Resolution of Generic Safety Issues: Task I.E: Analysis and Dissemination of Operating Experience (Rev. 3) ( NUREG-0933, Main Report with Supplements 1–34 )
The objective of this task was to establish an integrated program which involved participation by the licensees, vendors, NSAC, INPO, and the NRC, and which included foreign operations experience for the systematic collection, review, analysis, and feedback of operating experience to NRC licensing, inspection, standards, and research activities, and to licensees for all NRC-licensed activities. Appropriate corrective action was expected to be taken in response to feedback.
ITEM I.E.1: OFFICE FOR ANALYSIS AND EVALUATION OF OPERATIONAL DATADESCRIPTION
The purpose of this TMI Action Plan48 item was to establish an NRC office which would be responsible for: (1) analysis and evaluation of operational data associated with all NRC-licensed activities; and (2) development of specific recommendations for action by other NRC offices.
Systematic evaluation of operating data can identify potential significant safety problems or their precursors. Dissemination to NRC and industry of evaluation results which identify such problems, along with recommendations for their resolutions, can avoid occurrence of these problems at other plants of similar design. This item was initiated to improve the NRC capability to make independent assessments of safety and, therefore, was considered a Licensing Issue.
The Commission approved the establishment of AEOD in July 1979 and an interim office was established in October 1979. As of June 1984, AEOD was staffed and functioning in accordance with the purpose and scope described in Chapter 0143 of the NRC Manual. Thus, this Licensing Issue was resolved.
ITEM I.E.2: PROGRAM OFFICE OPERATIONAL DATA EVALUATION
The purpose of this TMI Action Plan48 item was to ensure that each NRC office conducted operational safety analyses. These analyses were to be coordinated and the results distributed as part of the integrated program on operating experience assessments. The work of each office was expected to complement the operational data evaluation activities conducted by AEOD under Item I.E.1.
Systematic evaluation of operational data can identify potential significant safety problems or their precursors. Dissemination to NRC offices and industry of such evaluations, along with their resolution, can avoid occurrence of these problems at other facilities of similar design that conduct similar operations. This item was initiated to improve the NRC capability to make independent assessments of safety and, therefore, was considered a Licensing Issue.
As of June 1984, each of the following NRC Offices had established responsibility and procedures for evaluating operational data:
(1) In OIE, the Events Analysis Branch had the lead responsibility for this activity using input from such sources as LERs, Preliminary Notices, 10 CFR 21 Reports, 10 CFR 50.55(e) Construction Deficiency Reports, and 10 CFR 50.72 Reports to the NRC Operations Center. Evaluations which identified potential significant safety problems were disseminated by means such as IE Notices or IE Bulletins.
(2) In NRR, the Operating Reactors Assessment Branch (ORAB) conferred daily with OIE on operational occurrences and made preliminary assessments of their safety significance as part of their functional responsibility described in the NRC Manual. Those occurrences which were considered to have potential safety significance were identified at weekly NRR management briefings on operational events conducted by ORAB with OIE participation. If deemed necessary, further evaluation was assigned to the appropriate NRR Division.
(3) In RES, the Reactor Risk Branch had lead responsibility for evaluating operational events and was responsible for issuance of periodic reports on Precursors to Potential Severe Core Damage Accidents,76 which were derived from a systematic review of LERs. These reports provided operational experience data and were available for use in Event Tree Analyses and PRAs conducted by RES.
(4) NMSS had issued a procedure305 for achieving a more formal review and evaluation of inspection and operational data and event reports to identify and correct generic problems. The procedure included criteria for identifying operational events that warranted detailed review and evaluation. Evaluation reports that identified safety significant operational events were distributed within NRC.
Based on the actions taken by the cognizant Offices, this Licensing Issue was resolved.
ITEM I.E.3: OPERATIONAL SAFETY DATA ANALYSIS
The purpose of this TMI Action Plan48 item was to conduct special operational safety data analyses to determine equipment failure rates and to develop error data analysis for nuclear plant operations. As of June 1984, the Reactor Risk Branch of RES was performing studies to: (1) determine equipment failure rates using LERs; (2) develop and use common-cause/common-mode analysis of LERs; (3) analyze data from the NPRDS to distinguish order-of-magnitude differences in component failure rates between such factors as plants, sizes, service environment, status at time of failure, and manufacturer; (4) identify potential reliability problems evident in the LER data; and (5) identify potential accident precursors.
The information obtained from this item was used to: (1) provide more reliable equipment failure rate data, including common-cause/common-mode failure statistics to support PRAs of nuclear power plants (see Items II.C.1, "Interim Reliability Evaluation Program," and II.C.2, "Continuation of IREP"); and (2) to identify potentially serious equipment reliability problems evident from LER data and provide feedback to equipment maintenance/surveillance programs to reduce equipment failure rates (see Item II.C.4, "Reliability Engineering"). This item was initiated to improve the NRC capability to assess safety and, therefore, was considered a Licensing Issue.
At the time of this evaluation, implementation of the program had resulted in the following:
(1) Publication of data summaries of LERs on pumps, control rods and drive mechanisms, diesel generators, valves, primary containment penetrations, and instrumentation and control components (See References 344, 345, 346, 348, 349, 350, and 351).
This item was an ongoing effort to collect and analyze data. While no quantified safety benefits could be directly assigned to it, the benefits occurred as the results of the equipment failure rate data and reliability analysis were used in assessing other specific related safety issues, including Items II.C.1, II.C.2, and II.C.4. The ongoing activities were to be conducted as described in Section 16.1 of the NRC Long Range Research Program.133 Thus, this Licensing Issue was resolved.
ITEM I.E.4: COORDINATION OF LICENSEE, INDUSTRY, AND REGULATORY PROGRAMS
The purpose of this TMI Action Plan48 item was to ensure coordination of licensee, industry, and NRC programs for evaluating plant operating experience. As part of the implementation of NUREG-0737,98 licensees established the capability for evaluating plant operating experience and procedures for providing feedback of the information to operations personnel and for incorporating it into training programs, in accordance with Items I.A.1.1, I.B.1.2, and I.C.5. Industry evaluation programs were to be conducted by INPO. AEOD was responsible for coordinating the NRC programs for evaluation of operational data with those of licensees and industry.
Licensee evaluations of plant operating experience, coordinated with industry and NRC evaluations using common data bases, were to ensure that licensee, industry, and NRC corrective action recommendations were properly coordinated and applied. Effective feedback of prioritized and analyzed event descriptions to plant operating personnel and incorporation into training programs can avoid occurrence of these problems at other plants of similar design. This item was initiated to improve the NRC capability to assess safety and, therefore, was considered a Licensing Issue.
The results of industry and NRC operating experience evaluations are shared under an NRC-INPO Memorandum of Agreement238 initially signed in June 1981 and revised in April 1982. Thus, this Licensing Issue was resolved.
ITEM I.E.5: NUCLEAR PLANT RELIABILITY DATA SYSTEM
This item was evaluated with Item I.E.6 below and resolved.
ITEM I.E.6: REPORTING REQUIREMENTS
Items I.E.5 and I.E.6 were combined and evaluated together.
The objectives of these TMI Action Plan48 items were to: (1) determine if there was a need to make licensee participation in NPRDS mandatory; and (2) establish improved reporting requirements for operating reactors.
NPRDS is a reliability-oriented data collection and reporting system for selected components and systems related to the safety of nuclear power plants. Periodic reports containing failure statistics are issued and licensee participation is voluntary. Improvements in reporting significant events of operating plants can identify potential significant safety problems or their precursors and can avoid occurrence of these problems at other plants of similar design. Improved reporting of system/component reliability data will increase the validity of operating experience assessments and PRA programs. This item was initiated to improve the NRC capability to assess safety and, therefore, was considered a Licensing Issue.
By affirmation of SECY-81-494,260 the Commission endorsed the following staff actions to resolve these issues: (1) develop a proposed rule to modify and codify the existing LER requirements and to assure consistency with 10 CFR 50.72 which covers the immediate reporting of significant events; (2) endorse the INPO plan to assume responsibility for the management, funding, and technical direction of NPRDS; (3) coordinate with INPO to minimize duplication between the LER and NPRDS systems and between subsequent NRC and INPO analysis of NPRDS data; (4) obtain INPO assurance that NPRDS receives, processes, and disseminates the reliability data needed by industry and the NRC to support PRA programs; and (5) monitor (by AEOD) INPO's management of the NPRDS and provide the Commission with semi-annual reports on the effectiveness of INPO management of NPRDS.
As of January 1982, INPO had assumed responsibility for the NPRDS and the NRC was represented on the NPRDS Users' Group and participated in various NPRDS work groups. AEOD submits semi-annual reports to the Commission on the effectiveness of the INPO management of NPRDS. A proposed rule on LERs was published in the Federal Register (47 FR 19543) on May 6, 1982, and the final rule597 was published in July 1983. Based on the actions described above, Items I.E.5 and I.E.6 are Licensing Issues that were resolved.
ITEM I.E.7: FOREIGN SOURCES
The purpose of this TMI Action Plan48 item was to supplement domestic operating experience of safety significance by obtaining operating and design information from foreign reactors. To obtain foreign experience in a more systematic manner, the Office of International Programs (IP) participated with nuclear regulatory agencies of other nations in a centralized exchange of incident information with the Nuclear Energy Agency (NEA); this NEA exchange was initiated in 1980. Supplementing this effort was the upgrading of information exchange on significant incidents through direct contact and correspondence between the NRC and its bilateral partners, and by additional formal bilateral information exchange agreements which were concluded or renewed in 1981 and 1982.
AEOD also sponsors a program by which the Nuclear Operations Analysis Center at ORNL screens and stores for ready access reports of foreign reactor incidents and provides monthly summaries of these events that are potentially significant and relevant to U.S. LWRs. This item was initiated to improve the NRC capability to assess safety and, therefore, was considered a Licensing Issue.
As of June 1984, foreign reactor incident and operating experience reports were being routinely received and disseminated to NRC technical staffs. IP was also routinely sending these foreign reactor incident reports to INPO for use by industry in evaluating plant operating experience under Item I.E.4. Foreign reactor incident and operating experience reports were being assessed by AEOD and affected NRC Offices, as described in Items I.E.1 and I.E.2, respectively, to identify potential significant safety problems or their procedures which may be applicable to U.S. plants. Dissemination within NRC and to industry of such assessments, along with their resolutions, can avoid occurrence of these problems at other facilities of similar design. Thus, this Licensing Issue was resolved.
ITEM I.E.8: HUMAN ERROR RATE ANALYSIS
The TMI-2 incident increased concern for the effect of human reliability on reactor safety. The lack of human reliability data applicable to nuclear power plants compared to hardware reliability data highlighted this concern in nuclear safety assessments and regulation. The purpose of this TMI Action Plan48 item was to continue research to: (1) complete analysis of field-collected data for human reliability in maintenance and calibration activities at operating nuclear power plants; (2) review abnormal occurrence reports, licensee event reports, and compliance reports to identify areas where human performance reliability is low; (3) develop probability models to predict error rates for multiple human errors occurring as a function of coupling influences; and (4) identify patterns and basic associative factors for the human-error rates determined for basic test, maintenance, and operator actions.
The information obtained from this item was to be used to: (1) identify necessary improvements in operator training and training aids to reduce human error rates (see Items I.A.2.6, "Long-Term Upgrading of Training and Qualification of Operating Personnel," and I.A.4.2, "Long-Term Training Simulator Upgrade"); and (2) provide quantifiable human error data and models to support PRA of nuclear power plants [see Items II.C.1, "Interim Reliability Evaluation Program" (IREP), and II.C.3, "Continuation of IREP"]; and (3) provide human engineering criteria for evaluating the design of new or modified systems.
While no quantified safety benefit could be directly assigned to this item, the benefits were expected to occur as the results of the human error rate and reliability analyses were used in assessing other individual related safety issues, including TMI Action Plan Items I.A.2.6, I.A.4.2, II.C.1, II.C.2, and Task Action Plan Item B-17. Therefore, this item was considered a Licensing Issue.
The Human Factors Branch of RES implemented an expanded Human Reliability research program to accomplish the purpose of this item and provide the human error information for its end use as described above. Major reports issued included: (1) a human reliability data bank338; (2) a draft handbook for human reliability analysis339; (3) procedures for estimating human error probabilities341,342; and (4) a workbook for conducting human reliability analysis.343 Future work included finalizing the handbooks, workbooks, and reliability models and maintaining the data bank; this work was described in Section 7.1 of the NRC Long Range Research Plan.133 Thus, this Licensing Issue was resolved.