major findings based on the Resident Inspectors' responses to the questionnaire on the current status of maintenance programs at U.S. nuclear power plants include the following: There are no formal predictive maintenance programs at one-fourth of the ; sites whose Resident Inspector responded to the questionnaire and an additional half of the plants have no written program guidelines for predictive maintenance. Only one-third of the nuclear power plants have maintainability improvement programs. The maintenance staff at a majority (70%) of the plants spend one-fourth or less of their time on preventive maintenance tasks. l The common addition of " informal labels" on plant equipment indicates f that a large number (40%) of plants need to improve the labeling programs at their plants. 4. Although nearly half (45%) of the plants do not review their maintenance procedures on a regular basis, most plants were reported to have an adequate systematic process for developing, reviewing, and updating procedures.
.Most plants (77%) require separate verification of maintenance tasks only on safety-related jobs.
One-fourth of the plants were reported to have a high backlog of , s maintenance work in the opinion of Resident Inspectors. I L
2.3 Other NRC Programs and Industry Activities Related to NPP Maintenance NRC programs and industry activities related to NPP maintenance were monitored during Phase I of the MSPP. Major findings are discussed in the context of the technical issues identified in the MSPP. 2.3.1 Human error in the performance of maintenance: NRC and industry sponsored studies have focused on several problems associated with human error in the performance of maintenance, such as contributors to wrong unit / wrong train events and personnel training. Personnel-related problems are reported to be a substantial secondary cause of reactor trips above 15% power, accounting for 28%. Personnel error is the cause for 14% of critical path delays as determined from a study of 20 BWR outages. A study of human error in 35 wrong unit and wrong train events at 10 sites has identified labelling of equipment to be among the most common contributing factors. Taken together, inadequacies in labeling, training and procedures were the primary contributors to human error in 60% of the wrong unit / wrong train events investigated. Although most wrong unit / wrong train events involved several contributors, any one of the 10 cited causes of events was judged sufficient by itself to cause an event. Technician performance is the leading source of variability in reliability of ultrasonic testing / inservice inspection. (One of the few predictive maintenance techniques used at NPPs.) Industry implementation of performtnce-based maintenance training and qualification programs is still in its infancy, but shows promise of improving personne1 training. Maintenance jobs are major contributors to occupational exposures for NPP personnel. 2.3.2 Indicators of maintenance effectiveness: Several NRC and industry initiatives to collect objective and reliable indicators of NPP maintenance performance and effectiveness have been ongoing during Phase I, such as, INF0/NUMARC' Indicators, Systematic Assessment of License Performance, Performance Appraisal Team (PAT) . inspections, Safety System Functional Inspections, Licensee Event Reporting System, In-Plant Reliability Data, and probabilistic risk assessment studies. 7 Although NRC and the industry are collecting an extensive set of data, current data sources do not provide the types of data that permit easy identification of strategies to improve maintenance performance at the plant level. The industry and the NRC are using performance measures to characterize I maintenance and NRC is collecting operational data to monitor safety. 2.3.3 The role of preventive maintenance in counteracting aging and service wear effects: The Nuclear Plant Aging Research (NPAR) sponsored by the NRC stresses the importance of inspections and surveillance testing to detect the onset of ; equipment aging as well as component specific methods to determine the role of maintenance in preventing or correcting equipment failures. One of the primary goals of plant maintenance activities should be to ensure the operability of equipment important to safety under all anticipated environmental conditions, especially during accidents. NRC audits have shown that the industry has not yet fully established programs to integrate equipment qualification requirements with equipment maintenance programs and activities. 2.3.4 Management and organization impacts on maintenance effectiveness: Management functions such as planning, scheduling, procurement o'f parts, materials and services are critical items for maintenance. Except for PAT inspections and SALP reviews, few NRC programs directly focus on the role of management and organization in the development and administration of effective maintenance programs. However, several industry, state, and international initiatives are useful examples of methods to improve the management of maintenance. One of the NUMARC's principal initiatives since its inception has been the development of plant-specific indicators of operational and maintenance effectiveness for use by utility management. Some state public utility commissions (PUCs) are in the process of developing and implementing incentive programs by which utility performance is measured and analyzed to determine the impact on productivity. The prescriptive maintenance program and practices regulated by the Japanese government are credited for low trip rates and long mean time between events in that country's nuclear industry. The International Atomic Energy Agency (IAEA) guidelines documents which prescribe many aspects of maintenance organization and administration and other aspects of plant operation are endorsed by the nuclear regulatory bodies in Finland and Sweden. l
The regulatory role of the Federal Aviation Administration (FAA) in the management of commercial aviation maintenance is extensive and well codified. 2.3.5 Maintenance program criteria and standards: The NRC currently has no specific comprehensive regulations governing maintenance programs and activities. While several NRC Regulatory Guides endorse diverse industry standards that contain some maintenance guidance, there is no specific NRC guidance which addresses maintenance programmatically. Licensing reviews do not programmatically address maintenance practices,
- organizations, or functions and the NRC Standard Review Plan contains no comprehensive review criteria for maintenance.
Several of the national standards groups have subcommittees or working groups devoted to maintenance, most prominently IEEE, ASME, and ANS. While the NRC has sponsored several initiatives for development of maintenance guidelines, the industry has taken the lead in such development efforts with the publication of INPO maintenance guidelines. 2.3.6 The maintenance and operations interface: A close relationship between maintenance performance and plant operations is essential for a variety of routine plant manipulation, test, inspection, and repair activities governed by plant Technical l Specifications (TS). The basis for the test frequencies required in the Technical Specifications are being evaluated by the NRC using probabilistic methods. Plant quality assurance programs and activities apply to plant operations and maintenance but often draw a distinction between activities and
- systems that are considered to be safety-related and those that are not.
Activities proposed under the NRC Quality Assurance Program-Implementation Plan (SECY-85-65) have been designed to address problems contributing to repeated cases of major quality related failures at nuclear power plants. The quality of vendor-supplied equipment and information to licensees is audited by the NRC vendor inspection program. Industry and NRC efforts to monitor and measure plant performance are also a potentially important contributor to improved quality assurance initiatives. 2.3.7 Regulatory framework for maintenance. The NRC's current regulations pertaining.to nuclear power plant maintenance provide no clear programmatic treatment of maintenance. Regulatory requirements for maintenance, as confirmed by the NRC General Operating Criteria Interoffice Working Group in 1984, do not adequately cover all aspects of maintenance at operating nuclear power plants. Diverse industry standards address maintenance of specific equipment'or caecific activities.
3.0 CONCLUSION
S As the findings and activities of Phase I show, the problems identified in the MSPP are still evident throughout the industry. 3.1 Needed maintenance is not being accomplished or is not performed effectively. An average of 64% af total industry forced outage time is due to component failure. The forced outage rate, a measure currently used by INP0 as a maintenance measure, reflects the overall ability of the licensee to maintain the reliability of systems and components. Maintenance-related LERs accounted for 39% of industry LERs in 1984 and 48% in 1985. The ratio of maintenance-related LERs to the total number of LERs constitutes another measure of the contribution of maintenance to plant risk. Current maintenance programs and practices, as determined from Phase I surveys, are fragmented and less than effective in some areas. 3.2 A high percentage of failures result from improper performance of maintenance. As described above, maintenance-related LERs comprise 48% of total industry LERs. Although the total number of industry LERs decreased in 1984 due to changed reporting requirements, the proportion of maintenance-related LERs to the total has subsequently increased. More than 30% of the abnormal occurrences reported to Congress each year since 1975 may be attributable to maintenance. 3.3 The maintenance and operations interface is inadequate. There has been no significant reduction in loss of safety system function events since 1981. Out of 133 loss of safety system function events, 87 were caused by human error, including errors occurring during maintenance. A study of 35 instances involving human error in wrong unit and wrong train events showed that operations personnel were frequently at fault in the incidents (75%), many of which occurred during preparation for maintenance work. Faulty communications between operations and maintenance contributed to wrong unit / wrong train events. 3.4 The number of maintenance-related challenges to safety systems is excessive. l About 75% of the industry engineered safety features actuations reported during 1984 and 1985 were attributed to maintenance, surveillance and component failures. The stable trend (i.e., lack of improvement) in loss of safety system function events between 1981 ar.d 1984 warrants further attention to mechanisms which ensure availability of safety systems. 3.5 The major portion of occupational radiation exposure and many radiological hazards occur to personnel performing maintenance activities. Maintenance-related radiological exposures represented about 46% of total exposures in 1984. 4.0 RECOMMENDATIONS NRC and industry attention is warranted in the near term to resolve the I following technical and regulatory issues identified in the MSPP and confirmed by Phase I activities. Improve the development and use of measures of plant maintenance performance in order to review industry performance improvement against a baseline and to focus resources on appropriate plants and issues. Determine the contribution of human performance in maintenance to plant reliability and identify the safety impact of alternative strategies for improving human reliability. Define maintenance needs, including the role of preventive maintenance for l achieving acceptable levels of reliability and safety. Define goals for nuclear power plant reliability that foster management involvement in ensuring effective maintenance, including the use of performance measures and incentives. Develop and implement performance-oriented maintenance criteria and standards as the basis for effective maintenance programs. Further evaluation of cur ent maintenance program practices is necessary to determine facto; s influencing maintenance effectiveness. Identify approaches to maintenance which enhance the interface between maintenance and operations and ensure reliability. Coordinate NRC and industry initiatives related to maintenance. i
- j. !
APPENDIX A SELECTED NRC PROGRAMS AND INDUSTRY ACTIVITIES RELATED TO NUCLEAR POWER PLANT MAINTENANCE
1.0 INTRODUCTION
This appendix describes some of the important results of selected NRC programs and industry activities related to NPP maintenance that were ongoing during Phase I of the Maintenance and Surveillance Program Plan (MSPP). This discussion is presented in the context of the following six maintenance topics identified in the MSPP: Human Error in the Performance of Maintenance; l Indicators of Maintenance Effectiveness; The Role of Preventive Maintenance in Counteracting Aging and Service Wear Effects; Management and Organization Impacts on Maintenance Effectiveness; l Maintenance Program Criteria and Standards; and The Maintenance and Operations Interface. 2.0 HUMAN ERROR IN THE PERFORMANCE OF MAINTENANCE j A variety of factors contribute to human error in maintenance and affect ! personnel performance on-the-job. These include but are not limited to characteristics of the work place, equipment used, training and ] qualifications of personnel performing the work, and procedures used by personnel or governing the performance of work. NRC studies have focused on identifying the causes and effects of human error in maintenance. Unplanned Trips. While hardware failures are reported to be the dominant cause of unplanned reactor trips above 15% power, personnel-related problems (i.e., human errors or manual actions) are reported to be a substantial secondary cause of reactor trips above 15% power, accounting for 28%. Nearly half of the trips caused directly by human error were traceable to unlicensed personnel. Thus, errors by unlicensed personnel account for at least 10% of all reactor trips above 15% power (AE0D P504, " Trends and Patterns Report of Unplanned Reactor Trips at U.S. Light Water Reactors in 1984," August 1985). Critical Path Delays. From a study of 20 BWR outages, General Electric has determined that personnel error is the cause for 14% of critical path delays (only equipment failures [49%] and inadequate planning [19%] accounted for more delays) (Maintenance Training Overview course by General Electric, Nov. 4-8, 1985, San Jose, California). i A-1
~
Wrong Unit / Wrong Train. An NRC investigation of human error in 35 wrong unit and wrong train events at 10 sites has identified the most common contributing factors. Inadequate labeling of plant equipment, components, and areas was the leading contribution to error, with training and procedures also accounting for many of the events investigated (see NUREG-1192, " Investigation of the Contributors to Wrong Unit / Wrong Train Events," April 1986). Together, labeling, training and procedures were the primary contributors to human error in 60% of the wrong unit / wrong train events. Operations personnel were most often responsible for the wrong unit / wrong train events in performing actions preparatory to maintenance work. The study concluded that human errors leading to wrong unit / wrong train events can be reduced by improving plant labeling of systems and components so that personnel can readily identify unit / train designations. This study also recommends improving NRC guidance of independent verification and improvement of reporting of human performance issues in LERs. Nondestructive Examination Techniques. Personnel performance problems and technician training were among factors studied in NRC-sponsored research on nondestructive evaluation. Nondestructive examination (NDE) techniques, particularly ultrasonic testing / inservice inspection (UT/ISI) of pipes is one of the few predictive maintenance techniques used at NPPs (reported to be used by licensees in 70% of safety-related applications according to an NRC survey: see Volume 2 of NUREG-1212, " Status of Maintenance in the U.S. Nuclear Power Industry, 1985," May 1986). In reliability studies, investigators have often highlighted the variability of technician performance. According to one study, technician training at the EPRI-NDE Center appeared to have little effect on improving crack detection (NUREG/CR-4600, " Human Factors Study Conducted in Conjunction with a Mini-Round Robin Assessment of Ultrasonic Technician Performance," May 1986). Several human engineering aspects of the UT/ISI equipment and equipment use may contribute to less than optimal technician performance (e.g., design of controls, viewing screens, and radiation contamination problems). The study recommended further investigation of the effects of training on technician performance and that the evaluation be done using the relative operating characteristic (ROC) technique. Maintenance Procedures. Inadequate procedures were one often cited cause of events involving wrong unit / wrong train errors or loss of safety system function (see NUREG-1192, " Investigation of the Contributors to Wrong Unit / Wrong Train Events," April 1986). Failure to use procedures while performing work is also an obvious potential cause for human error. An NRC-sponsored study of maintenance procedures practices in the U.S. nuclear power industry revealed that industry maintenance procedures suffer from a variety of technical and human factors deficiencies (see NUREG/CR-3817, 1985). Deficiencies identified stensned from lack of human factors and technical writing expertise, incomplete and outdated vendor and technical information, lack of incentives for and enforcement of the use of procedures, high cost of procedures development, and lack of control over procedure review and approval processes. The study also investigated maintenance A-2 ___ _ _ _ _ _ _ . \
procedure development, use, and control in the military and the chemical, maritime, and aerospace industries. Effective maintenance procedures programs in non-nuclear applications were attributed to several factors, including use of administrative controls and regulation to direct the development, verification, periodic review, update, and use of procedures. Personnel Training and Qualification. Industry efforts related to human ) performance and human error currently focus on personnel training. INP0 ; accreditation of industry training programs, including electrical, l mechanical, and instrumentation and control technician training may improve ! personnel performance in the long term (only 12 plants thus far have received i accreditation for these programs), but industry implementation of performance-based maintenance training programs is still in its infancy. Conclusion. Most of the NRC's studies of factors affecting human error and personnel performance in maintenance focus on individual causative factors or events resulting from errors rather than the complex relationships among factors affecting human performance. Industry initiatives currently focus on personnel training as a means of improving human performance. The high contribution of human error to reportable events and to risk warrants identification of optimum means to improve human performance as an important determinant of plant reliability. 3.0 INDICATORS OF MAINTENANCE EFFECTIVENESS Several NRC and industry initiatives to collect objective and. reliable measures of NPP maintenance performance and effectiveness have been on-going during Phase I. As part of the research conducted in Phase I of the MSPP, DHFT has identified 31 performance data elements, based upon publicly available data, which may be used as measures of maintenance performance and effectiveness. They may be used to identify industry maintenance performance trends, to identify particularly good or poor performers and to assist in allocating regulatory effort accordingly, and to assess the impact of NRC and industry initiatives in maintenance (see Section 2.1 of the main body of this report). Properly constructed measures of maintenance performance would allow the NRC and others to: Monitor maintenance performance in industry and identify favorable / unfavorable trends; Identify weak performers and allocate regulatory effort accordingly; Identify strong performers and their good practices; and Assess the impact of NRC/ industry initiatives in maintenance. A-3
INP0/NUMARC Indicators. The nuclear utility industry is presently reporting data to INPO that are to be used as measures of plant performance and maintenance effectiveness. These data, which are not available for use by the NRC or the public, are the primary means by which NUMARC and INP0 will monitor the impact of regulatory and industry initiatives to improve NPP maintenance. INP0 uses these data to show, for example, that average numbers of plant scrams continue to decline yearly. INP0 also uses the indicators to establish that improvements in equipment reliability, as related to scrams, l are important for achieving reliability improvenents. INP0's interpretation I of industry trends and their significance is consistent with conclusions based on NRC indicators. NRC Measures of NPP Performance. The NRC has used a variety of separate methods and measures to evaluate licensee performance such as Licensee Event Report data, Systematic Assessment of Licensee Performance (SALP), Performance Appraisal Team (PAT) inspections, Safety System Functional Inspections (SSFI), and other routine and event-initiated inspections and licensee-reported data. Probabilistic risk assessment (PRA) techniques have also been used to estimate risk associated with certain systems, failures, and operating and maintenance practices. AEOD Trends and Patterns Analysis. AE0D trends and patterns analyses draw upon the Licensee Event Reporting System (LERs) to determine causes of reactor trips. They are able to determine the number of instances in which post-trip failures have gone uncorrected as demonstrated by their repetition following subsequent trips. Based on 1984 data, about 20% (77 out of 494) of reactor trips had post-trip recovery complications due to equipment failures or personnel errors unrelated to the original cause of the reactor trip (see " Trends and Patterns Report of Unplanned Reactor Trips at U.S. Light Water Reactors in 1984," AE0D/P504, August 1985). These represent errors which may be avoidable. Hardware failure appears to be the dominant cause of unplanned reactor trips above 15% power; 59% of these disturbances originated in balance-of-plant systems. This exemplifies that the largest single class of events leading to reactor trip is historically and currently outside of regulatory purview. For 1981-1983 event data, AEOD reports that variation in both the quantity and quality of LERs between nuclear power plants seem to be extensive (see " Exploratory Trend and Pattern Analysis of 1981 through 1983 Licensee Event Report Data," draft NUREG/CR-4129, January 1985). Due to variations in LER reporting among plants, it is difficult to infer that plants having higher incidences of LER reporting have more reliability / maintenance problems than other plants. It is also difficult to determine what personnel were involved in almost three-fourths of the instances of personnel error reported in LERs. A , review of the LER abstracts is necessary to determine (where possible) what personnel may have been involved. Analysis of trends and patterns in LERS should be made consistent with other analyses of licensee events and performance data. Future analyses should include: number of scrams caused by hardware failure and by personnel errors; number of A-4
maintenance-related scrams (activity underway at time of scram); and number of post-scram recovery complications (both hardware and personnel errors). Performance Appraisal Team (PAT) inspections are performed at operating l reactor sites by the Office of Inspection and Enforcement (IE). One l objective of these inspections is to evaluate the effectiveness of licensee implementation of management control systems in selected areas important to operational safety. One area selected for evaluation is maintenance. The evaluation includes interviews with the staff and management involved with plant maintenance, review of the licensee's programs, and evaluation of licensee implementation of the program. The number of PAT inspections is limited by resources; 31 PAT inspections have been completed since the program began in 1979. Safety System Functional Inspections (SSFI). In September 1985, IE initiated Safety System Functional Inspections (SSFI). This team inspection, which uses PAT inspectors, involves a detailed assessment of the operational readiness of a s(lected safety system (s). The assessment includes a thorough evaluation of design controls, testing, operational training, and maintenance of the selected system. SSFI inspections provide the most comprehensive routine reviews of maintenance currently done by NRC. In-Plant Reliability Data. The NRC-sponsored in-plant reliability data base for nuclear plant components (IPRD) was developed to provide a component failure and repair data base for use in probabilistic risk assessment (FRA) (see NUREG/CR-2641, "The In-Plant Reliability Data Base for Nuclear Power Plant Components: Data Collection and Methodology Report," July 1982). IPRD was established through a cooperative effort with five nuclear power plants (3 PWRs and 2 BWRs) to collect descriptive failure and repair data for selected components (e.g., pumps, valves, diesel generators, inverters, battery chargers, and batteries). IPRD includes information on equipment degradation as well as incipient and catastrophic failures. Since the IPRD system includes non-catastrophic failure data, it has the potential for providing information useful in establishing plant-specific maintenance program requirements. The data base also contains information on safety-related and nonsafety-related components. The reported maintenance frequency of safety-related valves is approximately twice that of nonsafety-related valves, without a corresponding improvement in reliability estimates over those of nonsafety-related valves. Incipient failures dominate the reported failure data for all pump, valve and diesel generator failure data (incipient failures are defined as component still functional but detected deficiencies that would probably progress to degraded or catastrophic level if left unattended). The preliminary data show that for valves, pumps, and diesels, incipient failures predominate. Thus, the data demonstrate the potential role of maintenance in preventing equipment degradation. A-5
Probabilistic Risk Assessment. The importance of preventive and corrective maintenance functions and surveillance was analyzed as part of a probabilistic study for the NRC reliability research program (NUREG/CR-4377, " Evaluation and Utilization of Risk Importances," August 1985). The importance of proper maintenance for various systems shows five orders of magnitude of variation (i.e., a factor of 100,000) which affects risk astimates. For surveillance tests, there were four orders of magnitude variation (i.e., a factor of 10,000) effect on risk for some surveillances (e.g., emergency power systems vs. core flood , systems). The study did identify several systems for which maintenance l appears to have the greatest contributions to risk estimates. Estimated mdintenance importance and risk estimates (e.g., for probability of core
- r , serv melt)werehighes})for: emergency DC power (8.5),y)eactor protection system(3.1 system (8.5x x10-10 1),, emergency AC power (1.1 x 10emergencyfeedwatersy feedwater initiation and control system (2.7 x 1 -7), battery and switchgear emergegcy cooling system (2.4and x 10-2)0 safety relief system (2.3 x 10 ). Thelargestsurveillanceandtestcgntribution risk was from batte DC power (1.3 x 10-gy ), e emergency cooling gency AC power (1.1system x 10- (1.5
, andg)10-safety), emergency relief system (1.0 x 10-Risk estimates, based on a plant-specific probabilistic risk assessment study, are one means of identifying maintenance and surveillance needs on a plant-specific basis. This methodology can be used by the NRC to focus on the safety significance of maintenance and surveillance activities and can be used to derive safety and reliability goals for maintenance.
Systematic Assessment of Licensee Performance. The Systematic Assessment of Licensee Performance (SALP) is one mechanism used by the NRC to collect available observations on a periodic basis and to evaluate licensee performance based upon those observations. The objectives of SALP evaluations are to improve the NRC regulatory program, to permit sound decisions regarding NRC resource allocations, and to improve licensee performance. Maintenance (both preventive and corrective) and surveillance are 2 of 11 functional areas addressed in SALP evaluations. Both positive and negative attributes of licensee performance are considered. Management involvement and staffing are considered in these evaluations. The integrated SALP assessment is intended to be sufficiently diagnostic to provide a rational basis for allocating NRC resources and to provide meaningful guidance to licensee management. Occupational Exposures. Maintenance personnel received 46% of the total occupational exposures for NPP personnel in 1984. Maintenance jobs which were reported to be large contributors to BWR doses in 1984 included recirculation piping replacement, inspection for intergranular A-6
stress corrosion cracking and repair, Mark I torus modifications, and reactor vessel component inservice inspection. Steam generator maintenance and repair, including tube inspection, sleeving and plugging were major sources of exposure at PWRs in 1984 (LWR Occupational Dose Data for 1984," memorandum from F. Congel to D. Muller, NRC, September 27,1985). Based on a sample of plants surveyed in an NRC-sponsored study of high-dose jobs and occupational dose reduction, worker exposure data are seldom quantified for use during equipment selection (see NUREG/CR-4254,
" Occupational Dose Reduction and ALARA at Nuclear Power Plants: Study of High-Dose Jobs, Radwaste Handling, and ALARA Incentives," June 1985).
Application of dose data to improvement of preventive maintenance programs is not done routinely and systematically. Lack of time and difficulty in abstracting the necessary data are reasons cited for failure to apply exposure data to modify preventive maintenance programs. ALARA policy, data management and job reviews are recommended for improving direction of ALARA efforts on a plant-specific basis. Studies conclude that radiological data tracking and trending are necessary for coordinating operations and maintenance activities and for achieving improvement in occupational dose reduction. Conclusion. The NRC should continue to develop measures to characterize and prioritize critical maintenance functions and to collect data which improves the agency's ability to understand and derive measures of the relationship between plant functions and operational safety. The ability to ultimately predict plant performance trends and to allocate NRC resources accordingly, a goal consistent with 1986 Policy and Planning Guidance (NUREG-0885, Issue 5), should be supported by further refinements in operational data and their analyses. Current NRC data sources, as described above, do not provide the types of data that permit easy identification of strategies to foresee needed improvements in maintenance performance at the plant level. 4.0 THE ROLE OF PREVENTIVE MAINTENANCE IN COUNTERACTING AGING AND SERVICE WEAR EFFECTS Several NRC and industry efforts provide data and information related to aging and other equipment performance problems. Nuclear Plant Aging. The nuclear plant aging research (NPAR) sponsored by the NRC is investigating the effects of aging and service wear on systems, equipment, and components (see NUREG-1444, " Nuclear Plant Aging Research (NPAR) Pro Research NPAR) (gram Plan," Julyhave activities 1985). The results concentrated onofdescribing the Nuclear Plant failure Agingand modes aging effects on specific components and systems. The findings show the importance of inspections and surveillance testing as a method for detection of the onset of aging as well as to determine the role of maintenance in preventing or correcting failures. For example, the NPAR has reviewed valve A-7
and motor-operator failure data, manufacturers' recommended maintenance and surveillance practices, and operating experience data for motor-operated valves (MOV). Results indicate that testing activities had a major role in discovery of valve failures (54%) while only three percent of valve failures 1 were discovered during maintenance. A large proportion of failures were 4 discovered during normal operation (31%). The results of the many separate l NPAR products will need to be reviewed upon completion in order to identify j generic findings with applicability to industry activities and NRC regulatory l policy. NPAR has also provided a review of condition monitoring as a technique for ! providing a quantitative assessment of components and parameters for predicting equipment degradation. This technique allows for scheduled I maintenance and replacement _ intervals based on actual and predicted equipment performance. Condition monitoring requires data collection and analysis to determine equipment performance trends and to allow for appropriate scheduling of inspection, preventive maintenance, or replacement activities. Most utilities do not perform condition monitoring due to inadequate knowledge of degradation mechanisms and the relationships between measurable parameters and predicted functional capability. Monitoring Equipment Qualification. A specific type of equipment degradation of concern to the NRC is that due to the effects of the NPP environmental conditions. One of the primary goals of plant maintenance activities should be to ensure the operability of equipment important to safety in all anticipated environmental conditions, especially during accidents. As determined from NRC inspections, in general, the industry has not yet fully established firm programs to integrate equipment qualification requirements with equipment maintenance programs and activities. Environmental qualification requirements for electrical equipment of nuclear power plants (10 CFR 50.49) are intended to ensure that equipment important to safety is able to perform its design functions throughout its installed life. The effectiveness of industry activities to ensure environmental qualification of equipment can be evaluated by observing plant-specific component failure trends. Data to make such an evaluation are not currently available to the NRC. Sample data may be obtained and analyzed through NRC's in-plant reliability data system (IPRDS). However, equipment failure root cause is often difficult to determine from licensee data (the source of IPRDS data) as well as from NRC event reporting systems. Many utilities are not well equipped to collect, analyze and trend comprehensive component maintenance or failure data. Industry Research. Many individual industry initiatives are germane to the issue of plant equipment aging and maintenance. EPRI's research programs include studies of equipment diagnostic techniques and hardware and methods improvements. INP0's event reporting system and good practices documents provide mechanisms for documenting and disseminating information on equipment performance problems. The Nuclear Plant Reliability Data System (NPRDS) operated by INP0 is another industry method for evaluating equipment performance. A-8
Conclusion. Both NRC and industry initiatives in equipment aging and performance are producing data and information on specific equipment and techniques. These programs, while providing important data for industry use, should also be focused on common safety or reliability goals to enhance usability and compatibility of results. This would allow for consistent assessment of component performance and facilitate feedback of results of investigation for improving design, operation, and maintenance. 5.0 MANAGEMENT AND ORGANIZATION IMPACTS ON MAINTENANCE EFFECTIVENESS NRC has not traditionally identified nor analyzed management impacts on plant performance and maintenance except in response to industry events. In fact, the role of management and organization is hard to pinpoint because licensee operational data have not supported a detailed analysis of utility management impacts. However, management functions such as planning, scheduling, procurement of parts and materials and services are critical items for maintenance (in fact, inadequate planning has been cited as the cause of 19% of critical path outage delays at a sample of BWRs). The staff has compiled data which describe some of the characteristics of plant maintenance organization and administration from site surveys and questionnaires administered to all Resident Inspectors (see Section 2.1 of the body of this report). Outside the scope of the MSPP, few NRC programs, except as mentioned above, focus directly on the role of management and organization in the development and administration of effective maintenance programs. However, several industry, state, and international initiatives are useful examples of initiatives to improve management. NUMARC. The Nuclear Utility Management and Human Resources Committee (NUMARC) was formed in 1984 to bring management level attention to some of the operational safety and regulatory issues facing the industry today. One of NUMARC's principal initiatives since its inception has been the development of plant-specific indicators of operational and maintenance i effectiveness for use by utility management. l Public Utility Commissions (PUCs). State public utility commissions (PUCs) I are in the process of developing and implementing incentive programs by which utility performance is measured and analyzed to determine the impact on productivity. PUCs plan to use performance measures as the basis for establishing incentives for utility improvements in operating efficiency. Foreign Maintenance Practices. In Japan, maintenance is under the regulatory approval authority of the Ministry of International Trade and Industry (MITI). The prescriptive maintenance program, including disassembly and ) I inspection of individual components, and practices regulated by the government are credited for Japanese low trip rates and long mean time between events. International Atomic Energy Agency (IAEA) guidelines documents covering maintenance and other aspects of plant operation are utilized by several European countries (e.g., Finland and Sweden), thus < prescribing many aspects of maintenance organization and administration. l l A-9
Federal Aviation Administration (FAA). The regulatory role of the FAA in commercial aviation maintenance is very prescriptive and well codified (see Appendix B for a discussion of the FAA regulation of maintenance). Industry-developed maint ance program requirements, codified by the FAA, were developed in an ata 3;gre of management concern for safety and a preference for having re alii'[,ity-based maintenance intervals rather than scheduled maintenance. m Conclusion. The role of management and organization in U.S. nuclear plant maintenance is less well-defined than for some other countries and industries. Although the importance of management continues to be identified in event follow-up investigations as a major influence on plant reliability and safety, current regulations lack a sufficient basis for defining a management goal for maintenance and overall plant reliability. ' 6.0 MAINTENANCE PROGRAM CRITERIA AND STANDARDS The NRC currently has no comprehensive regulations governing maintenance programs and activities (see Appendix B for a more thorough discussion of NRC regulations and maintenance). While several NRC Regulatory Guides endorse diverse industry standards that contain some maintenance guidance, there is no NRC guidance which addresses maintenance programmatically. Licensing reviews do not address maintenance practices, organizations, or functions and the NRC Standard Review Plan contains no criteria to support programmatic reviews of maintenance. As previously noted, SALP reviews and PAT inspections are among the few methods NRC uses to conduct broad-based reviews of licensee programs. However, several of the national standards groups have subcommittees or working groups devoted to maintenance, most prominently IEEE, ASME, and ANS. IEEE Standard Group. IEEE Subcommittee 3, Operations, Surveillance, and Testing, has established a Maintenance Good Practices Working Group (WG 3.3) to develop industry guidance on maintenance good practices. The working group has the lead in the generation of maintenance standards for electrical equipment. A maintenance process document and multiple recommended practices documents for specific components are in preparation. The IEEE Standards Group appears to be the most active of the national standards groups currently working on maintenance and surveillance related issues -- a proposed revision to IEEE Standard 338-1977 " Periodic Surveillance Testing of Nuclear Power Generating Station Safety Systems" contains perhaps the most programmatic emphasis of the IEEE standards relevant to maintenance; however, this standard only briefly addresses the programmatic need _ for modification of test intervals to reflect equipment performance and periodic evaluations of machinery experience. ASME Committees. The ASME Committee on Nuclear Quality Assurance (NQA) is responsible for development and maintenance of standards covering quality assurance programs and practices including ANSI /ASME NQA-2-1983 " Quality Assurance Requirements for Nuclear Power Plants." The NRC staff is currently working on a regulatory guide to endorse NQA-2-1983 and the NQA-la-1985' A-10
Addenda. The standard contains QA requirements for specific work practices. Revisions to NQA-2 are in preparation that are intended to provide amplified requirements for certain quality assurance activities, including maintenance. The ASME Committee on Operation and Maintenance of Nuclear Power Plants has l published and is developing a number of standards for maintenance of specific ' mechanical components. ANS. The American Nuclear Society (ANS) standards are directed mainly at the design and construction of nuclear power plants. The operational aspects of nuclear power generation and the maintenance needs resulting from power plant operations over an extended period of time have been addressed in general terms by ANSI /ANS-3.2-1982. This standard is aimed to provide generic guidelines related to the administrative control and quality assurance of plant operations. Concerning maintenance, the standard identifies general rules for identifying the subject areas that are to be addressed by the maintenance program, but does not extend to providing detailed guidelines for the implementation or the evaluation of the effectiveness of the maintenance program. NRC Regulatory Guide 1.33 (Feb.1978) currently endorses ANSI N18.7-1976/ANS-3.2, " Administrative Controls and Quality Assurance for the Operational Phase of Nuclear Power Plants." ANS Working Group 3.9 initiated work on a maintenance management program document in late 1984. The intent of the Standard was to provide a description of maintenance in terms of major functional requirements. The activities of this ANS committee were deferred in consideration of INP0/NUMARC initiatives. A guideline cocument for maintenance has been developed and published by IAEA (Manual on the Maintenance of Systems and Components Important to Safety, Rev. 2, June 1985). This document is intended to prescribe the essential elements of an effective maintenance program and to define corrective and preventive maintenance. The IAEA document references numerous IAEA safety guides which contain further guidance on specific aspects of operation and quality assurance activities for nuclear power plants. The document provides guidance for many specific maintenance functions and activities. This manual is routinely used in other countries (e.g., Finland). INP0. The INP0 Maintenance Guidelines (INP0 85-038, October 1985) provide generic guidance for establishing multiple maintenance functions (e.g., ! maintenance organization, training, facilities, procedures, planning, l procurement, etc.) with emphasis on the purpose of such maintenance elements. The document references numerous other INP0 guidance documents without giving l detailed implementation guidance. The document also references the EPRI guide for Developing Preventive Maintenance Programs in Electric Power Plants but does not describe methods of analysis and trending appropriate for implementing reliability-centered maintenance. INP0 also applies some broad maintenance objectives and criteria during their plant evaluations. A-11
INP0 has also developed writing guidelines for maintenance, test, and calibration procedures (INP0 85-026). The guidelines provide the information necessary for development of plant specific procedures writing guidelines and for review of existing procedures. The document addresses development of procedures but not procedures use or control. The degree of industry implementation of the INP0 guidance is unknown, although several utilities surveyed by the NRC, reported using INP0's writing guidelines for development of new maintenance procedures. EPRI. Both EPRI (NP-4350, December 1985) and the NRC (NUREG/CR-3517, March 1985) have developed Human Engineering Guidelines for Maintainability of Nuclear Power Plants. These guidelines describe human engineering characteristics that should be applied to the design of new nuclear power plant facilities and equipment or to the review of operating plants to identify needed maintainability improvements. The guidelines cover physical, environmental and anthropometric aspects of the man / machine interfaces involved in power plant maintenance as well as preventive maintenance. The extent of industry application of such guidelines is unknown. EPRI also has developed and published a human engineering checklist for evaluatin This document (NP-2360, May 1982)g maintainability
, is a in nuclear useful tool which addressespower plants.of maintenance a variety functions, including organizational factors, equipment maintainability, and personnel selection and training. Maintainability guidance for industry use appears to be adequate. The extent of industry implementation is unknown.
EPRI's guide for Developing Preventive Maintenance Programs in Electric Power Plants (NP-3461, May 1984) presents a model for development and implementation of a preventive maintenance (PM) program for nuclear power plants. The model includes the identification of critical equipment items to be considered for a PM program, the establishment of the scope and recommended frequency of PM activities, and methods to determine cost-benefit and program optimization analyses. The model includes use of failure cause data in determining preventive maintenance program requirements. The guide presents a complete and accurate picture of the process for analyzing needs, and developing, implementing and evaluating a PM program. The degree of industry implementation is unclear, however, the site surveys performed for the MSPP did not reveal that the licensees have performed the analytic
- activities necessary to systematically determine or evaluate plant PM program requirements.
Other Industry Initiatives. Informal mechanisms of maintenance information exchange are utilized to some extent by the industry (e.g., the Nuclear Operations and Maintenance Information Service (NOMIS) marketed by NUS Corporation to over 61 domestic utilities). INP0 publishes significant events, good practices and compiles equipment reliability data (i.e., NPRDS) for industry use. The INP0 maintenance superintendent's workshops held annually provide another industry forum for informal exchange of maintenance experiences and practices. These workshops are well-attended and are an apparently effective means of disseminating maintenance information, usually by case studies. A-12
NRC Research. NRC-sponsored research on factors affecting ultrasonic testing (UT) of pipes for stress corrosion cracking (SCC) is expected to provide input to development of criteria, requirements, and processes for certifying l personnel, equipment, and procedures for UT. The improvement of UT l reliability, which is one of the industry's main predictive maintenance ' techniques, appears to be an arena where industry research and initiative continues to further the technology. Further NRC scope of concern should focus on dissemination of research results and support of industry standards development. Similarly, the Nuclear Plant Aging Research can contribute the results of many component-specific studies to industry standards development initiatives. Conclusion. The NRC has been actively participating in and encouraging industry standards groups' development of improved maintenance guidance documents. While the NRC has sponsored several initiatives for development of maintenance guidelines (i.e., procedures and maintainability), the industry has clearly taken the lead in such development efforts. As part of its regulatory responsibility, the NRC should continue to encourage and endorse programmatic maintenance standards and guidelines developed by the industry, as appropriate, to achieve an integrated, performance-oriented approach to maintenance. It is clear that a variety of standards, guidelines and research publications which define maintenance functions are available to the industry. Integration of diverse industry standards for maintenance is not yet an issue addressed by the industry or the NRC. The surveys of maintenance programs conducted at eight plants revealed that licensees are for the most part poorly equipped to compile and analyze the extensive amounts of information necessary to develop, implement and evaluate a reliability-based maintenance program. The extent of implementation of industry maintenance guidelines is unknown. The NRC should monitor industry use of standards and other industry guidance documents to identify any conflicts or issues not addressed by the current fragmented approach to maintenance standards and requirements. 7.0 THE MAINTENANCE AND OPERATIONS INTERFACE The goal of maintenance, as an integral function of nuclear power plants, is to ensure plant operability. In fact, the close relationship between maintenance performance and plant operations is evident from the role of both operations and maintenance personnel in a variety of routine plant manipulation, test, inspection, and repair activities governed by plant Technical Specifications (TS). Operations personnel are most frequently involved in performing the technical specifications surveillance requirements. However, if needed maintenance is identified during the routine surveillance, maintenance personnel are those most likely designated to implement the needed repairs. Therefore, A-13 l
maintenance personnel are those with the responsibility of restoring equipment to the design basis specified in the plant operating license and technical specifications. Most NRC activities addressing problems of the maintenance and operations interface have focused on human error rather than on planning, coordination and integration of maintenance and operating activities (see Section 2.0 of this Appendix). Technical Specifications. Analyses are needed to improve the basis for test methods and test frequencies required in the Technical Specifications. At present, there is no standard methodology to determine surveillance test intervels. An NRC research program is currently studying probabilistic methods to determine allowed outage times and surveillance test intervals. Research results from one PRA study estimate that current Technical Specifications allow an extremely wide range (six to seven orders of magnitude) of risk for various degraded modes of operation and test intervals (see " Evaluation of Risk Impacts of AN0-1 Technical Specifications," P. L Samanta et al, draft report, August 1985). In addition, about half of the Technical Specifications requirements were estimated to have minimal risk impact. The lack of documented bases for test frequencies, action statements, test types, and allowed outage times is a significant issue that the NRC is currently addressing. NRC Quality Assurance Programs. Quality assurance activities are performed by licensees to ensure that plant oper,ations and maintenance are performed according to technical specifications, applicable standards, requirements, and procedures. Plant quality assurance programs and activities apply to plant operations and maintenance but draw a distinction between activities and systems that are considered to be important-to-safety / safety-related (ITS/SR) and those that are considered important-to-safety /nonsafety-related (ITS/NSR). The issue of what NPP systems, equipment, and components should be considered ITS/NSR has not yet been resolved by the NRC. While ITS/NSR systems are to be identified during the licensing process, designation of and requirements for ITS/NSR systems at operating reactors has not been determined. As presently drafted, the proposed rule defining ITS and SR states that normal industry practice is acceptable for ITS/NSR equipment. Activities proposed under the NRC Quality Assurance Program Implementation Plan (SECY-85-65) are designed to address problems contributing to repeated cases of major quality-related failures at nuclear power plants. The key elements of the QA program plan include use of performance-oriented quality assurance programs to focus NRC and utility resources on identification and analysis of performance trends in order to apply corrective action to quality problems. Alternative means of optimizing both utility quality assurance resources and NRC inspection effort are being investigated. Plant performance and maintenance measures developed as part of the MSPP could be used as input to performance-oriented measures of QA effectiveness. For example, in a performance-oriented QA program for plant operation a method may be developed to measure and analyze operational data indicative of plant QA program performance, such as errors in surveillance testing, repetitive unplanned reactor trips from similar causes, and amount of rework in safety system maintenance. A-14
NRC Vendor Inspection Program. Activities of the NRC vendor inspection program are contributing to the improvement of equipment quality by ensuring, on a sample basis, that the procurement of equipment and vendor products / services meet NRC requirements. The quality of vendor-supplied equipment and information to licensees was a problem addressed in part by NRC Generic Letter 83-28, requesting licensees to commit to programs which would ensure that vendor infornation for safety-related components is available. 2 The NRC vendor program activities focus on specific operational safety issues resulting from vendor products / services and, therefore, do not generally have broad, programmatic applicability. Rather, the inspection activities are primarily concerned with specific problems resulting from equipment procurement, production, and utility implementation of vendor products / services and information affecting quality of plant operations and reliability. Methods of consistent and reliable evaluation of industry activities affecting reliability are needed for NRC use, l Conclusion. The maintenanct and operations interface is not being directly addressed by NRC initiatives. NRC research may result in an identification
, of more performance-oriented bases for technical specifications. Other NRC initiatives affecting quality of both maintenance and operations currently do not focus on common goals for plant reliability and safety to ensure consistency of regulatory activities. This illustrates the need for performance-oriented approaches to ensure reliability.
i a l i i A-15
APPENDIX B CURRENT NRC REGULATION OF MAINTENANCE IN COMMERCIAL NUCLEAR POWER PLANTS 1.0 THE CODE OF FEDERAL REGULATIONS AND THE REGULATION OF MAINTENANCE IN COMMERCIAL NUCLEAR POWER PLANTS 1.1 Introduction
< The Atomic Energy Act, Energy Reorganization Act, and Title 10 of the Code of Federal Regulations, predominantly Part 50 (10 CFR 50), establish the legal basis for the NRC's regulation of commercial nuclear power plants (NPPs).
When a utility applies to the NRC for a construction permit (CP),10 CFR 50 l is the major legally binding document to which they must conform. However, various other NRC documents (Regulatoyy Guides, NUREGs, etc.) and industry standards, e.g., ANSI /ANS, IEEE, ASME , become binding as a result of their incorporation in the CP application (and amendments thereto), Final Safety Analysis Report, and subsequent incorporation into the provisions of the operating license. 10 CFR 50 is the basic regulatory structure upon which the licensing process is built. Weakness in this structure affect the overall regulatory framework governing operation of licensed nuclear power plants, j An examination of 10 CFR 50 shows that it contains no major reference to maintenance. Maintenance is only mentioned once in the body of 10 CFR 50: Contents of applications; technical information, paragraph (b)(6)(iv), which refers to the Final Safety Analysis Report (FSAR), states that the applicant shall include " plans for conduct of normal operations, including maintenance, surveillance and periodic testing of structures, systems, and components" (10 CFR 50.34). 1.2 FAA Regulation of Maintenance I In contrast, in the Code of Federal Regulations, Aeronautics and Space,
; Parts 1-199, dealing with the Federal Aviation Administration (FAA) regulations, maintenance occupies the entirety of Part 43 of Title 14, which consists of 17 sections and 5 appendices. A recent FAA fine of $9.5 million against a major airline for maintenance violations also shows the seriousness with which the enforcement of maintenance regulations is taken. The 1
I ANSI /ANS - American National Standards Institute /American Nuclear Society IEEE - Institute of Electrical and Electronic Engineers ASME - American Society of Mechanical Engineers B-1
l 1
,1 commercial aviation industry maintenance program requirements began to evolve after WW II with reliability measurement as a criterion for an air carrier to i gain relief from restrictive FAA overhaul requirements. The collection of )
data by the airlines subsequently led to the more precise specification of goals and objectives for aircraft reliability. Focus on preventive maintenance became a major element of the FAA approach to safety to ensure continued in-service reliability. The aviation industry's prescriptions for maintenance activities were developed and promulgated by industry groups which produced maintenance manuals and maintenance programs. An industry steering group authored a policy and procedures handbook which became the ' basis for FAA guidance in development of preventive maintenance programs for all certified transport aircraft. The basis for FAA regulation of maintenance evolved from designating prescriptive overhaul intervals to the current use of reliability methods and condition monitoring as the basis for maintenance program development. A good summary of the evolution of the FAA regulations can be found in EPRI report NP-3364, Commercial Aviation Experience of Value to the Nuclear Industry, January 1984. The FAA has authority for the initial preventive maintenance program for each newly certified aircraft. The NRC does not specifically approve maintenance programs as part of facility licensing. Utility FSARs do not include a full maintenance program description. NRC instead discusses maintenance, surveillance, and testing of structures, systems, and components in accordance with component specifications and Technical Specifications. In the aviation industry, maintenance requirements for critical items differ (e.g., required independent inspection of work) from maintenance for other aircraft components. NRC requirements for activities affecting systems "important to safety" also differ from other activities, although maintenance requirements in either case are very general. Effective preventive maintenance programs in the air transport industry ensure operational reliability of aircraft and have been emphasized because of the direct relationship to economics as well as safety (EPRI NP-3354). While a utility's maintenance and surveillance activities for structures, systems, and components are required to be described in the Final Safety Analysis Report (FSAR) by 10 CFR 50.34, the NRC does not require a preventive maintenance program and does not prescribe requirements for specific maintenance program functions. In fact, no formal integrated maintenance program review and approval is conducted by the NRC for licensing. The airline industry has an active reliability control program which relies on extensive collection, evaluation, and use of reliability and maintenance event data. Each airline is able to develop and operate a program to meet its individual needs and is monitored by the FAA. In contrast, commercial nuclear power plants may or may not develop and use reliability data, and no approval or control of maintenance related data is performed by the NRC. A voluntary industry data reporting system exists (NPRDS) but is not used by the NRC to determine maintenance needs. The NPRD system has now improved to the point where it may warrant further use by the NRC as a source of reliability data to meet agency needs (AE0D Annual Report, April 1986). l l B-2
1.3 NRC Regulation of NPP Maintenance The subject of NPP maintenance is not treated programmatically in NRC regulations nor is it defined nor described as it is in the airline industry. General performance objectives for maintenance are not provided. Clearly, equipment experience history data plays a major role in reliability and maintenance in the aviation industry. The central feature of maintenance program' development, that is, information collection and feedback, is not well developed among the nuclear utilities. The aviation industry emphasizes use of reliability data starting with the design phase, following through to aircraft operation and continual monitoring by the FAA. The nuclear industry has evolved without the extensive use of reliability data as the basis of regulation of design and operating activities. Additional observations which can be made regarding the lack of NRC focus on NPP maintenance are as follows: Maintenance is not mentioned in requirements for the contents of the Preliminary Safety Analysis Report (PSAR). No mention is made of maintenance in the regulations on construction permits. No mention is made of maintenance in regulations concerning technical specifications (although " surveillance requirements" as one of the categories required to be covered in technical specifications is defined in the Code as preventive measures). In-service inspection requirements are covered under 10 CFR 50.55a, Codes and Standards, but no attempt is made to connect this constituent of preventive maintenance to the concept of a maintenance program. No mention is made of maintenance in 10 CFR 50, Appendix A, General Design Criteria, although maintaining system reliability within design limits is the purpose of maintenance. Inspection and Testing are mentioned, but not as part of a comprehensive maintenance program. The 1985 Commission Policy Statement on Training and Qualification endorses the Institute of Nuclear Power Operations (INP0) accreditation of industry training programs, including training for electrical, mechanical, and I&C technicians. The Policy Statement endorses the essential elements of performance-based training but does not contain prescriptive guidance. 1.4 Conclusion In conclusion, NRC regulations do not provide a basis for a coherent approach to regulating NPP maintenance, and this is also clearly reflected in the way maintenance is treated in ancillary guidance (Regulatory Guides, NUREGs, etc.) promulgated by the NRC. B-3
.__ ___a
2.0 NRC GUIDANCE ON MAINTENANCE OF COMMERCIAL NUCLEAR POWER PLANTS 2.1 Introduction This section describes guidance promulgated by the NRC staff to assist the applicants and licensees in interpreting and complying with 10 CFR 50. Such guidance is not binding unless incorporated into the provisions of the operating license as previously discussed. It also describes the guidance on methods acceptable to the NRC staff for implementing maintenance regulations through " Regulatory Guides" which endorse standards developed by professional societies, such as ANS, IEEE, and ASME. 2.2 Guidance on Maintenance to Applicants Preparing Preliminary (PSAR) and Final Safety Analysis Reports (FSAR) Basic guidance on the required contents of the PSAR and FSAR is contained in 10 C"R 50. However, applicants take their cues on what to include in these reports from the Standard Format and Content of Safety Analysis Reports for Nuclear Power Plants (Regulatory Guide 1.70) and the Standard Review Plan (SRP) (NUREG-0800, Rev. 0-4, July 1981). None of the chapters in the SRP deal directly with maintenance. Chapters 13 and 17 contain the few references to maintenance in the SRP: Section 13.1.1. This section states that a description of responsibilities for various areas should be included in a SAR review and the " development of plant maintenance programs" is one of the 11 areas mentioned. Nowhere is a description of the actual maintenance program required. This paragraph also states the FSAR should provide " educational and background experience" for certain categories of management / supervisory personnel, including maintenance support. Sections 13.1.2 - 13.1.3. These sections state that the applicant should
" describe the structure, functions, and responsibilities of the organization j established to operate and maintain the plant." It goes on to state that ;
education, training, and experience should be described for " management, operating, technical, and maintenance positions." Section 13.2.2. This section of the SRP addresses " Training for Nonlicensed Plant Staff." Maintenance is not addressed directly in this section; but Regulatory Guide 1.8 and ANSI /ANS-3.1 are referenced under acceptance criteria and these documents do address maintenance personnel. Section 13.5.2. This section on " Operating and Maintenance Procedures" states that the PSAR should describe " preliminary schedules" for the preparation of operating and maintenance procedures and that the FSAR should provide descriptions "as to the nature and content of procedures," including instrument calibration / test, and maintenance and modification procedures. This section of Chapter 13 contains the only acceptance criteria that specifically address maintenance. The first,13.5.2-IIA, requires a
" generally acceptable" target date for the completion of operating and maintenance procedures. The second, 13.5.1-IIC, references staff positions in Regulatory Guide 1.33 and ANSI /ANS-3.2.
B-4 t
Chapter 17. This chapter of the SRP deals with the ways that the applicable requirements of Appendix B of 10 CFR 50 are satisfied, focusing only peripherally on the issue of maintenance. The most relevant reference is found in SRP 17.1, dealing with acceptance criteria for " activities related to corrective action" during design and construction. This section states that.such activities are acceptable if:
- 1. Procedures are established and described indicating an effective correc'tive action program has been established. The QA organization reviews and documents concurrence with the procedures.
, 2. Corrective action is documented and initiated following the determination of a condition adverse to quality (such as'a ! nonconformance, failure, malfunction, deficiency, deviation, and defective material and equipment) to preclude recurrence. The QA organization is involved in the documented concurrence of the adequacy of the corrective action.
- 3. Follow-up action is taken by the QA organization to verify proper implementation of corrective action and to close out the corrective action in a timely manner.
- 4. Significant conditions adverse to quality, the cause of the conditions, and the corrective action taken to preclude repetition are documented and reported to immediate management and apper levels
; of management for review and assessment.
Conclusion. Only isolated references to maintenance are scattered throughout Chapters 13 and 17 of the SRP. Also, no definition is provided for corrective, preventive, or predictive maintenance. Furthermore, acceptance criteria in the SRP are grossly incomplete for judging the adequacy of a NPP's maintenance program. Finally, a review of FSARs for near-term operating licenses (NT0Ls) reveals
- that these documents contain only minimal reference to maintenance and, as currently constituted, do not provide the NRC staff with information necessary to judge the adequacy of plant maintenance programs.
2.3 Guidance on Methods Acceptable to the NRC Staff for Implementing
]
Maintenance Regulations: Regulatory Guides Regulatory Guides (RGs) are not legally binding on utilities seeking licenses for commercial reactor operation unless they are incorporated in the utility's operating license application and that application is accepted by NRC. Regulatory Guides 1.33 and 1.8, which describe quality assurance and personnel qualifications respectively, contain the most complete treatment of maintenance program requirements. They are reviewed in the next two sections of this report. J 8-5
Regulatory Guide 1.33, Quality Assurance Program Requirements. Criterion 1, Appendix A, 10 CFR 50, concerning General Design Criteria (GDC), requires , that a quality assurance (QA) program be established and implemented in order to provide adequate assurance that NPP structures, systems, and components will satisfactorily perform their safety functions. Appendix B of 10 CFR 50 establishes QA requirements for the design, construction, and operation of those structures, systems, and components that prevent or mitigate the consequences of postulated accidents that could cause undue risk to the health and safety of the public. Regulatory Guide 1.33 has been promulgated to describe acceptable methods for complying with the provisions of these appendices of the Code for the operational phase of NPPs. The most recent official issue of Regulatory Guide 1.33 is dated February 1978 and endorses ANSI N18.7-1976/ANS-3.2, Administrative Controls and ' Quality Assurance for the Operational Phase of Nuclear Power Plants. This I standard has since been updated and reissued as ANSI /ANS-3.2-1982. 4 Regulatory Guide 1.33 is also undergoing revision and several draft versions have been issued. The current proposed revision to RG 1.33 endorses ANSI /ANS-3.2-1982. 4 ANSI /ANS-3.2-1982 has appended a list of " typical procedures" for pressurized ' water reactors (PWRs) and boiling water reactors (BWRs). This list was taken from the February 1978 issue of Regulatory Guide 1.33. Section A6 of the - Appendix deals with maintenance procedures and A8 deals with calibration and test procedures. These, and an updated list of references, constitute the only substantive maintenance-related differences between ANSI /ANS-3.2-1982 and the NRC-endorsed ANSI N18.7-1976/ANS-3.2. ANSI /ANS-3.2-1982 contains one of the few treatments of NPP maintenance thus far considered for NRC endorsement. Perhaps the most important reference to maintenance is contained in paragraph 5.2.7.1, " Maintenance Programs," which states: s A maintenance program shall be developed to maintain structures, systems, and components important to safety at the quality required for them to perform their intended functions. The paragraph also requires maintenance planning and scheduling, the , development of maintenance procedures, malfunction evaluation and documentation, equipment deficiency reporting, and preventive maintenance (PM) of safety-related equipment and systems. Regarding the latter, the standard states: A preventive maintenance program including procedures as appropriate for structures, systems, and components important to safety shall be , established and maintained which prescribes the frequency and types of maintenance to be performed. The standard further states that a preliminary preventive maintenance program should be developed before fuel loading begins. Paragraph 5.3.5, " Maintenance Procedures," describes the general content categories for s maintenance procedures, which include: 1) preparation for maintenance, i
- 2) performance of maintenance, 3) post-maintenance check-out and return to service, and 4) supporting maintenance documents. Policy on the use of B-6 o
l' , maintenance procedures is addressed in paragraph 5.2.7, which states that written procedures and instructions will be available for maintenance or modification of equipmen? " appropriate to the circumstances." Section 5.2.7 also states that skills ilormally possessed by qualified maintenance personnel i may not require detailed step-by-step written procedures. Summary. -ANSI /ANS-3.2-198E provides the only compendium of NPP maintenance ! requirements in the regulatory lit.erature. The Regulatory Guide does not i contain measurable criteria by wnich to assess the adequacy of a maintenance program. Finally, the outdated version of Regulatory Guide 1.33 and the many subsequent versions of published ANSI QA standards adopted by the licensees
, c{eateaninconsistentbasisforregulationofoperatingreactors.
Regulatory Guide 1.8, Personnel Qualification and Training. Regulatory Guide 1.8, September 1975, reissued May 1977 endorsing ANSI N18.1-1971 prescribes that maintenance repairmen in NPPs have 3 years of experience in ons.or more crafts and a "high degree" of manual dexterity and ability, plus i the'" capability to learn." Technicians are required to have 2 years' working l exp5rience in their specialty and 1 year of related training. No specific training or requalification requirements were given for technicians or maintenance repairmen. ANSI /ANS has reissued the standard on Selection, Qualification, and Training for Personnel for Nuclear Power Plants - the latest published version being ANSI /ANS-3.1-1981. The 1981 version of ANSI /ANS-3.1 expands on the training requirements of the
- previous version of this standard
l A training program shall be provided for job functions that could affect the' quality of structures, systems, and components important to safety. i The program shall include applicable administrative controls, special complex system and component instruction, and demonstrated performance ( capability. The special training above the journeyman level of technician and maintenance personnel shall be based on a task analysis of the individual's assigned functions (par. 5.3.4). l The 1981 version of ANSI /ANS-3.1 is a more comprehensive and detailed l treatment of maintenance , training than are previous versions. Proposed Regulatory Guide 1.8, Revision 2, scheduled for publication in 1986, endorses ANSI /ANS-3.1-1981. Operating reactors and applicants currently subscribe to N one of three different versions of the industry standard, resulting in less than con:;istent industry maintenance training requirements. S w ary. Current NRC guidance on training of maintenance personnel is in a state of transition. Regulatory documents are outdated and endorse outdated Standards. Industry maintenance training programs are in a state of change while accreditation is being implemented. B-7
l 3.0 I&E INSPECTION PROCEDURES ACDRESSIfKi MAINTENANCE OF COMMERCIAL NUCLEAR POWER PLANTS A useful regulatory system requires a feedback loop to assess regulatory compliance, the efficacy of regulatory initiatives, and the need for additional guidance or modifications to current guidance. The I&E arm of the NRC is the primary means for providing this feedback. I&E inspection procedures (IP) for hPP maintenance are contained primarily in I&E Manual Chapters 62700-62703. The IPs covering maintenance activities adhere primarily to the quality assurance standard, ANSI N18.7-1976, that is current'y endorsed by the NRC. The IE Manual contains several IPs that dddress the maintenance area. IP 62700 is a performance-based procedure that requires the inspector to assess the licensee's implementation of the maintenance program. IP 62703 requires the performance of a monthly observation of maintenance on selected safety systems to ensure, on an on-going basis, that the licensee is adhering to his maintenance program. IPs 62704 and 62705 (I&C and Electrical Maintenance, respectively) require thorough observation of maintenance activities in each of these discipline areas from a technical perspective, as well as a review of documentation associated with the maintenance activities observed. l B-8
M o._....._.,0.._._ i..-T_......._._.-.,,
.t b :E',"E BIBUOGRAPHIC DATA SHEET NUREG-1212, Vol. I t LEE INSTRUCT *0Ns ON THE Rtb ER$t 2 T*[L[.Y $ustiTLE 3 LE.vf SL.NE Status of Maint nance in the U.S. Nuclear Power Industry 1985 f Volume 1: Findin and Conclusions f oos,~.o.ri...oafcou aTio f j ....
G. Cwalina, . Grenier, J. Jankovich, J. Koontz fMay 1986 N. Le, P. McL ughlin and D. Persinko I oo~,- = o.' e a * 'oa ' '55"o l 7 PER* opMiNG oRG. Nil.fiON %.wt .No w. L - 's .oDa t sS tt irsse le Com/
/ June PMoJECT T.sn wore LNe t hvW9ER 1986
~ Division of Human Factor Technology ' Office of Nuclear Reacto egulation ...s m a.~1 w .ia U.S. Nuclear Regulatory Co. mission Washington, D.C. 20555 2
,o s osso..so 0 0.s,z.1,0~ s... .so ..,L ~a .x u .o, Division of Human Factors Tech logy
,,,,,,,C, ,,.1 .,0.,
g Office of Nuclear Reactor Regul ion Technical Report U.S. Nuc1 ear Regulatory Commissi *a'"***"'o"~~~~~' Washington, D.C. 20555
,2 SUPPLE MENT.n, NoTt 5 13 . 578* aci (200 weeve or wu/ '
M This report presents the results and co lus' ns derived from activities performed under Phase I of the NRC (1aintenance and u eillance Program (MSP). Findings are z_ based on trends and patterns derived from erational data compiled by the NRC for the period 1980 through 1985, site surveys nducted at eight plants, and question-naires administered to NRC Resident Insp t s to characterize nuclear power plant
- maintenance programs and practices. The ac vities have shown that plant main-tenance programs and practices are high' vari le from plant-to-plant and are currently undergoing major changes. W le meas ed plant performance has improved 5 overall since 1980, the maintenance-r ted contr ution to reportable events and challenges to safety systems remains gh and is i reasing by some measures. The
- results of Phase I of the MSP confi ad a number o roblems in nuclear power plant maintenance which warrant further N and industry a ention.
.. wcuomr .~.< ,s s . . ..os oisca..Te s is v,.g,g1, maintenance and surveillance orogram (!1SP) nuclear industry Unlimited
'. SECURtTy CL.SSie sC. TION tin.s peper ..
..oisT,.aso,n naioTe s Unclassified IYM 9000!t) 1 Uncl.assified o ~uo . o, ..c,. .
k b
UNITED STATES mem rouarn etass natt NUCLEAR REGULATORY COMMISSION 'o8$d'y5' WASHINGTON, D.C. 20555 gas,H,o g., OFFICIAL BUSINESS PENALTY FOR PR:VATE USE. 8300 20 5507E877 1 1;N11y11gicg1 ADP-DIV Cr TICC POLILY 5 PUB 'GT FR-PDR 3, U g g s a-501 W A SH I NS T 0 *J OC 2C555 E 1 y _ _______. _}}