Preventive Maintenance Basis Project Overview Report Update TR-106857- R1 Final Report, November 1998

EPRI Project Manager John P. Gaertner

EPRI • 3412 Hillview Avenue, Palo Alto, California 94304 • PO Box 10412, Palo Alto, California 94303 • USA 800.313.3774 • 650.855.2121 • [email protected] • www.epri.com

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THIS REPORT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS REPORT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS REPORT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS REPORT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS REPORT. ORGANIZATION(S) THAT PREPARED THIS REPORT Applied Resource Management G&S Associates

ORDERING INFORMATION Requests for copies of this report should be directed to the EPRI Distribution Center, 207 Coggins Drive, P.O. Box 23205, Pleasant Hill, CA 94523, (925) 934-4212. Electric Power Research Institute and EPRI are registered service marks of the Electric Power Research Institute, Inc. EPRI. POWERING PROGRESS is a service mark of the Electric Power Research Institute, Inc. Copyright © 1998 Electric Power Research Institute, Inc. All rights reserved.

CITATIONS This report was prepared by Applied Resource Management 313 Nobles Lane Corrales, New Mexico 87048 Principal Investigator D. H. Worledge G&S Associates 9420 Holly Bend Lane Huntersville, North Carolina 28078 Principal Investigator G. R. Hinchcliffe This report describes research sponsored by EPRI. The report is a corporate document that should be cited in the literature in the following manner: Preventive Maintenance Basis: Project Overview Report Update, EPRI, Palo Alto, CA: 1998. Report TR-106857-R1.

iii

REPORT SUMMARY A variety of preventive maintenance (PM) tasks are being implemented at different frequencies at U.S. nuclear plants for nominally the same equipment, with little documented basis to support the tasks, or their intervals beyond fundamental vendor manual information. The Preventive Maintenance Basis project was developed in response to the requests of the EPRI nuclear membership to provide a technical basis and recommendations for the assignment of PM tasks that go beyond the available vendor information. Background A large fraction of U.S. nuclear plants is in the process of reducing preventive maintenance (PM) costs and improving equipment performance by more closely matching PM tasks with the functional importance of equipment. For this to succeed, utilities require information on the most appropriate tasks and task intervals for the important equipment types, while accounting for the influences of duty cycle and service conditions. This data has not previously existed in an accessible form, often resulting in arbitrary and unsuitable tasks and intervals that increase maintenance costs and diminish reliability. Objectives The Preventive Maintenance Basis will serve the nuclear utility maintenance community as an essential reference for preventive maintenance task selection on common major components. The PM Basis reports contain a database of maintenance task contents, task interval recommendations, and a synopsis of the associated technical basis for 39 major components. The technical basis states the reasons for each task and the relationship between the equipment’s failure locations, failure mechanisms, the influences on equipment degradation, and the timing of failures. Approach Expert panels composed of knowledgeable individuals from EPRI, EPRI member utilities, and manufacturers were employed to formulate the bases and range of PM task options presented for the selected equipment. Most of the expert panels addressed a small number of closely associated component types, e.g. three types of pressure relief valves. A utility oversight committee, the PM Basis steering committee, was established to oversee and direct this process. While the inclusion of any equipment-specific recommendations was made by the individual expert panels, the steering committee maintained purview over the project structure and process, selected and prioritized component types, assisted with expert panel member recruiting, and approved the methodology employed for establishing the PM tasks, v

task intervals, and rationales. The PM Basis project steering committee was also accountable to the EPRI Operations and Maintenance Cost Control Target Steering Committee for ensuring all utility needs were considered and met. Results This report contains an overview of the objectives, project organization, and the process employed to develop, describe, and use the 39 component type PM programs and supporting technical bases. Each component type report is individually presented as a separate volume of this series. EPRI Perspective The PM Basis Reports is one of numerous related EPRI products to support PM optimization at nuclear plants. Other products include Reliability Centered Maintenance (RCM) demonstrations and methods, RCM assessment software, methods for identifying predictive maintenance opportunities, integration of PM with plant work processes, and methods for Maintenance Rule implementation. A design requirement of the PM Basis project was that the results be consistent with and complementary to these other inter-related products. EPRI further required that the PM Basis Reports must have credibility among nuclear plant maintenance engineers. Therefore, the unique approach for this project was crafted; that is, using a highly structured facilitation to enable small groups of utility component- experts to develop PM tasks, frequencies, and their bases. Not only was this approach designed to produce widely accepted results; but alternative approaches—including engineering analysis of failure modes, failure data analysis, or industry consensus surveys—were considered to be prohibitively expensive. Wide acceptance and use of the PM Bases to date have borne out the success of the approach. Also, EPRI believes that the depth and breadth of the supporting information on degradation mechanisms enables the PM Basis reports to be used for a wide range of engineering applications. Many of these applications have already been demonstrated by utilities. These and others are discussed in this overview report. EPRI is committed to continue support of PM Bases applications to reduce maintenance costs, increase plant performance, and maintain safety and efficient regulatory compliance. TR-106857-R1 Interest Categories Nuclear component reliability Nuclear plant operations and maintenance Keywords Preventive Maintenance Maintenance optimization Component reliability Power plant reliability vi

ABSTRACT

Preventive maintenance (PM) programs in US nuclear plants have evolved from strict compliance with the vendor’s general recommendations which are likely to be conservative, to more flexible tasks which are intended to accommodate the plant service conditions. These PM programs have evolved piecemeal from the vendor recommendations and historical experience, with the result that their logical structure and coherent basis may be technically weak and scarcely recognized. The historical reasons for the performance of specific PM tasks are usually poorly documented, if at all. The result is that relationships and dependencies between the tasks, their limits of applicability and justifications for task intervals, are not well known. Utilities have expressed the desire for a technical basis and rationale capable of supporting the maintenance tasks for each major component type and of facilitating appropriate changes to them. The PM Basis project provides the utility user with the technical basis for PM tasks and task intervals and also gives the necessary information to adapt the tasks and intervals to plant conditions. A recommended program of PM tasks, a synopsis of the task content and intervals, and the reasons why these choices are technically valid in a variety of circumstances, are presented for 39 major component types. Each component type is treated in a separate volume of this series. This project overview report presents the project objectives and organization, describes the content, interpretation, and use of the PM Basis database, and explains the process by which the information was obtained.

vii

ACKNOWLEDGMENTS

EPRI and the authors gratefully acknowledge the input, guidance, and support provided by the PM Basis project Utility Steering Committee and the members of each component expert panel. The steering committee was comprised of the following utilities: x

John Arnold, Commonwealth Edison Company

x

Edwin Rogers, Entergy Operations, Inc.

x

Mark Forsyth, Houston Lighting and Power Company, chairman

x

Kevin Higgins, Northeast Nuclear Energy Company

x

David Rollins, Omaha Public Power District

x

Brain Ferguson, Ontario Hydro

x

Howard Arnold, Pacific Gas and Electric Company

x

Rod Sorrell, Texas Utilities Electric Company

x

Bruce Boyum, Washington Public Power Supply

ix

EXECUTIVE SUMMARY

Need and Objective Utilities have expressed the desire for a technical basis and rationale capable of supporting the maintenance tasks for each major component type. The PM Basis project provides the utility user with the technical basis for PM tasks and task intervals and gives information to adapt the intervals to plant conditions. A recommended PM program, a synopsis of the task content and intervals, and the reasons why these choices are technically valid in a variety of circumstances, are presented for 39 major component types. The project had two major objectives. The first was to summarize a sample of industry experience on which tasks and task intervals comprise a sound, cost-effective PM program for a large number of major component types. The second objective was to outline the relationships between a component’s degradation mechanisms and the factors that influence them, the time scale of progression to failure, and the opportunities to discover and prevent them, so that utilities can adapt the programs to plant conditions. The PM Basis Product PM Basis reports for the 39 component types is available as a series of volumes under the TR-106857 report number. All of these reports have been completed at the time of writing this overview report. The overview report is presented as a master report for the entire project to detail the process by which the PM Basis was developed, and to outline its potential and current uses. The PM Basis was tasked by the steering committee to be a body of information that supports and includes an interpretive summary of utility power plant experience on preventive maintenance for each component type. Each report should contain, as much as practical in one place, all the PM tasks, task intervals, task rationales, and the most important influences on equipment degradation and maintenance for that specific component. This body of information should provide additional context and perspective when plant personnel are seeking to interpret vendor requirements in the xi

light of plant specific conditions, or when they need to understand the motivations for and limitations of PM tasks when modifying task content or extending task intervals. In many instances the information can be used in concert with input from the manufacturer to understand how vendor requirements might be relaxed in specific circumstances. The information in the PM Basis reports is intended to complement, not to replace the PM instructions in vendor manuals. Project Organization To promote utility buy-in and support, a utility oversight committee, the PM Basis project steering committee, was established. The steering committee was comprised of the EPRI project manager, a utility chairman, and 10 other utility members supported by the project staff. The steering committee, through the EPRI project manager, was also accountable to the Utility Advisory Committee for ensuring all utility needs were considered and met. The steering committee was granted purview over the structure of the PM Basis project, the prioritization and selection of the component types to be analyzed, the composition of the expert panels, and the methodology employed for the development of the component PM rationales. For each component type an expert panel of utility component experts provided the raw data on equipment degradation and the range of PM task options for the selected equipment. Each panel, as far as practical, was composed of knowledgeable individuals from EPRI member utilities, manufacturers, EPRI, and others. Most of the expert panels addressed a small number of closely associated component types, e.g. three types of pressure relief valves. The panels developed the preferred PM practices and raw material on equipment degradation in tabular form. The contractors developed the task rationale from all the data supplied, and submitted the component report to the expert panel for comment and approval. Process To construct the rationale for a PM program it has been found that a large amount of very specific information is required. It was essential that the meetings of component experts were used to maximum efficiency to provide this information in a short time. For this reason the meetings quickly became highly structured, closely following a multi-step process to ensure disciplined coverage of all the required aspects. Basic data obtained from the expert panels consisted of the hardware locations that are the sites of degradation and failure, the kind of degradation experienced and the main factors that influence it, the time development of the degradation processes and failures, the means to detect equipment condition and to intervene, the higher level PM tasks that would be used to implement these measures, and the task content. Supplementary information was added such as examples taken from nuclear power plant systems, definitions of applicable duty cycle and service conditions that influence xii

task intervals, a discussion of the risk to reliability of doing intrusive maintenance, the most common or dominant failure locations and mechanisms taken from industry sources, the principle focus of each PM task, and the availability of design modifications that can improve reliability or decrease dependence on preventive maintenance. The most important elements of the process for developing the rationale for each task are shown in the flow chart of Figure 1. PM Task Rationale When the basic data had been obtained and documented it was screened in various ways to discover the more common types of degradation and failures addressed by each task, which of them is most responsible for the timing of the task, the tasks that are aided by other PM tasks which also address the same failure locations and conditions, and the tasks that are most relied on because they do not have a backup from other tasks. This analysis also identifies a logical time interval for each task that is determined by the time scales of occurrence of the failures which are addressed. The technical basis for each task is presented under the following three headings: Failure Locations and Causes: This section states the failure locations and degradation mechanisms that the task is designed to deal with, generally with the most commonly encountered situations described first. This section provides an overview of the main focus of the task. Progression of Degradation to Failure: This section summarizes the information on times of occurrence of the dominant failures. It shows whether the failures should be expected to be predominantly random in nature and indicates the time scale over which the chance of a failure should become appreciable, or whether a wearout behavior is anticipated, and the relevant time period that is expected to be free of failures. It must be emphasized that this is a synthesis of many contributing processes over all the failure locations addressed by the task. This section also shows which of the failure locations and mechanisms is most likely to cause trouble if the interval is extended too far. Fault Discovery and Intervention: This section explains remaining aspects of the choice of one task over another, the interaction among tasks and the reconciliation of the expert panel’s choice of intervals for the task with the contractor’s analysis of the basic data that the expert panel provided. Despite the uncertainties in the data, the result was almost always a clear validation of the judgment of the expert panel in the limiting intervals that they xiii

assigned to the task. This result is a significant validation of industry PM assumptions and programs. The PM Template Each PM Basis report includes a PM Template which summarizes the program of tasks and task intervals for the equipment type. The program displayed in the Template is a technically defensible PM program, but it may not be the optimum for a particular plant. Each plant should base its PM program on appropriate vendor recommendations and its own history of preventive and corrective maintenance. For a plant that already has a PM program that is based on its own history, the Template can serve as a baseline for comparison, and the rationale section will probably indicate why their program is appropriate, or if it may not be appropriate in some aspects. For a plant that does not have an extensive operating history with a particular component type, the Template can be used in conjunction with vendor recommendations directly as a default program, with gradual changes anticipated as information is fed back in the future from a living program. Maintenance Risk and Correlations with Reliability The term maintenance risk is used to represent effects that accompany preventive maintenance that tend to increase failure rates rather than to decrease them. The point was demonstrated during the early part of this project when utility data from nuclear power plants on Air Operated Valve (AOV) reliability was correlated with the PM tasks being performed on them. A significant negative correlation was discovered between the frequency of intrusive preventive maintenance (i.e. internal inspections, parts replacement, and overhauls) and the reliability of the valves. Members of the expert panels were therefore routinely asked whether or not the component type being considered presented a significantly higher risk of further failures if intrusive preventive maintenance was performed too frequently. Their responses provided input to a discussion of the level of conservatism represented by the task intervals, and the kinds of failures they had experienced with the equipment that were attributable to maintenance error. Uses of the PM Basis The information in the PM Basis reports was originally intended to be used during large scale updating of a whole plant PM program such as an RCM analysis. In this application it would make task selection more efficient, promote consistency between analysts, and support justifications for relinquishing vendor recommendations or for making other changes. It could also be used when modifying any individual PM task as a result of changes in performance, modifications to equipment, or equipment aging. xiv

EPRI has developed quick vertical slice audits of utility PM programs carried out by sampling equipment in two or three systems at a plant. This provides a snapshot in a few days of how the plant PM program compares to industry practice as exemplified in the EPRI PM Basis. The PM Basis may also find use for validating corrective actions and goals in the maintenance rule when these consist of changes to preventive maintenance. Important aspects could be the reasons the task or interval change is the correct response, and the PM activities the task should include. Many other uses of this information have been suggested by utilities including the systematic optimization of PM task intervals, the evaluation of the relative benefit obtained by performing a particular task, and the development of a repository of industry experience on equipment degradation and PM effectiveness as an updatable electronic database containing guidance for its use in applications.

xv

Figure 1 Flow Chart: PM Development Process

Hardware locations that degrade or fail 1

Actions that could detect or prevent the condition 3

Degradation mechanisms for each location 4

PM tasks for each location that include the above actions 6

Degree of reliance on each task 7

Failure locations covered by each PM task 9

Principle motivation for performing each task 10

Duty cycle aspects that affect task timing 12

Outline of PM task content

Service conditions that affect task timing 15

Task intervals for critical/duty cycle/service conditions 16

13

Influences on the degradation that start or promote it 2

Timing features of each degradation mechanism 5

Most likely failure causes 8

How long it takes before the failure point is reached 11

Time constraints on the task intervals 14

CONTENTS

REPORT SUMMARY ..................................................................................................... v ABSTRACT ................................................................................................................. vii ACKNOWLEDGMENTS............................................................................................... ix EXECUTIVE SUMMARY .............................................................................................. xi CONTENTS ................................................................................................................ xvii LIST OF FIGURES ...................................................................................................... xix 1

INTRODUCTION .....................................................................................................1 1.1

Background .........................................................................................................1

1.2

Objectives ...........................................................................................................2

1.3

The PM Basis Reports and Their Interpretation..................................................2

1.4

Definition of Preventive Maintenance..................................................................5

1.5

Major Components..............................................................................................6

1.6

Project Organization............................................................................................7

1.7

Component Expert Panels ..................................................................................8

1.8

Project Learning Curve .......................................................................................8

References...................................................................................................................9 2

USE OF THE EPRI PM BASIS.............................................................................11 2.1

PM Optimization................................................................................................11

2.2

PM Audits..........................................................................................................12

2.3

Evaluation of the Benefit of PM Changes..........................................................13 xvii

2.4

Improving The Use And Integration of Predictive PM Tasks.............................14

2.5

Craft Information Feedback...............................................................................15

2.6

Electronic Database Of The EPRI PM Basis Data ............................................15

3

PROCESS FOR DEVELOPING THE PM RATIONALE ........................................17 3.1

Overview ...........................................................................................................17

3.2

The Expert Panel Meeting.................................................................................18

3.3

Failure Locations...............................................................................................25

3.4

Degradations and Influences ............................................................................28

3.5

Time Information for Degradation and Failure ..................................................31

3.6

Discovery Opportunities and PM Strategies......................................................32

3.7

The PM Template..............................................................................................38

3.8

The PM Task Basis and Content ......................................................................41

3.9

Duty Cycles .......................................................................................................45

3.10 Service Conditions ............................................................................................46 3.11 Maintenance Risk..............................................................................................48

xviii

LIST OF FIGURES

Figure 1 Figure 3-1 Figure 3-2 Figure 3-3 Figure 3-4

Flow Chart: PM Development Process....................................................xvi PM Template: Medium Voltage Motors ...................................................24 PM Basis Table 3.1 – Medium Voltage Switchgear .................................26 Failure Time Characteristics Taken Into Account by the PM Basis..........33 Table 3.2 – Low Voltage Motors...............................................................42

xix

1 INTRODUCTION

1.1

Background

Many U.S. nuclear plants are in the process of reducing preventive maintenance (PM) costs and improving equipment performance by more closely matching PM tasks with the functional importance of equipment. For this to succeed, utilities require information on the most appropriate tasks and task intervals for the important equipment types, while accounting for the influences of functional importance, duty cycle and service conditions. This data that not exist previously in an accessible form, often resulting in arbitrary and unsuitable tasks and intervals, that increased maintenance costs and diminished reliability. EPRI has long recognized the need to provide its members with maintenance optimization tools and was the first to promote the use of reliability centered maintenance (RCM) in optimizing PM activities in the power industry. RCM was developed in the 1960’s by the commercial airline industry to apply reliability concepts to maintenance and the design of maintenance programs. It was hoped and later proven, that the RCM approach to preventing equipment, and ultimately aircraft, failures through maintenance specifically aimed at preventing equipment failure mechanisms, would maintain the high reliability desired for commercial aircraft and would in turn be cost effective. As good a tool as RCM proved to be in optimizing PM activities, it was not completely embraced by the nuclear utility community. Nevertheless, utilities continue to seek to improve maintenance at reduced cost. As deregulation of the electric power industry approaches, many utilities, especially those with nuclear units, recognize the fact that they must reduce operational and maintenance costs to be a low-cost producer of electric power. One area of opportunity is the optimization of preventive maintenance programs, with the aim of making them technically sound and as cost-effective as possible. Compounding the relatively straightforward reasons driving PM Optimization, i.e. reliability, cost control and cost reduction, is the US Nuclear Regulatory Commission’s issuance of 10CFR50.65, “the maintenance rule”1. In brief, the maintenance rule requires US nuclear utilities to develop a reliability and availability monitoring program for the systems, structures, and components (SSC’s) considered to be within

Introduction

the scope of the rule. Such monitoring will be designed2 to determine the effectiveness of the maintenance performed on these SSC’s. In addition, the maintenance rule also requires the utility to evaluate industry-wide operating experience and to account for that experience when modifying their maintenance activities. PM modifications are expected to be the most likely corrective action when maintenance is found not to be effective under the maintenance rule. Despite these motivations, most PM optimization techniques, including RCM, are not designed to develop the time to failure relationships required to justify the selection of PM task intervals. The US nuclear power industry suffers from a general lack of this kind of component data. An additional factor missing in PM optimization programs is the understanding of the factors that influence the progression of the degradation mechanisms for the given component. Not understanding these relationships and how they affect PM tasks and intervals, has made the development of cost-effective and technically appropriate PM programs difficult if not impossible. At the end of 1993 EPRI published a report “Guide for Determining Preventive Maintenance Task Intervals”3. This report presented task intervals for nominally equivalent PM tasks for major equipment types from about two dozen nuclear power plants. The data showed a wide variation in task intervals among the plants with little preference for particular intervals for similar tasks on similar equipment. Most utility maintenance personnel thought that these results reflected a general lack of specific technical basis for the components and tasks addressed. As a result of the need to reduce maintenance costs and to comply with the maintenance rule, utilities are now experiencing the need to assess both the effectiveness and the underlying bases of their PM programs. One aspect of these assessments is evaluating how their current PM tasks and task intervals are aligned with industry judgments of the most effective program of such tasks.

1.2

Objectives

EPRI’s PM Basis project had two major objectives. The first was to summarize a sample of industry experience on which tasks and task intervals comprise a sound and costeffective PM program for a large number of major component types. The second objective was to outline the relationships between a component’s degradation mechanisms and the factors that influence them, the progression of these mechanisms to failure, and the opportunities that are available to discover them.

1.3

The PM Basis Reports and Their Interpretation

The PM Basis project emanated from utility requests for technical information to support the assignment of preventive maintenance (PM) tasks to nuclear power plant equipment. Currently PM tasks are assigned on the basis of vendor recommendations 2

Introduction

and experience at the plant, but this is often unduly biased by custom, by conservatism, and by vendor recommendations. The project was designed to capitalize on the different experiences at US nuclear plants, and at the same time to make minimal demands on the time of plant personnel. Thirty nine component types were prioritized for development and are listed in Section 1.5. PM Basis reports for the 39 component types are available as a series of volumes under the TR-106857 report number. This overview report is presented as a master report for the entire project to provide a project overview that details the process by which the PM Basis was developed. Section 1 of this overview report provides the motivation, organization, and essential information for understanding the reports. Section 2 considers how the component reports and the technical bases can be used in utility projects such as PM optimization, PM audits, maintenance rule compliance, developing a plant PM Basis and in optimizing PM task intervals. Section 3 contains the process description; it also contains a description of what is to be found in each of the component reports, and how this information is to be interpreted. The PM Basis was tasked by the steering committee to be a body of information that includes an interpretive summary of utility power plant experience on preventive maintenance for each component type. Each report contains, as much as practical in one place, all the PM tasks, task intervals, task rationales, and the most important influences on equipment degradation and maintenance for that specific component. This body of information provides additional context and perspective when plant personnel are seeking to interpret vendor requirements in the light of plant specific conditions, or when they need to understand the motivations for and limitations of PM tasks when modifying task content or extending task intervals. In many instances the information can be used in concert with input from the manufacturer to understand how vendor requirements might be relaxed in specific circumstances. The information in the PM Basis reports is intended to complement, not to replace the PM instructions in vendor manuals. If changes to PM tasks are being introduced that conflict with the vendor recommendations, all the relevant industry information sources such as Vendor Bulletins, Generic Letters, Information Notices, SER’s, and SOER’s should be consulted, as well as the history of performance and asfound condition reports at the plant in question. Each component report in the 39 volume series is divided into 3 sections: Section 1 explains the approach used to develop the PM basis and the strategic considerations that influenced the expert working group. This section provides, 1) a description of the PM Basis process steps and how they are integrated to develop the final report, 2) cautions on the effective use of vendor manuals, and on the determination of task intervals, 3) a component boundary description that bounds the area of application, 4) specific component design changes that may be used to increase 3

Introduction

existing task intervals, and 5) current Industry references of which the end user should be aware, that may have a bearing on PM task content and intervals. Section 2 presents the recommended PM program in a concise format. The selected PM tasks are grouped appropriately and presented as a template, covering eight sets of application conditions such as “functionally critical components with a high duty cycle and severe service conditions”. Explanatory notes provide the definition, scope, focus, and rationale of each PM task. Examples are provided of components in nuclear power plants that fall into the application categories used in the template. This section also contains the definitions for component functional criticality, duty cycles, and service conditions that are used to establish task intervals and the need for each task. Section 3 contains a tabular summary of degradation and failure mechanism information obtained by direct interviews with the expert panel members. The data contained in Table 3.1 of each component report entitled, Failure Locations, Degradation Mechanisms, and PM Strategies, represents the panel’s opinions of the factors that influence failure: 1) where failures are most likely to occur, 2) the degradation mechanisms, 3) the factors that influence the degradation, 4) how the degradation progresses over time, 5) when failures are likely to occur, 6) the opportunities to recognize the on-set or status of the degradation, and 7) the PM actions and strategies that can be employed to prevent a failure from occurring. Additionally, this section contains a list of the component’s dominant or most common failure locations and mechanisms to assist the user in understanding the principal influences that drive maintenance actions. When reading Table 3.1, users should note that items in the 3rd, 4th, and 5th columns are carefully aligned with each other across the page. In columns 4 and 5, entries are assumed to repeat going down the columns, unless a new item is entered. This feature makes the table much easier to read but requires the user to mentally “fill in” blank lines with the last entry above. Data extracted from Table 3.1 is presented in a second form, Table 3.2 entitled, PM Tasks and their Degradation Mechanisms. Table 3.2 recompiles the data in Table 3.1 removing repetitive information, and leaving out most of the timing information. Table 3.2 focuses on the PM tasks themselves, showing very quickly which failure causes are addressed by each task, whether or not a task covers a broad group of potential failure locations, and which tasks are backed up by other tasks. The expert panels are most efficient when the panel size is no more than 8 to 10 members. The expert members are experienced and knowledgeable maintenance, component, and system engineers, and providing they consider a wide range of applications, the information and technical basis they provide should be valid and adaptable for all nuclear power plants. Even when fewer experts were present, their input spanned a significant segment of industry experience. It is not feasible or 4

Introduction

practical to develop a true consensus for the PM programs among the approximately 100 US nuclear plants without a project that consumes vastly more resources. Consensus may eventually be gained as the task bases, and task intervals contained in each component report are reviewed and utilized by the utility community. Plans have been made and resources allocated to maintain and periodically improve the information developed by the PM Basis project by distributing it as an electronic database in MSAccess97“, with update features that use Internet communications.

1.4

Definition of Preventive Maintenance

When developing the PM Basis project, project staff quickly realized that there is no consistent definition for preventive maintenance among plant maintenance personnel. In general, preventive maintenance is the conduct of preplanned (i.e. scheduled) tasks necessary to ensure safe and reliable operation of the equipment. PM is the total aggregation of these scheduled tasks along with their assigned task intervals, and includes such tasks as oil sampling, vibration monitoring, visual inspection, lubrication, and the scheduled removal and replacement of parts prior to equipment failure. Even when an overhaul is not scheduled (i.e. scheduled and completed at some fixed time interval), it could be considered to be a PM task, if it is performed as the result of another task (e.g. vibration monitoring or oil sampling) whose primary purpose is to monitor equipment condition and to detect the on-set or initiation of failure. Providing the failure has not already occurred the overhaul then becomes part of the action taken to prevent a failure. However, overhauls resulting from the occurrence of failures are not preventive but corrective maintenance. Each preventive maintenance task has a primary purpose. Some provide a warning of impending failure, some include specific actions to prevent undesirable levels of degradation, and some PM tasks are designed to discover whether standby equipment has already failed. In the latter case, the task prevents the accumulation of extended periods of unavailability. This is the only sense in which it is still a preventive task. The PM Basis project utilizes three distinct types of PM task for each component PM Template, see Section 3.7 of this report. The three types are “Condition Monitoring”, “Time Directed”, and “Failure Finding”. Together these three task types comprise the basic set of preventive maintenance strategies employed by the majority of nuclear utilities. The PM Basis project chose these PM task types because they provide a practical terminology that is reasonably consistent with current usage. To maintain a consistent focus and to provide a common starting point, the following definition for PM was adopted for the PM Basis project:

5

Introduction

Preventive maintenance is considered to be any task or group of tasks that is planned and scheduled, whose purpose is to prevent the unanticipated failure of a component by the monitoring or inspection of equipment condition, replacement or refurbishment of pre-specified parts or subcomponents prior to their failure, or the functional testing of such equipment to determine its ability to function upon demand.

1.5

Major Components

Thirty nine reports have been completed. Those components are: Project Overview Report – TR106857 V1 Air Operated Valves V2 Medium Voltage Switchgear V3 Low Voltage Switchgear V4 Motor Control Centers V5 Check Valves V6 Motor Operated Valves V7 Solenoid Operated Valves V8 Low Voltage Electric Motors (600V and below) V9 Medium Voltage Electric Motors (between 1kV and 5kV) V10 High Voltage Electric Motors (5kV and greater) V11 Direct Current Electric Motors V12 Vertical Pumps V13 Horizontal Pumps V14 Reciprocating Air Compressors V15 Rotary Screw Air Compressors V16 Power Operated Relief Valves - Solenoid Actuated V17 Power Operated Relief Valves - Pneumatic Actuated V18 Pressure Relief Valves - Spring Actuated V19 HVAC - Chillers and Compressors V20 HVAC - Dampers and Ducting V21 HVAC - Air Handling Equipment V22 Inverters V23 Battery Chargers V24 Battery - Flooded Lead-Acid V25 Battery - Valve-Regulated V26 Battery - Nickel-Cadmium (NICAD) V27 Liquid-Ring Rotary Compressor and Pump V28 Positive Displacement Pumps V29 Relays- Protective V30 Relays- Control V31 Relays- Timing V32 Heat Exchangers 6

Introduction

V33 V34 V35 V36 V37 V38 V39

Feedwater Heaters Condensers Main Feedwater Pump Turbines Terry Turbines Main Turbine EHC Hydraulics Transformers- Station Type Oil Immersed I&C Components

The project scope did not extend to equipment such as emergency diesel generators, control rod drive mechanisms, and reactor coolant pumps because vendor manuals and special industry projects have already treated preventive maintenance for this equipment in significant detail. However while the above specialized equipment were not the subject of group meetings, some guidance on their PM may be developed by looking at the practices for applicable subcomponents groups, e.g. motors, pumps, SOV’s, and AOV’s.

1.6

Project Organization

To promote utility buy-in and support, a utility oversight committee, the PM Basis project steering committee, was established. The steering committee was comprised of the EPRI project manager, a utility chairman, and 10 other utility members supported by the project staff. The steering committee, through the EPRI project manager, was also accountable to the Operations and Maintenance Cost Control Target Steering Committee for ensuring all utility needs were considered and met. The steering committee was granted purview over the structure of the PM Basis project, the prioritization and selection of the component types to be analyzed, the composition of the expert panels, and the methodology employed for the development of the component PM tasks and rationales. From the onset the steering committee acknowledged that they were not the technical experts, and deferred the inclusion of any equipment-specific recommendations to the individual component expert panels. Both the steering committee members and EPRI recognized that project success hinged on continuing utility support through participation in the component expert panels. Securing utility support became one of the committee’s primary functions. The Component Expert Panel of utility component experts provided the raw data on equipment degradation and the range of PM task options for the selected equipment. Each panel, as far as practical, was composed of knowledgeable individuals from EPRI, EPRI member utilities, manufacturers, and others. Most of the expert panels addressed a small number of closely associated component types, e.g. three types of pressure relief valves. The panels developed the preferred PM practices and raw material on equipment degradation in tabular form. The contractors developed the task rationale from all the data supplied, and submitted the component report to the expert panel for 7

Introduction

comment and approval. After their comments were incorporated, the reports were reviewed by the EPRI project manager.

1.7

Component Expert Panels

The steering committee recommended that each expert panel should consist of a minimum of 4 utility members to assure a reasonably diverse experience level. It was found that panel sizes of 8 total members were ideal and panel sizes up to 10 were manageable. Each panel met once for 4 days on average and initially began with a minimal level of understanding of the process. Each panel was provided with an outline of the PM Basis process, and experienced its own learning curve while implementing the process, facilitated by two contractor personnel. By the end of the four day meeting they were usually quite efficient at compiling the required information. It should be noted that although a few members were skeptical at first, all panel members have indicated concurrence with the process. Experience showed that between 2 and 2 1/2 days were required to complete the basis for the first component to be considered by a panel; subsequent components were much faster, taking advantage of the panel’s learning curve, and the results previously documented. On rare occasions, because of last minute emergencies, an expert panel has had as few as two members. Although the project did not plan for this small number it is felt the two experienced engineers could still provide a high quality set of data sufficient to meet project needs. The recommendations of these expert panels represent a consensus of that group on time-directed, condition monitoring, and failure finding preventive maintenance tasks, and task intervals with bases information that supports their conclusions. Consensus here does not necessarily mean finding the “best” PM task, it means reaching agreement on a range of tasks and task intervals with an agreed upon technical basis derived from industry experience. Where possible, each expert panel was composed of utility personnel with collective experience in component and systems engineering, reliability engineering, and PM improvement, to endeavor to produce recommendations that are focused on PM program improvement and not just on individual component maintenance.

1.8

Project Learning Curve

A change in project methodology and content occurred after the first component (AOV) was completed. The AOV report has more of an RCM flavor and endeavors to describe component failures and the influences on them in terms of the equipment’s failure modes that are used in safety analysis. The focus on failure modes has subsequently changed to a focus on failure locations and mechanisms. Despite these 8

Introduction

differences, the PM program for AOV’s and the supporting information in the AOV report, volume 1, is felt to present a technically sound and practically useful PM program. Future updates will bring the AOV report into compliance with the current process and format. After the AOV report was completed, Table 3.1, Failure Locations, Degradation Mechanism, and PM Strategy was added. Table 3.1 provided the depth of information necessary to describe how a given component fails and what PM strategy can be employed to defend against the failure mechanisms. However, Table 3.1 did not completely answer how to extract the rationale for each task. What was needed was a way to summarize the data and to connect it with information the expert panel had provided on the most dominant failure causes, the focus of each task, and the task interval. Table 3.2 was developed from Table 3.1 to remove the repetitive information and to connect tasks to failure locations more directly. The process described in Section 3.8 of this report, was then used to provide a complete and coherent view of the rationale for each task, and to provide assurance that task intervals suggested by the expert panel were soundly supported by the detailed data. Table 3.2 also provides the utility user with all the failure locations and causes addressed by each task, displays whether or not a task covers a broad group of potential failure locations, and reveals which tasks are backed up by other tasks.

References 1. “Requirements for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants” (the maintenance rule), 10 CFR 50.65, Federal Register, July 10, 1991. 2. “Industry Guideline for Monitoring the Effectiveness of Maintenance at Nuclear Power Plants, NUMARC 93-01, Revision 2, April 1996, Nuclear Energy Institute. 3. “Guide for Determining Preventive Maintenance Task Intervals”, EPRI TR-103147, December 1993. 4. EPRI Training course material, “Reliability Centered Maintenance”, Module 4.

9

2 USE OF THE EPRI PM BASIS

2.1

PM Optimization

In the years leading up to this project the majority of US utilities were engaged in optimizing PM programs at nuclear power plants using various types of Reliability Centered Maintenance approaches. This project was conceived as a response to requests for logically defensible PM tasks and intervals that had been found by the industry to be both technically applicable and cost-effective. Today, PM optimization can be performed at several different levels, 1) in the traditional way as a project-oriented activity in which a large part of the plant PM program is reviewed in a relatively short time, 2) in a similar process whose main goal is to update previous optimization results and to bring them into conformity with the plant’s Maintenance Rule program, 3) in a process that may be focused on special aspects such as task interval adjustment and/or improved use and integration of predictive maintenance, and 4) through a Living PM process which is a gradual optimization over a period of several years in response to equipment condition, maintenance experience, performance monitoring, and the evolution of maintenance technology. The EPRI PM Basis reports can directly support all of these approaches in various ways. For example, the updating or selection of PM tasks for a large amount of equipment benefits from the use of a standard baseline of tasks such as the EPRI PM Basis, which also recommends variations to accommodate different duty cycles and service conditions. The EPRI PM Basis provides the necessary justification for moving from vendor recommended tasks and intervals to technically sound and more costeffective options. Access to the EPRI PM Basis will also promote consistent assignment of tasks to the same type of equipment in different systems, especially when these are selected by different analysts or by system engineers. Consistency with a plant’s Maintenance Rule program may involve integration with SSC functions, risk significance, and performance criteria. For example, components whose failure would directly cause the failure of a risk significant SSC with reliability performance criteria limited to only one or two failures would need to be protected by 11

Use of The EPRI PM Basis

a level of PM aimed at preventing all failures, i.e. it would be classed as a critical component. A component whose failure has only an indirect effect on such performance criteria, or whose failures would contribute to more relaxed performance criteria will need PM tasks to prevent just the most common failures, rather than all failures, i.e. would be classed as a non-critical component. Similar considerations would apply to preventing repetitive failures from the most common causes. The data tables in the EPRI PM Basis reports enable these choices to be made with some precision and with a clear technical basis. Although such a PM basis is not required initially for Maintenance Rule programs, a plant-specific PM Basis will be a de-facto requirement to support PM changes when these are the substance of corrective actions and goals to regain (a)(2) status. This is the reason some utilities are preparing a documented plant-specific PM Basis. The most efficient way to establish a plant PM Basis is to define plant PM standards using the EPRI templates and the experience of evaluating the first few systems. Such plant PM standards speed the task selection process, encourage consistency in task and interval selection, and efficiently embody much of the plant PM Basis. A focus on specific areas of improvement such as task intervals and predictive maintenance can make use of the time scales for development of various degradation mechanisms, and the specification of appropriate predictive techniques listed in the EPRI PM Basis data tables. For example, some PM task intervals on certain component types are fairly tightly constrained by known time scales of degradation. These are not worth the expenditure of plant resources in attempts to extend the task intervals. Other task intervals invite extension, with potential limits that can also be determined from the EPRI data. These data tables are to receive additional enhancement in the near future to become even more comprehensive and accurate.

2.2

PM Audits

An informative starting point for any PM optimization is a direct audit of an existing PM program. This can provide plant management with a means to judge how the plant PM program compares to standard industry recommendations, and where there is most need for improvements. Typically this can be done very inexpensively by a sampling approach in which a vertical slice audit is performed on specific equipment types in a small number of systems, chosen because of their diversity, because of recurring equipment problems, or because of other suspected weakness in the PM program for those systems. These audits can be conducted on site by one contractor in four or five days, aided by a utility PM coordinator who has responsibility for accessing the plant databases and for arranging interviews with plant personnel. 12

Use of The EPRI PM Basis

The audit proceeds by listing the PM tasks for each component of the agreed types in the systems under examination, and comparing with the EPRI PM Basis recommendations. Where there are significant differences, component or system engineers are interviewed to discover the reasons. Cases where an adequate technical basis can not be discovered for the differences are written up as audit findings. The audit report consists of a disposition of each PM task and the degree to which it is in concordance with the EPRI PM Basis recommendations. The report is written immediately following the audit, completing an efficient process in a minimum of time. Tools that EPRI has developed to facilitate the audits include data forms for each component type already loaded with the EPRI Template data. These provide for a directed search for specific PM tasks in the plant databases and are instrumental in completing a thorough detailed review. Interviews are conducted with system and component engineers to confirm preliminary assignments of functional importance, duty cycle, and service conditions, and to discover component history and the basis for plant PM tasks and intervals. The only barrier to a high efficiency audit is the quality of plant databases. When PM tasks are not comprehensively contained in the databases or are distributed over several databases the audit can be less efficient. However, with the assistance of application guidelines which EPRI is building into an electronic database form of the EPRI PM Basis reports, it will be possible for utilities to conduct self audits without the assistance of a contractor.

2.3

Evaluation of the Benefit of PM Changes

Utilities do not currently have a methodology and data to enable the most important benefits of performing a PM task to be assessed and compared to alternatives. Currently, the benefits of PM changes are evaluated solely on the basis of preventive maintenance cost reduction, simply from the change in the amount of PM performed, and not at all in terms of a reduction in the number of failures. But the latter is the only reason that PM resources are deployed at all in any industrial plant. The EPRI PM Basis reports contain almost all the information needed to assess the reliability benefit of performing a specific PM task at a given interval in the context of the other PM tasks that are being performed. This information is contained in the breakdown of failure locations, degradation mechanisms, and the times to first failure, combined with the assignments of particular tasks that are applicable to address each mechanism. A software tool is being prepared called Maintenance Benefit Analysis, or MBA, that will numerically roll up this information to give simple evaluations of the marginal benefit of a PM change. 13

Use of The EPRI PM Basis

The benefit evaluation alone will be a valuable tool when selecting PM tasks and intervals. However, when combined with estimates of the costs of performing PM tasks, the costs of unplanned failures, and the costs of corrective maintenance, and other cost impacts such as radiation exposure, MBA will constitute a powerful costbenefit tool for maintenance improvement and O&M cost reduction. This step will require integration of the MBA tool with other software products that EPRI is preparing for utility cost evaluations.

2.4

Improving The Use And Integration of Predictive PM Tasks

One of the principle benefits of introducing predictive PM tasks is to lessen the need for intrusive activities which have a high cost and which may also have a high chance of causing additional failures. Time directed tasks such as internal inspections, scheduled replacements of components, and overhauls might be largely eliminated in this process, or their time intervals might be extended. There is a risk that a newly introduced predictive capability might not be completely effective at detecting incipient failures while there is enough lead time to plan for corrective action at a convenient time, so that the cost of the resulting unplanned plant outages exceeds the cost savings from performing fewer time directed tasks. This concern has been a major factor slowing the introduction of predictive technology in nuclear power plants. A fully informed decision on this topic would need to consider if there are any failure mechanisms that the previous time directed task covers, that might not be addressed by the full program of predictive tasks and other time directed tasks that remain in the program. The PM Basis can be used in precisely this way, reading this information directly from the task rationale and from EPRI PM Basis Tables 3.1 and 3.2. It can also add context and perspective by showing the compromises that are made in the original PM program, even among the time directed tasks. Not all failure mechanisms are always addressed in a comprehensive way, the intrusive task interval will carry a specific risk of failing to address random failure events that occur at short times, and it is likely to involve a high degree of risk of introducing new failures. Additionally, the new Maintenance Benefit Analysis software tool will be able to assess the interplay between predictive tasks that are performed frequently, but possibly with reduced effectiveness in identifying causes of failure, compared to the more intrusive tasks which may have been more certain to identify impending failure mechanisms but which were performed too infrequently to be as effective as the predictive tasks. When the PM Basis is used in this way the relative benefit and risk of the new predictive task can be viewed in explicit relationship to the existing tasks. 14

Use of The EPRI PM Basis

2.5

Craft Information Feedback

To sustain a credible PM improvement strategy over the long term it is absolutely essential that the condition of the equipment, observed during maintenance, be recorded and directed back to influence changes to the PM program. The PM Basis can be the starting point for the development of information feedback forms to be used by craft personnel when executing PM and CM work orders. These forms can contain checkboxes narrowly focused on the dominant failure locations, degradation mechanisms, and driving influences specific to the equipment type and task at hand. Generic feedback forms are of lesser value because they are too lengthy and detailed for maintenance personnel to use efficiently, containing too many questions of little or no pertinence to the equipment and task being performed. The design of condition information feedback forms and the method of electronic information capture and retrieval from them is the subject of ongoing investigation, and has not been carried to the point of application at the time of writing this report. However, feedback forms for 20 component types have already been prepared using the EPRI PM Basis reports. These forms require customizing for use at a plant by screening to retain only the most important items, and to be tuned to the specific hardware design at the plant. In principle, the derivation of information on equipment condition can be an essential feedback to a Living PM Process once the problem of electronic recording of the information is solved. EPRI has already developed an evaluation process for the use of such information in making changes to a PM program.

2.6

Electronic Database Of The EPRI PM Basis Data

The EPRI PM Basis reports are also to be distributed as a series of computerized data tables with associated application software on an MSAccess platform. Availability of the electronic form of this data will launch three developments: 1. A user interface designed to facilitate utility applications such as those described above. Included in the interface will be Help files that constitute an Applications Handbook, containing guidance on querying the data, using the interface with associated software products, and implementation insights from relevant utility projects. 2. The ability of utility personnel to review and update the data and to submit their revisions to EPRI via the Internet. Even a small number of independent utility reviews for each component type will significantly improve confidence in the universal applicability and accuracy of the data. EPRI personnel will have the

15

Use of The EPRI PM Basis

capability to incorporate new utility information from the reviews in updates to the data which can be redistributed via the Internet. 3. Gradual transformation of the initial EPRI PM Basis data, simply via the above reviews over time, into a repository of utility PM experience and data on equipment degradation, retaining the exact format of the EPRI PM Basis data. This repository of industry information will be extremely efficient to create and to retrieve information from because it will not be an event-based database. The information in the database will already be in a summary form, available for direct use through the Applications Software. EPRI would take steps to incorporate EPIX data into this process. Transfer of the EPRI PM Basis report data into Access tables will be complete by the end of 1998, and the Application Handbook Help files will be completed during 1999.

16

Process for Developing the PM Rationale

3 PROCESS FOR DEVELOPING THE PM RATIONALE

3.1

Overview

Section 3.1 will describe the process that was followed to elicit information from plant engineers, and to develop and describe the rationale for each PM task for a given component type. The basic premise is that the rationale for why a PM task has a certain content, focus, and time interval can be understood by asking which PM activities can most effectively address each of the causes of degradation and failure for the component. This information is well known by component engineers, system engineers, and by maintenance engineers and technicians who are experienced in the types of failures and deterioration that have been observed in the industry during the performance of preventive and corrective maintenance over a long period of time. This information can be retrieved by first discussing with them which parts of the equipment typically degrade or fail, which mechanisms are usually responsible for the degradation, which factors in the physical or operational environment have the most effect in initiating the degradation or in making it more severe, and how long the deterioration can be expected to progress before it becomes unacceptable, or results in a failure. Subsequently, the discussion can move on to the kinds of PM techniques or activities that have the best chance of discovering the degraded conditions, if they exist, and which higher level PM tasks should include these activities. When this data has been obtained and documented it can be screened in various ways to discover the more common types of degradation and failures addressed by each task, the tasks that are aided by other PM tasks which also address the same failure locations and conditions, and the tasks that are most relied on because they do not have a backup from other tasks. This analysis also identifies a logical time interval for each task that is determined by the time scales of occurrence of the failures which are addressed. There are three key parts to this process for each component type. The first consists of extracting the above data from the experiences of a small number of utility plant engineers in a workshop-like format, the second is the development of the rationale for each task and the presentation of a summary of the whole program of such tasks in a table called the PM Template, and the third is the addition of supplementary information such as examples taken from nuclear power plant systems, definitions of the component boundary and applicable duty cycle and service conditions, a discussion 17

____________________________________________________________________________________________ Process for Developing The PM Rationale

of the risk to reliability of doing intrusive maintenance, the most common or dominant failure locations and mechanisms taken from industry sources, and the availability of design modifications that can improve reliability or decrease dependence on preventive maintenance. In what follows it will be observed that terms such as degradation, degradation mechanism, deterioration, failure cause, failure mechanism, and failure type are used more or less interchangeably to indicate the process that leads to failure. Degradation is usually the preferred usage. The terms “task basis”, “technical basis”, and “task rationale” are also used interchangeably. It has been found that no purpose is served in this work by attempting to be more precise about the employment of these terms. Failure modes of equipment, such as those used by NPRDS, and familiar in Probabilistic Safety Analysis, e.g. “fails closed”, “fails to run”, “fails open”, etc. are not used extensively in this work. Preventive maintenance tasks primarily address the degradation of equipment (e.g. corrosion or a bent valve stem) and it matters little to a PM task in which mode the equipment eventually fails. Most degradation mechanisms can affect more than one failure mode, perhaps most or all of them as is the case of corrosion or a bent valve stem. Although consideration of failure modes can be of value in deciding whether certain component failures can be tolerated or not (e.g. “fails open” can be functionally critical, whereas “failed closed” can be the fail-safe mode), there are generally no PM tasks that are specific to “fails open” but not to “fails closed”. Consequently, apart from a short experiment with the use of failure modes in the AOV report (Volume 1), they have been found not to be useful in developing a PM Basis. The PM Basis has been prepared for components which are generally quite complex pieces of equipment. The general term adopted in these reports for the hardware at this level of description (e.g. Medium Voltage Switchgear), is “Equipment”. Subcomponents such as bearings, shafts, coils, gaskets, and stator windings, which are the sites of specific degradation processes and failures are referred to as “Failure Locations”. The term “Component” is used interchangeably with either of these meanings, depending on the context.

3.2

The Expert Panel Meeting

The information required was obtained for each component type during an intensive three to four day meeting with a small group of industry experts drawn from EPRI member utilities. As indicated in Section 1.7, even one or two experienced engineers were able to provide the majority of the information required when it was difficult to assemble a larger group owing to time constraints or other emergencies. However, for most component types, interaction between the members of a somewhat larger group provided a wider coverage of equipment designs and manufacturers, a wider base of industry experience, and opportunities for discussion of the more difficult areas. 18

Process for Developing The PM Rationale

This Section gives an overview of the steps involved in the expert panel meeting. To construct the rationale for a PM program it was found that a large amount of very specific information is required. It was essential that the meeting of component experts was used to maximum efficiency to provide this information in a short time. For this reason the meetings quickly became highly structured, closely following a multi-step process to ensure disciplined coverage of all the required aspects. The following is an outline of the steps that were used. STEP 1. In many cases failure cause and maintenance data from obtained applicable EPRI NMAC guides were briefly reviewed by project personnel before the expert panel meetings to provide input to the panel on an appropriate equipment breakdown into component categories such as “Kingsbury type bearings”, or “pistons and cylinders” that could be used as the hardware locations of failures and the site of specific degradation mechanisms. This input also indicated the relative proportions of different failure causes such as “misadjusted switches”, or “aging of elastomers”. The expert panel always determined the final choice of hardware breakdown and the list of the likely failure locations and causes. STEP 2. The expert panel determined if the component type needed to be sub-divided into logical groups by design characteristics. For example, it was decided that pumps should be divided into a “vertical pump” group, a “horizontal pump” group, and a “positive displacement pump” group because these types of pumps have enough design and maintenance differences between them to warrant separate treatment, yet each addresses a sufficiently generic set of equipment to enable a single treatment to be made for each group without further subdivision. In a many cases the need for further subdivision was dealt with on an ad hoc basis, generally by including alternative components and expecting the user to decide which ones apply. An example would be reciprocating compressors with oil bath air filters in relation to those with dry element filters. Each of the various major groupings of equipment, e.g. “vertical pump”, was then treated to the multistep process separately, and written up as a separate PM Basis report. STEP 3. The definition of the equipment boundary and the components and subcomponents to be included when considering PM tasks, was made so that it is clear whether auxiliary devices such as external lubrication systems, interfacing components such as pump/driver couplings, and various control and instrumentation components are included or excluded from the PM Basis reports. STEP 4. The PM Template is organized so that particular duty cycles such as standby operation versus continuous operation, and service conditions such as being exposed to the outside environment and weather, can have an influence on the PM task intervals or on the need for a PM task. The service conditions and duty cycles that impact PM strategies were established at this point in the process whenever possible, although it was sometimes necessary to revisit the definitions at a later stage. For consistency in 19

____________________________________________________________________________________________ Process for Developing The PM Rationale

using the Template there was always provision for two duty cycles termed “High” and “Low”, and two sets of service conditions termed “Severe” and “Mild”. This adequately covered all the cases encountered, although occasionally two choices for each were not needed. For example, it was determined that all spring actuated safety relief valves fell into the class “Low Duty Cycle”, and are operated under “Severe Service Conditions”. Some components required very careful definitions of these terms, occasionally involving an interaction between duty cycle and service conditions. For example, motor operated valves incorporate a pressure drop consideration in the duty cycle definition. These valves also require a separate use of the Template for task intervals for the actuator and the valve whenever there is a significant pressure drop. An additional consideration addressed at this time is the division of components into two groups, those that are functionally so important that it is worth spending considerable preventive maintenance resources to avoid failures, and those for which only the spending of considerably lower maintenance resources can be justified. The former are referred to as “critical” components, the latter as “non-critical”. Critical components can be critical because of their functional importance to safety or to electricity generation, or both (or on the basis of certain other user-defined criteria). Generally, a comprehensive level of PM has to be developed for critical components so that all failures are prevented. Non-critical components require some level of preventive maintenance rather than permitting them to fail. The PM objective is to control reliability at an appropriate level, generally by preventing the most common types of failures. A third group of components can be referred to as the “Run-to-Failure” group for which preventive maintenance is neither functionally nor economically justifiable. The PM programs addressed by the PM Basis project obviously do not apply to the “Runto-Failure” components. However, the critical and non-critical types are both addressed by the PM Basis project. The definitions of critical and non-critical components are almost always unchanged for different component types. STEP 5. At this point it was also necessary to develop a preliminary list of PM tasks. Without such a list there would be frequent confusion in the expert meeting as to whether a specific PM activity would be performed on one occasion or another, accompanied by other tasks or not. For example, “inspection” was often identified as a possible means of detection of a degraded condition, meaning that if an observer cared to look in the right place, and knew what he was looking for, it would be possible to see the condition. To this was added a statement of the higher level PM task that would provide the opportunity for such observations. Such a higher level statement or PM task could be “External Visual Inspection”, “Partial Disassembly”, “Partial Refurbishment”, “Overhaul”, “Elastomer Replacement”, “Functional Test”, or even 20

Process for Developing The PM Rationale

“Calibration”, depending on the degree of disassembly required. Sometimes the relevant PM task would depend on who would be performing the task; for example, I&C technicians might be the only people qualified to observe a certain condition, and they would perform the appropriate inspection during calibration. The higher level task labels such as “Refurbishment”, or “Functional Test” are normally referred to in the reports as PM tasks or PM strategies. Line items that might be included in these PM tasks, such as “inspect and clean filter” are normally referred to as “activities”. Occasionally, a single activity such as “inspect and clean filter” is the main focus of the task, and practically the only thing that is done in the task. Such a case would then be made into a higher level PM task such as “Filter, Clean and Inspect”. More information is provided in Sections 1.4 and 3.6 on the kinds of PM tasks and activities in the reports. At this early point in the expert meeting only the main headings for the PM tasks and a preliminary idea of the task content was required. STEP 6. For the purpose of organizing the breakdown of the component into convenient parts for consideration during the meeting and for reporting, the component was often divided into major maintainable subgroups such as “Actuator” and “Valve” for an air operated valve, or even “Mechanical Components” and “Electrical Components” for electric motors. This was not always necessary or possible, as in the case of check valves and horizontal pumps. It should be noted that this was done only for convenience; on these occasions the subgroups corresponded more to how maintenance is performed on the component, than on design features. STEP 7. For a specific type of equipment the point of departure for consideration of degradation processes, the timing of failures, and the choice of PM tasks was the list of failure locations (e.g. packing and mechanical seal) for the equipment. This list was developed by the expert panel and corresponded to the points of hardware degradation and failure that constituted their knowledge of where failures occur, including input from the prior review of industry sources. Each failure location was subsequently considered in turn, with information on degradation, timing and appropriate PM tasks documented in a tabular format. The expert panel did not include extremely rare kinds of failures. When such an inclusion was made inadvertently, a PM task was only assigned if it would have been performed for other reasons (i.e. to address a different failure location). STEP 8. For each failure location the group assigned the main degradation processes, the factors that most influence the degradation, and the time characteristics of the progression to failure. This information was documented in a table (Table 3.1) by the meeting facilitators, on a large screen so that everyone could see it, discuss it, and modify it as necessary. Hardcopies were made as the screen became filled so that all attendees at the meeting retained a complete record of the information for reference. 21

____________________________________________________________________________________________ Process for Developing The PM Rationale

Sections 3.4 and 3.5 contain further information describing this process and the table entries. STEP 9. The table was completed by adding the discovery opportunities for each of the degrading subcomponent failure locations. Sometimes this corresponded to observing the degradation process in progress before the failure point is reached, as in monitoring vibration levels. At other times it corresponded to observing deficiencies such as leaks that require maintenance action before a functional failure occurs, and at other times it corresponded to finding failures that have already occurred. The purpose of these discovery opportunities was to identify the physical act (e.g. inspection) or measurement, but not the PM task during which the action would be taken (e.g. refurbishment), although sometimes these would coincide (e.g. vibration monitoring). STEP 10. Following the identification of the discovery opportunities, the PM strategies or tasks during which these opportunities would arise were added. For example, “direct thrust measurement” as a discovery opportunity would be performed during “mechanical off-line test” as the PM strategy for an MOV. These entries were drawn from the preliminary list of PM tasks developed in step 5. Changes were made to the PM list as necessary. STEP 11. After all the hardware failure locations had been processed, members of the expert panel completed the list of PM task contents, i.e. the set of line items representing the scope of all actions to be taken during each PM task. This information was drawn from their experience and also from Table 3.1 just completed, so that all the requirements for each PM task were included. This list of task contents was not intended to be exhaustive in the sense of including PM actions that are specific to a particular make or model, and clearly does not correspond to a procedure, or explain how to perform the actions. Instead, it is an outline of the generic items that should be included to encompass all the opportunities to identify deterioration of the equipment and to provide appropriate intervention. STEP 12. Information had earlier been presented to the expert group members from industry sources such as NMAC analysis of failure causes. This information often identified the failure locations and failure mechanisms or degradation processes that were most prevalent in the industry databases such as LER’s and NPRDS. The expert group used this information and their experience to provide a short list of the most commonly encountered failures for the equipment being considered. This information was used to construct the rationale for each PM task. For example, if a task was the only task capable of addressing a certain failure mechanism, and the mechanism was prominent among failure occurrences, it would be important to recognize that the task fulfilled this special function, and to make observation of the failure mechanism a focal point of the task. The most dominant failure locations and causes, combined with the knowledge of which tasks could address them provided a means to distinguish the most important objectives of each task from the bulk of the data provided. Further 22

Process for Developing The PM Rationale

information on how the rationale for each task was developed can be found in Section 3.8. STEP 13. A list of component examples was then prepared for the eight combinations of criticality, duty cycle, and service conditions. For example, a feedwater pump motor would be one example of a motor that was considered to be critical, to have a high duty cycle, and to experience severe service conditions for medium voltage electric motors. These examples are provided to give the reader a better understanding of the meaning of the eight sets of conditions addressed by the Template and to give the expert panel members a specific focus in constructing the Template. In many instances a given system component may appear in more than one set of conditions. This is perhaps confusing, but demonstrates that plant-specific factors can alter the assignments. For example, a motor in a turbine building cooling water system could be inside the turbine building at one plant but outside at another, or a compressor cooling water system could operate on a closed, and hence clean, cooling water system at one plant but on a raw water system at another. It also demonstrates the value of having eight sets of conditions depicted on the Template, instead of a single uniform set of recommendations being made for all condensate pumps, or for all 6.9KV switchgear. No explanation is offered for each instance where a component appears in multiple conditions, as it is considered reasonably obvious and of little specific importance for the report as a whole. STEP 14. The expert panel next developed the PM Template, having completed the list of PM tasks and their scope, and having considered in some detail the time scales involved in the development of failures, the most important influences on this development, as well as examples of equipment with particular functional criticality, duty cycles, and service conditions. This involved the assignment of time intervals to each task appropriate for the eight sets of circumstances addressed by the Template. The Template for medium voltage electric motors is shown in Figure 3-1. The sets of circumstances cover all the combinations of a critical or non-critical component, with a high or low duty cycle, operating in severe or mild service conditions. If the time intervals are properly assigned they are a logical integration or roll-up of, 1) the factors that were most influential in affecting failures and determining maintenance and which were included in the definitions of the duty cycles and service conditions, 2) the degradation mechanisms, influences, and development time scales entered in the table, 3) the prevalence of these mechanisms in failure history, and 4) the effectiveness of the PM tasks in addressing them. It is obvious that all these factors can not be taken into account explicitly during the expert group meeting. Instead, the group of experts made the assignments from their experience in their own or other plants, and on what they believe would be suitable improvements to that experience. Subsequent analysis by project personnel reconciled the experts’ experiential views with the logical basis that can be extracted from the tabular data. It is a striking feature of the project results that these two viewpoints can 23

____________________________________________________________________________________________ Process for Developing The PM Rationale

almost always be shown to be in very good or acceptable agreement, showing that good industry practice, represented by the PM programs depicted on the Template, has a logical basis. Critical

Yes

1

2

3

4

X

X

X

X

* No Duty Cycle

High X

X X

Low Service Condition Severe X

6

7

8

X

X

X

X

X X

X

Mild

5

X X

X

X

X X

X X

X

Thermography

See PM Application Note 2.3.1

6M

6M

6M

6M

6M

6M

6M

6M

Vibration Monitoring

See PM Application Note 2.3.2

6M

6M

6M

6M

6M

6M

6M

6M

Oil Analysis

See PM Application Note 2.3.3

1Y

1Y

1Y

1Y

1Y

1Y

1Y

1Y

Electrical Tests - On-Line

See PM Application Note 2.3.4

3Y

4Y

3Y

4Y

3Y

4Y

3Y

4Y

Mechanical Tests - On-Line

See PM Application Note 2.3.5

1Y

2Y

1Y

2Y

1Y

2Y

1Y

2Y

Electrical Tests - Off-Line

See PM Application Note 2.3.6

2Y

3Y

2Y

3Y

2Y

3Y

2Y

3Y

Mechanical Tests - Off-Line

See PM Application Note 2.3.7

2Y

4Y

2Y

4Y

2Y

4Y

2Y

4Y

External Visual Inspection

See PM Application Note 2.3.8

1Y

1Y

1Y

1Y

1Y

1Y

1Y

1Y

Partial Disassembly and

See PM Application Note 2.3.9

AR

AR

AR

AR

AR

AR

AR

AR

Partial Refurbishment

See PM Application Note 2.3.10

AR

AR

AR

AR

AR

AR

AR

AR

Refurbishment

See PM Application Note 2.3.11

10Y

15Y

10Y

15Y

10Y

20Y

10Y

20Y

Functional Tests

See PM Application Note 2.3.12

AR

AR

AR

AR

AR

AR

AR

AR

Figure 3-1 PM Template - Medium Voltage Electric Motors

STEP 15. It was necessary to revisit the Template at this point to perform a sanity check on the cost-effectiveness of the “not critical but important” template columns to verify that the proper intervals and tasks were selected when compared to those in the critical columns. There is a tendency of the expert panel to assign similar tasks and intervals for columns 5 to 8 of the Template as for columns 1 to 4, until they have completed the examples (Step 13) and have an opportunity to reconsider the different economic and/or regulatory impact of failures on these two groups of components. STEP 16. In order to provide added focus to the PM task rationales, and an additional checkpoint for the logical basis for the task, the expert panel was next asked to list the main objective of each PM task in terms of the key failure locations and component degradation processes that are most responsible for the task intervals assigned to the template. After the meeting this information is reconciled with the tabular data. It 24

Process for Developing The PM Rationale

usually throws light on the types of failures that are likely to be experienced if the task interval is extended too far, and may provide insight into compensating actions that can be taken as insurance against such failures when extending task intervals. STEP 17. It was also necessary to elicit the expert panel members’ views on the risks of doing maintenance too frequently, or too intrusively. The components vary greatly as to the degree to which disassembly and reassembly can lead to additional problems that are not present before the maintenance is performed. Only in the case of air operated valves (AOV’s) was it possible to complete detailed analysis of failure data and to correlate it with the PM tasks being performed, before the expert meeting took place. For the other component types the opinions of the expert panel were consulted as to the kinds of failures they had found to be caused by maintenance error. STEP 18. Some of the last items of information required from the expert panel are not directly associated with the construction of a rationale for the PM program. They address design improvements that are either currently available or which industry organizations should consider developing in order to improve reliability or to decrease the cost of preventive maintenance. Typically these are modifications which make tasks quicker to perform, or less intrusive, or which overcome particular degradation mechanisms. STEP 19. Finally, the panel provided information on relevant vendor bulletins, and EPRI, NRC, INPO, or owners group reports that had been issued in the previous two years, and other activities such as the recent formation of industry user groups. STEP 20. After the meeting, the project personnel analyzed all the data to investigate the degree to which dominant failures are adequately addressed, the tasks that are important because they are the only means available to address certain failure mechanisms, and the mechanisms that are most responsible for the task intervals being assigned as they are.

3.3

Failure Locations

In the PM Basis report for each component type the list of failure locations can be found in column 1 of Table 3.1, entitled “Failure Locations, Degradation Mechanisms, and PM Strategies”, as shown here in Figure 3-2 for Medium Voltage Switchgear.

The failure locations for Medium Voltage Switchgear are: 25

Failure Locations, Degradation Mechanisms, and PM Strategies Failure Degradation Degradation Degradation Location Mechanism Influence Progression 1. Construction Lubrication Operating . Contamination failure Mechanism 2. Structure maintenance 3. Otherwise continuous 4. Improper breaker maintenance, random or systematic

. Age (time not cycles) .Temperature

Continuous

. Inactivity

Even a few cycles significantly improves condition

Figure 3-2 PM Basis Table 3.1 - Medium Voltage Switchgear

Failure Timing 1 & 2 - Trouble-free period for perhaps 3 years after random contamination. 3. Trouble-free for up to 10 years if contamination is continuous 4. Early failure (1-2 years) after improper action Trouble-free period for a number of years (up to 10 years) 6 year inactivity period leads to failure

Discovery Opportunity . Inspection of exposed surfaces

. Integral periodic

PM Strategy . Visual Inspection . Detailed inspection . Overhaul

tests performed during detailed inspections and overhaul, including: Close timing test Minimum voltage test Trip load test “Feel” of manual operation As above

As above

As above

As above

____________________________________________________________________________________________ Process for Developing The PM Rationale

-

Operating Mechanism Racking Mechanism Main Current Components, Arc Quench, and Insulation Electrical Devices

It should be noted that Table 3.1 may be arranged in sections, each with a heading denoting the major maintenance oriented division of the equipment, where this division was used to organize the work. For example, the MOV Table 3.1 has the first four pages headed “Actuator”, and the next two pages headed “Valve Body Assembly”. Within the “Actuator” Section the failure locations are:

-

Wiring (including insulation, terminal blocks, lugs) Motor Drive Train Switches Spring Pack Fasteners

For horizontal pumps the failure locations are: -

Impeller Diffusers, Volutes, and Channel Rings Balancing Device Wear Rings and Surfaces Shaft Mechanical Seal (where appropriate) Fixed Breakdown Bushing (where appropriate) Packing Gaskets and O-Rings Stuffing Box Flange to the Horizontal Joint External Pump Casing (Barrel and Closure Head) Pump/Driver Coupling (where appropriate) Pump/Driver Coupling (Lubricated Gear Type) (where appropriate) Pump Base Plate and Foundation Discharge and Suction Flanges Bearing Seals Breather Caps and Sight Glass Vents Metering Orifice Connections and Piping Slinger Rings 27

Process for Developing The PM Rationale

-

Anti-Friction Bearings Sleeve Bearings Kingsbury Type Bearings Pump Casing - Horizontal Split Bearing Heat Exchanger Internal Bearing Coolers Lube Oil System

These examples show some variation in the level at which the hardware is treated, partly a result of the choice of failure locations being at the discretion of the expert group, partly a result of a learning curve during the project (the Switchgear was treated at a somewhat higher level and was completed early in the project); the examples also reflect the size and complexity of the equipment. An extreme example of the variation possible is the occasional inclusion of “Lubrication Failure” as a failure location when it was desirable to focus attention on the ways this could occur, even though it is not itself a hardware location. There is no obvious advantage in promoting the standardization of the hardware breakdown between different component types because even the same failure locations can be subject to different degradation mechanisms, influences, and time scales for deterioration in different components. In general there is a high degree of similarity between the failure locations for components comprising different members of the same type (see STEP 2). For example, the failure locations for “High Voltage Electric Motor” differ from those for “Low Voltage Electric Motor” only with respect to components that are present in one motor but not in the other.

3.4

Degradations and Influences

In the PM Basis Report for each equipment type the list of “Degradation Mechanisms” and “Degradation Influences” can be found in columns 2 and 3 respectively of Table 3.1, entitled “Failure Locations, Degradation Mechanisms, and PM Strategies”. Degradation mechanisms are the means by which the equipment is brought to the failure point at the specified failure location. Aspects of the environment, plant operations, maintenance, or design that cause the initiation of degradation processes or which can affect the rapidity with which they develop are simply referred to as “influences” on the degradation. For example, the degradation mechanism of “misalignment of worm to worm gear” in an MOV drive train can be initiated or influenced by “personnel error” and “manufacturing tolerances of individual components”. The degradation “wear” on valve guides is influenced by “horizontal orientation” and also by “chronic operation under a pressure drop in any orientation”. In Table 3.1 entries in columns 3, 4, and 5 are lined up with each other across the page to indicate their association. Successive items in columns 4 and 5 are only entered when they change. So the user must “fill in” the blank spaces with the last entry above. 28

____________________________________________________________________________________________ Process for Developing The PM Rationale

This convention prevents the table from being filled up with data, and makes it much easier to read. A partial list of these degradation mechanisms and influences is reproduced below to illustrate their range and applicability. In contrast, the degradations and influences below are simply listed in two columns and are not associated with each other in a one-to-one relationship. Degradation

Influences

Insulation breakdown Misadjusted Pinched insulation Improper crimping Change of spring constant Crud buildup on seat Damaged seat Sliding wear Cracking Sticking Low oil level Incorrect lubricant Low oil flow Clogged water cooling ports Failed gasket Stuck Wear Inadequate clearances Failed sensor Loose connections Plugged orifices

High temperature Moisture ingress Personnel error Number of cycles Cleanliness of process medium Debris Flow conditions of process medium Radiation Age Clogged/crushed lines Leaking sight glass Clogged air filter Aging of pump Silt accumulation Improper torquing Wear Run time Misalignment Vibration Contamination Moisture from gasket failure

The above items are only a very small fraction of the total number of degradation mechanisms and influences encountered. The list shows that some of the degradations are themselves subcomponent failures (e.g. failed sensor, failed gasket), while often they are an association of hardware and a mechanism, such as plugged orifices, or damaged valve seat. The hardware in such cases is usually a piece-part of the failure location as in “burnt contacts” for the degradation mechanism in pressure switches on a rotary screw compressor. In that case the influence on the degradation could be misalignment of contacts or contamination . Similarly, what appears as a degradation in one place can appear as an influence in another as in “wear” being the influence which causes sticking as the degradation in a compressor unloader valve, but “wear” as the degradation mechanism for bearings, caused by a variety of influences such as “lubrication failure”, “misalignment”, or “manufacturing defect”.

29

Process for Developing The PM Rationale

The point of the above discussion is to illustrate the fact that a division of the failure process into hardware failure location, degradation process, and influences on the degradation is obliged to defy prescriptiveness and to invalidate tight definitions if it is going to be reasonably realistic and efficient. For the purpose of understanding what a PM task is trying to achieve this latitude in definitions does not have any serious consequences. It is more important to have a practical description of the process that component experts feel comfortable with, and which above all, facilitates the extraction of information from their experience in an expeditious manner. To enhance the description, additional information is sometimes appended as a list to the degradation mechanism as in, Leakage, (for Gasket and O-Ring failure as failure location) - Erosion, - Corrosion, - Inappropriate Material and to the influences as in, - Installation Error, (an influence for “Wear” of a pump/driver coupling) - Shaft Fit - Gap Setting - Key Sizing The objective of describing the degradation mechanisms and the influences on them is to alert the user to conditions which might be particularly applicable in his plant. For example, when moisture ingress and contamination are known drivers of insulation breakdown, and the equipment is in a damp and dirty location, this could be recognized as a vulnerability. The vulnerability might be a consideration if a task interval is being extended. The degradation mechanisms and influences also provide information that might be significant for improving craft training by showing what most to look for. They could also be an indication of the value of adequate procedures and training, especially in cases where equipment is subject to many kinds of personnel error across a wide range of failure locations. A particularly high or low potential for personnel error, manufacturing defects, or installation errors might also correlate with the risk of doing maintenance and affect decisions to reduce the amount of intrusive maintenance being performed in favor of condition monitoring. The description is also a starting point for designing information feedback processes from the crafts in a living program, because in combination with history at the plant it can indicate the particular aspects to look for when performing PM tasks.

30

____________________________________________________________________________________________ Process for Developing The PM Rationale

3.5

Time Information for Degradation and Failure

The type of timing information that is useful is that which may have a bearing on the task intervals. It has been found to be more productive to ask first for the time characteristics of the degradation, and then for how this affects the failure times or the point at which the condition would become unacceptable. The “Degradation Progression” in column 4 of Table 3.1 in each PM Basis Report gives an idea of whether the degradation process is present most of the time (“Continuous”), or whether it would not normally be present but might exist or initiate randomly. “Continuous” or “Random” are the most common entries in column 4 of the table. The “Failure Time Distribution” in column 5 of Table 3.1 refers to the expected distribution of failure times to which the degradation leads. Two general possibilities are recognized. One is typical of a predictable pattern of wearout, where a period of time is expected to elapse after a new or refurbished item is placed in service, before any failures appear. This is referred to in column 5 as “Expect a failure-free period ....“, or words to that effect. The failure-free period is then stated in months or years, or as a range of these values. When a failure-free period is expected it is acknowledged that actual failure times will be random, but none are expected to occur before the stated time. The other situation envisaged is where there is no expectation of a failure-free period so that failures could occur soon after the equipment is placed in service. In this case it is supposed that the chance of a failure in any given time period is more or less the same whether or not the equipment has been in service a short time or a long time. This is referred to in column 5 as “Random”. Even if the failures can occur without a failurefree period, there is still an important difference between equipment that is highly likely to fail in a short time compared to equipment whose probability of failure is spread out over a much longer time. For example, this could simply be the difference between equipment with a mean time between failures of 5 years rather than 15 years. It is not usually possible to be precise about the mean time to failure because accurate statistics are not available for specific failure locations, degradation mechanisms and influences. However, component and maintenance engineers often have a general sense of the time scale on which they would expect to see the first few failures from various causes, and so this time scale is entered to provide an approximate guide to task intervals. Column 5 therefore frequently has entries such as “Random on a scale of 3 years”, meaning that failures could occur at any time but the chance of having had a failure only becomes appreciable after 3 years. The intention is to indicate failure time characteristics as depicted in Figure 3-3 of this report. These statements are applied as specifically as the expert members experience allows. For example, in the Low Voltage Electric Motor Report for “Stator Windings” as the failure location and “Insulation Degradation” as the degradation mechanism, eleven different influences are listed. “Vibration” and “Age” appear as “Continuous” types of 31

Process for Developing The PM Rationale

degradation influence with “Failure free for >6 years” for vibration and “Failure free for