Equipment reliability process improvement and ...

Equipment Reliability Process Improvement and Preventive Maintenance Optimization Messaoudi Darragi*, Abdulnour Georges Université du Québec à Trois Rivières Cp.500, Trois Rivières, Qc, G9A5H7 Raynald Vaillancourt, Dragan Komljenovic, Michel Croteau Gentilly 2 Nuclear Generating Station, Hydro Quebec

Abstract The Gentilly-2 Nuclear Power Plant wants to optimize its preventive maintenance program through an Integrated Equipment Reliability Process. All equipment reliability related activities should be reviewed and optimized in a systematic approach especially for aging plants such as G2. This new approach has to be founded on best practices methods with the purpose of the rationalization of the preventive maintenance program and the performance monitoring of on-site systems, structures and components (SSC). A rational preventive maintenance strategy is based on optimized task scopes and frequencies depending on their applicability, critical effects on system safety and plant availability as well as costeffectiveness. Preventive maintenance strategy efficiency is systematically monitored through degradation indicators. Key words: Equipment, Reliability Process, Preventive Maintenance, Optimization 1 – Introduction At the beginning of 2002, the Management of the Thermal and Nuclear Production studied the possibility of increasing planned outages interval from 12 to 18 or 24 months at the nuclear power plant Gentilly-2 (G-2) [1]. For applying this decision, it is necessary to ensure that: - Risks for environment and population will not increase. - Safety is maintained at higher levels. - Forced outages, between planned outages, will not raise the number and/or duration. Such an increase in outages interval may have an impact on the management and planning of preventive maintenance activities (time directed tasks, predictive tasks and surveillance tests). In a mean and long term, negative impact may also affect equipment reliability due to the decrease in preventive maintenance workload. At the same time, the plant was involved in an equipment reliability improvement process defined by both the Canadian Nuclear Safety Commission CNSC (Reliability program S98) [2] and World Association of Nuclear Operators (AP-913) [3]. A WANO expert team has visited G-2 in April 2004 for conducting a survey based on the AP-913, and to ensure a support in the preventive maintenance optimization activities related to AP-913. These projects led to new requirements with regard to an optimization of the preventive maintenance program in order to meet plant reliability and availability objectives.

Correspondence: [email protected]

2 – Purpose and objectives The purpose of these projects is to define and validate an equipment reliability process focusing on two principal objectives: - The rationalization and the optimization of the preventive maintenance program: New preventive maintenance strategy should be based on rational task contents and frequencies depending on their applicability, cost-effectiveness, and especially critical effects on system safety and plant availability. - Monitoring on-site system, structure and components (SSC) reliability and performance through degradation indicators. Such equipment reliability process should fulfill all plant requirements. This process has to be integrated into the current reliability related processes without significant modification and/or increase in the workload. This process should conciliate between the equipment reliability process AP-913 suggested by WANO, nuclear industry best practices (Streamlined RCM Processes and the state-of-the-art maintenance optimization methods) [4], G2 processes and the S-98 regulatory requirements. 3 – G 2 Project Methodology The equipment reliability process and the preventive maintenance optimization are based on the following methodology: 1- Study the Preventive Maintenance Optimization in the nuclear industry: Make synthesis on methodology, methods and tools of reliability centered maintenance, Streamlined RCM and Preventive Maintenance Optimization processes applied in the nuclear industry. 2- Study the AP-913 equipment reliability process: Compare it to streamlined RCM processes and regulatory requirements (reliability program S-98 of the CNSC): The aim is to design a generic process compiling those knowledge sources. 3- Elaborate an Equipment Reliability Process for SSC at G2 including the Preventive Maintenance Optimization. 4- Carry out Gap Analysis between AP-913 (supporting all G2 Equipment Reliability Process activities) and G2 Reliability Related Processes. 5- Plan the implementation of G2 Equipment Reliability Process through the Preventive Maintenance Optimization Project. 4 – Preventive Maintenance Optimization For reducing equipment failures and reaching reliability targets, consequences of increased planned outages interval have to be controlled. Solution will be based on long term SSC reliability improvement. An ideal solution lies in building an adequate process to optimize maintenance and reliability. Reliability Centered Maintenance was the first concept to arise [5,6,7]. It consists in responding to 7 principal questions: 1- What are the functions and desired performance standards for the asset? 2- In what ways can the asset fail to deliver the functions and desired performance standards for the physical asset? 3- What are the root causes (failure modes) for each functional failure? 4- What are the effects of each failure mode?

Page 2 of 10

5- Does the consequence of each failure mode matter (therefore is it worth doing a task to prevent the failure mode?) 6- Can anything be done to predict or prevent the failure mode (What type of predictive or preventive task)? 7- What if we cannot predict or prevent the failure (Can we do a failure finding task for a hidden failure? Do we need to redesign the equipment or how we use the equipment? Is no scheduled maintenance the correct strategy?). The RCM process starts with the identification of the system limits, functions, functional failure, failure modes and causes, component criticality to conclude with preventive maintenance recommendations and tasks comparison. In a practical approach it is based on 3 essential steps: 1- Functional and dysfunctional analysis. 2- Failure mode and effects analysis. 3- Selection of maintenance strategies through logic tree analysis. Despite its rigor and consistency this process proved insufficiency to fit nuclear industry requirements (example: time and resources consuming, highly conservative environment). Studies were undertaken to streamline this process through new simplified approaches called "streamlined RCM". Functional and dysfunctional analysis step was simplified to focus only on system important functions. In fact, RCM process works out the FMEA for all the functions of the system independently of their importance, while in Streamlined RCM approach, only important functions are analyzed. In another hand, time used to conduct FMEA analysis for choosing the maintenance strategy was saved by the use of Maintenance Templates (automated process). Some approved streamlined RCM processes were experimented by the Electric Power Research Institute EPRI [4] such us: - Classical Streamlined RCM. - Criticality Checklist Streamlined Process. - Plant Maintenance Optimization. Finally, a new streamlined concept of Preventive Maintenance Optimization called PMO was experimented. PMO seems more like a concept rather than a unified formal process. This concept has changed over years and was adopted in different ways in the nuclear industry [8,9]. It is considered as one of the most viable methods for equipment reliability improvement. PMO is more streamlined using different personalized processes to fit real needs and requirements of the nuclear industry. While RCM is based on functional and systems analyses and built of a new preventive maintenance program, PMO concept is directed towards the analysis and optimization of the existing preventive maintenance program. Furthermore, it focuses only on dominant failures modes targeted by the existing PM tasks. In a weak preventive maintenance program, many failure modes are not addressed by PM tasks. A PM task could also be evaluated in accordance with: - Applicability, efficiency, intrusive risks and cost effectiveness.

Page 3 of 10

-

Interval adequacy. Scope: failure mode targeted (evident / hidden, critical, degradation mechanism)

The first phase in a PMO project consists in performing a SSC Criticality Analysis. The second phase focuses essentially on the following steps: - Make a data gathering and collect all pertinent information on a SSC (all PM activities, historical data, design manuals and vendor recommendations, etc.) - Group all PM tasks by component - Find out their targeted failure modes, causes and degradation mechanism or simply compare them to the component maintenance template. - Make the recommendations on tasks (addition, deletion, scope modification, interval increase / decrease) based on a logical and systematic approach ( compare to vendor recommendations and expert judgments, verify SSC historical data, emphasis on predictive tasks, etc). Consequently, PMO gained more popularity due to its simplicity of use and reduced costs. Project costs decreased significantly based on streamlining Functional and FMEA analysis. PMO reached its maturity with the initiation of project EPRI PM basis Database [10,11,12,13,14]. PM basis Database was developed in 1998 by EPRI. The team carrying out the project was built of EPRI consultants, maintenance experts and manufacturers. They compiled information on 60 generic components of 49 American nuclear power plant over 20 years of nuclear experience. The database recommends optimal PM tasks, their intervals according to: Criticality, Service conditions and duty cycle. It offers also an exhaustive generic database helping to analyze degradation mechanisms, stress factors, failure timing, discovering preventive opportunity, etc (Figure 1). Figure 1: Degradation Mechanisms and Maintenance tasks relationship Degradation location

Degradation mechanism

Degradation influence

Degradation progression

Failure timing

Maintenance task

5 – A-913 Equipment Reliability Process A whole Equipment Reliability process deals with preserving operational reliability level asymptotic to the intrinsic reliability level defined in the design stage of the equipment. However, PMO process demonstrated limits to achieve this ambitious objective. The lack is caused by overlooking the integration of several PM related activities essential to preserve high equipment reliability.

Page 4 of 10

In March 2000, World Association of Nuclear Operators (WANO) and Institute of Nuclear Power Operations (INPO) worked out a generic Equipment Reliability Process entitled AP-913 [3]. AP-913 is a process integrating 6 principal activities: - Scooping and Identification of Critical components. - Performance Monitoring. - Corrective Actions. - Continuing Equipment Reliability Improvement (analogous to a Preventive Maintenance Optimization). - Long –Term Planning and Life Cycle Management. - PM Implementation. Numerous North American nuclear power plants have been trying to implement this new process through the modification and the improvement of their own processes [15]. Regarding Canadian perspectives, AP-913 answers without additional efforts the greater part of S-98 reliability program requirements (CNSC). The AP-913 implementation was also facilitated through the use of EPRI PM basis Database. 6 – G2 Equipment Reliability Process The Gentilly 2 Nuclear Power Plant already possesses its reliability and maintenance program. However, based on new knowledge and tendencies presented above, the plant management wants to optimize these programs. This approach should allow to: - Rationalize and optimize maintenance activities. - Monitor Systems, Structures and components degradations before point of failure (inside PF interval). - Make a living equipment reliability process through internal and external experience feedback. A project is created and conducted to reengineer and provide a new process able to fit: - G2 reliability and maintenance needs. - WANO best practices requirements. - CNSC regulatory requirements. - State of the art knowledge in Preventive Maintenance Optimization in nuclear industry. The new G2 Equipment Reliability Process is composed of three principal activities based on: 1- Preventive Maintenance Optimization PMO. 2- Performance monitoring. 3- Living program The PMO process has been so far the first targeted activity by the management. The different steps of the PMO process were defined and the action plan was scheduled specially for I&C components in first. The approach adopted was as follow: Stage one: Efforts are focused on the optimization of the existing PM program. All equipments targeted are in the Preventive Master Equipment List (PMEL). After 20 years of operating experience at G2, PM activities are mature enough to address degradation mechanisms causing important functional failures.

Page 5 of 10

Stage two: Efforts are focused on equipments in the Master Equipment List (MEL) but not covered by PM activities. The aim of this phase is to make sure that all important equipments of this list are covered by appropriate PM activities. However, the following steps have to be previously completed: - Define"G2 Criticality Criteria" for components and align them to both AP-913 and "EPRI Criticality Criteria". - Work out a "Decisional Logic Diagram" for the "G2 PMO process". - Make a model for "G2 PM basis". - Create "As-Found Conditions" codes for equipments at G2. - Develop a database to support all those activities. Flowchart of this process is shown in figure 2. Some routine activities are performed in a systematic way every week as follows: 1- PM Tasks Priorization and Components Selection (1.1): Select PM bundles fixed in the schedule. Priories PM tasks based on "lowest interval criteria" to define components list of the week. Other criteria may be considered to close the list (example: consider identical components) in order to ensure a fast optimization and a quick win approach. 2- Data Collection (1.2): Collect information and data on selected components such as: PM tasks, historical data, vendor’s recommendations, EPRI Maintenance Template, general references (EPRI, WANO, SOER, etc). 3- Component Criticality Analysis (1.3): Evaluate components criticality and define list of Critical, Non Critical and Run to Failure (RTF) I&C components according to G2 criticality criteria. 4- PM Tasks Elimination (1.4): For RTF components eliminate PM tasks. 5- Service Conditions and duty Cycle evaluation (1.5): For Critical and Non Critical components evaluate Service Conditions and Duty Cycle using EPRI criteria [11]. 6- G2 Decisional Logic and PM Tasks Comparison (1.6): Apply G2 Decisional Logic (proper to PMO project scope): a. Make comparison (task interval and/or scope) between current G2 PM tasks and those proposed by EPRI PM Basis Database and other internal land external sources. b. Consult corrective and preventive work history done on the component for the last years (10 years for I&C). c. Consult general references (EPRI, WANO, SOER, etc.) if necessary. d. Make change on interval and / or scope of PM tasks based on recommendations such as EPRI Templates Maintenance, component history, vendors PM programs, external experience and expert judgements. Plan changes on PM tasks scopes in next step to accelerate the quick win process. e. Emphases on the use of predictive maintenance technologies if possible and priories such tasks for Non Critical Components to manage risk. Page 6 of 10

f. Add new PM tasks if necessary based on component historical data (dominant failure modes and degradation mechanisms experienced in the past at G2 and not targeted by PM tasks) and on external experience feedback (Serious failure modes and degradation mechanisms experienced elsewhere and not targeted by PM tasks at G2). 7- Process results documentation (1.7): Document all steps of the process and the results. 8- Action plans definition (1.8): Ensure definition of action plans and PM modifications to support implementation. The success of PMO process implementation is conditional to ensuring the two essential activities of the whole G2 Equipment Reliability Process: -

Performance Monitoring (2): Defining Monitoring Plans, Component Performance Monitoring through Degradation Indicators, Cross-Systems Analysis, Indirect Monitoring through Maintenance and Reliability Ratios.

-

PM implementation and "Living Program" (3): Follow-up action plans, Corrective actions and apparent/root cause analysis, equipment as-found conditions, equipment performance criteria adjustment and finally measure PMO process performance and associated profits.

Page 7 of 10

Figure 2: PMO Process Flowchart at G2

G2 Equipement Reliability Process 1-Preventive Maitenance Optimization (PMO)

1.1- PM tasks priorization and components selection

1.5- Service Conditions and duty Cycle evaluation

1.2- Data Collection

1.6- G2 Decisional Logic and PM tasks comparison

1.3- Component Criticality Analysis

1.7- Process results documentation

1.4- RTF components: PM tasks elimination

1.8- Action plans definition

2- Performance Monitoring

3- PM implementation and Living program

7 – Gaps analysis and PMO project action plan A gap analysis was carried out with WANO technical assistance for improving preventive maintenance and equipment reliability. This analysis allowed to: - Evaluate G2 reliability related processes according to AP-913 activities. This action helped to confirm the steps defined by the G-2 Generic Equipment Reliability Process. - Determine the action plan for the PMO process: I&C components are first analyzed because they are considered as a potential source of improvement . PMO project is synchronized with Maintenance Schedule and begins with more frequent PM tasks (weekly PM tasks first) two meetings per week. Principal project elements such as team structure, schedule and process performance indicators are as follows: Page 8 of 10

Team structure - Process owner: Technical services. - System engineers: Concerned by PM scheduled in 16 weeks. - Reliability engineer (safety aspect): PMO coordinator. - Reliability engineer (maintenance aspect): Maintenance advisor. - Component engineer: Component expert. - Maintenance supervisor: I&C Supervisor. - Maintenance scheduler. Schedule - At T-16:  System engineers identify criticality (AP-913) and technical basis.  Reliability identifies criticality (S-98) and Maintenance Templates.  Maintenance reviews files (history) and scope.  Maintenance scheduler reviews criticality. - At T-15: PMO team meeting: review and recommendations. - At T-14: Risk decision by Technical Service Manager. - At T-13: System engineers review PM tasks and updates Computerized Maintenance Management System (SGE & SIE). - At T-12: Modify schedule, if frequency reduced or PM task removed (RTF). - At T-0: Execute I&C PM tasks and document As-found condition. - At T+1: Feed back for As-found Condition (next PMO team meeting). Process performance indicators - Number of man-hours spent on PM (monthly, annualized). - Number of man-hours spent on corrective maintenance (for non R-T-F). - Number of PM task optimized, dropped, added, frequency reduced, frequency increased. - Review nuclear safety comity indicators. - Review improvement comity indictors to ensure reliability and nuclear safety. 8 – Conclusion G-2 has defined its specific Equipment Reliability Process witch will help optimizing the preventive maintenance program especially after increasing outage intervals. This Equipment Reliability Process is based on the integration of a broad range of equipment reliability activities. It has the ability to discover equipment degradations “As-found equipment conditions” before the failure occurrence, to rationalize and optimize preventive maintenance activities and to monitor equipment and system. All those activities were integrated into the living program, which is constantly updated through internal and external experience feedback. In general, the use of an Equipment Reliability Process to optimize preventive maintenance is considered very important for aging nuclear power plants such as G 2 and for the nuclear industry in general. The need to measure, maintain and optimize equipment reliability led quickly to a world tendency in general and particularly in the North American Nuclear Industry. This tendency was confirmed by the elaboration of generic processes (AP-913 of World Association of Nuclear Operators), best practices methods (Streamlined RCM and Preventive Maintenance Optimization projects),

Page 9 of 10

confirmed tools (EPRI PM basis Database) and regulatory requirements (Reliability Program S-98 of CCSN). In conclusion, the use of EPRI PM Database will help to evaluate results and will be an essential tool to implement in the future in a practical way, such a process. References 1- Vaillancourt R., D. KOMLJENOVIC, Assessment of the impact on safety with regard to change in outage interval from 12 to 18 or 24 months at gentilly-2 nuclear generating station, CNS 2003. 2- Commission Canadienne de Sûreté Nucléaire, Norme d'application de la réglementation S-98, Programme de fiabilité pour les centrales nucléaires, Ottawa, 2001. 3- INPO, Equipment Reliability Process Description, AP-913, Rev. 1, March 2001. 4- EPRI, Comprehensive Low-Cost Reliability Centered Maintenance , EPRI TR – 105365 September 1995. 5- Moubray J., Reliability Centered Maintenance, Industrial Press, April 1997. 6- Zwingelstein G., La Maintenance basée sur la fiabilité - Guide pratique d'application de la RCM, HERMES, 1996. 7- Messaoudi D., G. Abdul-Nour, La maintenance basée sur la fiabilité, Université du Québec à Trois-Rivières, April 2003. 8- Laurence J. (Fractal Solutions, Inc), Improving Equipment Reliability and Plant Efficiency through PM Optimization, 1998. 9- Turner S. (OMCS International), Introducing PM Optimization: A Tool for Getting Design and Maintenance Right First Time. 10- EPRI, The EPRI PM Basis Database, Version 4, EPRI TR-106857, November 1998. 11- EPRI, PM Basis Version 5.0 with Vulnerability Analysis Module, EPRI TR1009275, 2003. 12- EPRI, The EPRI PM Basis Database: User Manual, EPRI Product 100148, January 2001. 13- EPRI, Reliability and Preventive Maintenance: Balancing Risk and Reliability: For Maintenance and Reliability Professionals at Nuclear Power Plants, EPRI Product 1002936, 2002. 14- EPRI, “Guide for Predicting Long-Term Reliability of Nuclear Power Plant Systems, Structures and Components”, EPRI, Palo Alto, CA, and Wolf Creek Nuclear Operating Company, Burlington, KS: 2002.1002954. 15- EPRI, AP-913 Industry Capabilities Gap Analysis Results, EPRI Product 1003478 Palo Alto, 2002.

Page 10 of 10