MAINTENANCE, MODELING AND OPTIMIZATION

Transcription

1 MAINTENANCE, MODELING AND OPTIMIZATION

2 MAINTENANCE, MODELING AND OPTIMIZATION Edited by MOHAMED BEN-DAYA SALIH O. DUFFUAA ABDULRAOUF Department of Systems Engineering College of Computer Sciences & Engineering King Fahd University of Petroleum & Minerals Springer Science+Business Media, LLC

3 Library of Congress Cataloging-in-Publication Data Maintenance, Maintenanee, modeling, and optimization 1 / edited by Mohamed Ben-Daya, Salih O. Duffuaa, Abdul Raouf. p.cm. p.em. Includes bibliographical bibliographieal references referenees and index. ISBN ISBN (ebook) DOI / Plant maintenance. maintenanee. 2. Factory Faetory management. 1. I. Ben-Daya, M. (Mohamed) II. Duffuaa, S. (Salih) III. Raouf, A. (Abdul), TS192.M '02--dc '02--de Copyright 2000 by Springer Science+Business Media New York Originally OriginaHy published by Kluwer Academic Publishers, New York in 2000 Softcover reprint ofthe hardcover 1st Ist edition 2000 All AH rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher, Springer Science+Business Media, LLC. Printed on acid-free paper

4 Contents List of Figures List of Tables Preface Acknowledgment Contributing Authors xi xvi xvii xxi xxiii Part I INTRODUCTION 1 Maintenance Modeling Areas M. Ben-Daya and S. O. Duffuaa 1. Introduction 2. Maintenance Policies 3. Maintenance Planning and Scheduling 4. Spare Parts Provisioning 5. Integrated Models 6. Future Research Directions Part II MAINTENANACE PLANNING & SCHEDULING 2 Mathematical Models in Maintenance Planning and 39 Scheduling Salih O. Duffuaa 1. Introduction Maintenance Capacity Planning Allocation of Resources Scheduling Preventive Maintenance Scheduling Maintenance Jobs 46 v

5 vi MAINTENANCE, MODELING AND OPTIMIZATION 6. Conclusion and Further Research 52 3 Electric Generator Maintenance Scheduling 55 Ibrahim El-Amin 1. Introduction Maintenance Goals and Model Formulation Solution Methods The Application of Fuzzy Systems to the GMS Problem The Integrated Generation & Transmission Maintenance Sched- ~ ro 6. Conclusions 77 4 Reliability Based Spare Parts Forecasting and Procurement Strategies 81 A. K. Sheikh, M. Younas and A. Raoul 1. Introduction Scope of the Work Reliability Characteristics of the Unit Spare Parts Calculations for Failure Replacement Spare Part Calculations for Scheduled Replacements Purchasing and Maintenance Link Spare Part Classification Practical Spare Parts Management Extending the Ideas to Repairable Systems Concluding Remarks Maintenance Service Contracts 111 D. N. P. Murthy 1. Introduction Maintenance Warranty and Extended Warranty Maintenance Service Contracts Mathematical Modeling of Service Contracts Conclusions and Topics of Future Research Simulation Metamodeling of a Maintenance Float System 133 Christian N. Madu 1. Introduction Maintenance Float Systems The Metamodel Cost Optimization Steps for Developing a Metamodel Conclusion Part III PREVENTIVE MAINTENANCE 7 Basic Preventive Maintenance Policies and their Variations 155

6 Contents vii T. Dohi, N. Kaio, and S. Osaki 1. Introduction Mathematical Preliminaries Block Replacement Age Replacement Order Replacement Conclusion A General Framework for Analyzing Maintenance Policies 185 V. Makis, X Jiang, A.K.S. Jardine and K. Cheng Introduction Background The Average Cost Optimal Policy for a Repair/Replacement Problem Optimal Policy in the Discounted Cost Case Imperfect Preventive Maintenance Models 201 Toshio Nakagawa 1. Introduction Analysis of Policy 1 Analysis of Policy Analysis of Policy Conclusions Single Machine Inspection Problem 215 Moncer Hanga Mohammad A. AI-Fawzan 1. Introduction Notation, Assumptions, and Preliminary Results Infinite Horizon Model Finite Horizon Model with Continuous Inspection Times Finite Horizon Model with Discrete Inspection Time Random Horizon Model Conclusions A General Approach for the Coordination of Mainte- 245 nance Frequencies R. Dekker, R.E. Wildeman, J.B. G. Frenk and R. Van Egmond 1. Introduction Problem Definition Analysis of the Problem Solving Problem (Q) Numerical Results Conclusions Maintenance Grouping in Production Systems 283 Gerhard van Dijkhuizen 1. Introduction 283

7 viii MAINTENANCE, MODELING AND OPTIMIZATION Literature Review A Multi-Setup Multi-Component Production System Modeling Framework A Mixed Integer Linear Programming Formulation An Illustrative Example Concluding Remarks Part IV CONDITION BASED MAINTENANCE 13 On-Line Surveillance and Monitoring 309 Hai S. Jeong Elsayed A. Elsayed 1. Introduction Reliability Prediction and The Limit Value Determination in Condition Monitoring Vibration Monitoring Sound Recognition and Acoustic Emission Temperature Monitoring Fluid Monitoring Corrosion Monitoring New Approaches in Condition Monitoring Conclusions Maintenance Scheduling Using Monitored Parameter 345 Values Dhananjay Kumar 1. Introduction Reliability Models Proportional Hazards Model Proportional Intensity Model Accelerated Failure Time Models Some Other Models Selection of Models Maintenance Scheduling under the Age Replacement Policy Total Time on Test (TTT)-Plotting Example Maintenance Scheduling under the Block replacement Policy Example Conclusions 370 Part V INTEGRATED MAINTENANCE, PRODUCTION AND QUALITY MODELS 15 A General EMQ Model with Machine Breakdowns 377 V. Makis, X. Jiang and E. Tse 1. Introduction Model Description The Structure of the Optimal Policy An EMQ Model With Lost Sales Numerical Example 388

8 Contents ix 6. Model With Preventive Replacement After the First Production Run Conclusions Stochastic Manufacturing Systems: Production and 397 Maintenance Control E. K. Boukas Q. Zhang 1. Introduction Production Control Problem Production and Corrective Maintenance Control Production and Preventive Maintenance Control Numerical Methods and Examples Conclusion 428 Part VI MAINTENANCE & NEW TECHNOLOGIES 17 JIT and Maintenance 439 G. Waeyenbergh, L. Pintelon, and L. Gelders 1. Introduction The JIT-Philosophy Maintenance Meets Production From Preventive Maintenance to Productive Maintenance The TPM-Philosophy The Pillars of TPM The Need for Supporting Maintenance Concepts The Ideal Base to Start With TPM Modeling Tools for Improved Maintenance Illustrative case: Failure Analysis and TPM in a Brewery Conclusions 467 Index 471

9 List of Figures 2.1 Classification of maintenance planning and scheduling The relationship between the coefficient of variation K = a I J.L and shape parameter,,8 of the Weibull model [,8 = 11K] Failure rate function of Weibull modelfor,8 = 1,,8 < 1 and,8 > Probability density function of Weibull model for,8 = 1 (constant failure rate),,8 < 1 (decreasing failure rate) and,8 > 1 (increasing failure rate) Weibull reliability analysis of 54 extrusion eies life :die life Xi is expressed as Kg of metal extruded until failure One realization of renewal or counting process of part replacement Renewal function of Weibull model Renewal rate function of Weibull model Variance of number of renewals for Weibull model Illustration of inventory model and related variables Effect of normal demand during lead time Customer's optimal strategies Maintenance float network Metamodel construction procedure The optimal repair cost limit Cash flow structure over one cycle Graphical form of P (T, 00). 222

10 Xll MAINTENANCE, MODELING AND OPTIMIZATION 11.1 An example of the function gi (. ). The thin lines are the graphs of the functions <I>i (lit), <I>i (2/t),..., <I>i (5 It). The (bold) graph of gi( ) is the lower envelope of these functions An example of the objective function of problem (Q). There are three local minima: at Tl, T2 and T An example of the function g;r)(.). Notice the similarity with the graph of gi( ) in Figure An example of the objective function of problem (R) Maintenance initiators, activities, and packages: an overall perspective Tree-like structure of set-up activities and maintenance jobs (example) Multiple interrelated set-up and maintenance activities with associated costs for a tramcar (Sculli and Suraweera, 1979) How to create maintenance opportunities from maintenance frequencies ( example) Fatigue-crack-growth data from Bogdanoff and Kozin (1985) f(x; t) versus t for MIDP Ball bearing rotation Outer race defect. The ball pass frequency on the outer race is 172 Hz. The spectral line at 172 Hz is the representation of defect on the outer race of the bearing (Taylor, 1980) Signal analysis by spectrum Information for the AE event Prognostic and health-management system An illustration of the pattern of the log cumulative hazard rate The pattern of the hazard rate function assuming a good-as-new condition A TTT-plot to find the optimal replacement time intervals TTT-plot when parameter z= TTT-plot when parameter z=o TTT-plot when parameter z= An illustration of the block replacement policy with minimal repair. 363

11 List of Figures xiii 14.8 The optimal planned maintenance cycle length in hours The optimal planned maintenance cycle length in hours The plot is based on failure times only Long run average cost as a function of the lot size for several values of preventive replacement rate Ap Long run average cost as a function of the number of production runs before a preventive replacement for several values of preventive replacement rate Ap Optimal lot size as a function of the preventive replacement rate Ap Optimal number of production runs before a preventive replacement n* as a function of the preventive replacement rate Ap Long run average cost as a function of the lot size for several values of the failure rate A Long run average cost as a function of the number of production runs before a preventive replacement for several values of the failure rate A Evolution of the value function versus the stock level in mode 1 when the corrective maintenance is considered Evolution of the value function versus the stock level in mode 2 when the corrective maintenance is considered Evolution of the production rate versus the stock level when the corrective maintenance is considered Evolution of the corrective maintenance rate versus the stock level when the corrective maintenance is considered The bathtub curve The six big losses Problem solving approach with mathematical programming(mp) and simulation(sim) Steps of a simulation study. 463

12 List of Tables Unit Problem Data Weekly Peak Load and Gross Reserves Failure Modes and Wei bull Shape Parameter {3 [5] Loss matrix to determine ordering policies Full factorial experimental design plan Regression analysis of the maintenance flaot problem with no runs deleted Stability analysis of the maintenance float problem Statistical analysis of the maintenance float problem with statistical significance tests Validation of the maintenance float regression metamodel AEU values for defined intervals of Sand R obtained through the metamodel Cost table for the "Optimum" Rand S Optimal pm times T* Optimal replacement numbers N* Optimal inspection times T* for a and C2 when Cl = Effect of discounting on the inspection interval for (A=O.Ol) Effect of discounting on the inspection interval for (A=0.02) Effect of discounting on the inspection interval for (A=0.03) Effect of discounting on the equivalent uniform worth (A=O.Ol) Effect of discounting on the equivalent uniform worth (A=0.02). 229 xv

13 xvi MAINTENANCE, MODELING AND OPTIMIZATION 10.6 Effect of discounting on the equivalent uniform worth (A=0.03) Results for the finite horizon example Results of the dynamic programming procedure for the example problem Effect of the mean planning horizon on Tr and P(T, H) Values for the parameters in the four models Results of 4200 random examples for the special case of Goyal and Kusy,the minimal-repair model and the inspection model Percentage of the examples where the IFP and the algorithm of Goyal and Kusy generated deviations of more than 1% and 5% Results of 4200 random examples for the block-replacement model Example of a maintenance cycle for a production system consisting of m = 3 set-up activities and n = 3 components, where component j requires all set-up activities i ~ j to be carried out in advance Reductions in problem size for several period sets JC = {1,...,n} Maintenance opportunities, corresponding clusters and frequencies of occurence for the example of Figure The inter arrival times at which pressure gauges were replaced on failure or as part of planned maintenance (denoted by asterisks) Result of the PHM analysis Data from the cement plant, taken from Love & Guo (1991) Result of the proportional intensity analysis The optimal PM cycle length in hours for different values of the MPV B Notation Summary for the Modified EMQ model Numerical Data for Figures Simulation data Definitions and examples of the six big losses Comparison of five maintenance design concepts Definitions and examples of the six big losses Definitions and examples of the six big losses. 466

14 Preface Production costs have been coming down over time due to automation, robotics, computer integrated manufacturing, cost reduction studies and more. On the other hand, these new technologies are expensive to buy, repair and maintain. So the demand on maintenance is growing and maintenance costs are escalating. This new environment is compelling industrial maintenance organizations to make the transition from being repair departments for fixing broken machines to that of high level business units for securing production capacity. On the academic front, research in the area of maintenance management and engineering is getting tremendous interest from researchers. Many papers appeared in the literature dealing with the modeling and solution of maintenance problems using operations research (OR) and management science (MS) techniques. This area represents an opportunity for making significant contributions by the OR and MS communities. This book provides in one volume the latest developments in the area of maintenance modeling. Prominent scholars contributed chapters covering a wide range of topics. We hope that this humble contribution will serve as a useful informative introduction to this field that may permit additional developments and useful directions for more research in this fast growing area. This volume is divided into six parts and contains seventeen chapters. Each chapter has been subject to review by at least two expert. in the area of maintenance modeling and optimization. In the first chapter which forms Part I, M. Ben-Daya and S.O. Duffuaa provide an introduction to major maintenance modeling areas illustrated with some basic models. Areas covered include maintenance planning and scheduling, spare parts provisioning, preventive and con-

15 XVlll MAINTENANCE, MODELING AND OPTIMIZATION dition based maintenance, and integrated maintenance, production, and quality models. Directions for future research are outlined. Part II contains five chapters (Chapters 2 to 6) dealing with maintenance planning and scheduling. In Chapter 2, S.O. Duffuaa presents an overview of maintenance planning and scheduling models. Models presented are mathematical programming based models. 1. El-Amin focuses on electric generators maintenance scheduling in Chapter 3. He presented a general model and elaborated on various solution techniques. Chapter 4 by D.N.P. Murthy is about maintenance service contracts. This chapter provides an overview of the literature on maintenance service contracts, the issues involved and the models used to study these issues. In Chapter 5, A. Sheikh, M. Younes, and A. Raouf address issues related to spare parts procurement strategies based on reliability engineering. Simulation meta-modeling of the maintenance float system by C. Madu is the object of Chapter 6. Part III deals with preventive maintenance in six chapters (Chapters 7-12). Preventive maintenance policies and their variations are discussed in Chapter 7 by T. Dohi, N. Kaio, and S. Osaki. Important preventive maintenance optimization models, which involve age replacement and block replacement, are reviewed in the framework of the well known renewal reward argument. A general framework for analyzing maintenance policies is discussed in Chapter 8 by V. Makis, X. Jiang, A.K.S. Jardine, and K. Cheng. A wide class of maintenance decision problems are analyzed using the optimal stopping framework. The necessary mathematical background from martingale dynamics and the optimal stopping theory are summarized and the approach is illustrated by analyzing a general repair/replacement model. Chapter 9 by T. Nakagawa deals with imperfect preventive maintenance. Three maintenance policies are analyzed. In Chapter 10, a single machine inspection problem is discussed by Hariga and Al-Fawzan. The inspection problem is formulated using discounted cash flow under different assumptions regarding the length an nature of the planning horizon. A general approach for the coordination of maintenance frequencies in cases with a single set-up is presented in Chapter 11 by R. Dekker. An optimal approach for solving the multi-component maintenance model is discussed. The approach is also extended to deal with more general maintenance models like minimal repair and inspection, and block replacement. The last chapter of Part III deals with maintenance grouping in multi-setup multi-component production systems. A general mathematical modeling framework is presented for this class of problems. Part IV focuses on condition based maintenance and contains two chapters. The first one, Chapter 13, deals with on-line surveillance and

16 PREFACE XiX monitoring. The second one, Chapter 14, presents an approach for maintenance scheduling using monitored parameter values (MPV). One of the advantages of the discussed approach is that the importance of the MPV in explaining the failure characteristics can be identified without making assumptions about the distribution of the hazard rate. The approach can be used for both repairable and irreparable systems. Part V deals with integrated production and maintenance models and contains two chapters. Chapter 15 by V. Makis, X. Jiang, and E. Tse considers a general economic manufacturing quantity problem with machine breakdowns. The structure of the optimal policy is obtained using semi-markov decision processes. In Chapter 16, E.K. Boukas and Q. Zhang deal with production and maintenance control for manufacturing systems using continuous flow models. Various types of models that incorporate production, corrective maintenance, and preventive maintenance are described. Part VI addresses issues related to maintenance and new technologies and contains one chapter by G. Waeyenbergh, L. Pintelon, and L. Gelders dealing with Just-In-Time (JIT) and Maintenance. The chapter describes the consequences of the JIT philosophy on maintenance. THE EDITORS

17 Acknowledgment The Editors would like first to acknowledge the authors for their valuable contributions. This book would not have been in this form without their participation and cooperation throughout the stages of this project. We are also grateful to all the referees who reviewed the material submitted for this book. The critique and suggestions they made have definitely lead to substantial improvements. It takes a lot of skill and patience to do all the word processing, text conversion, and type-setting, and our thanks go to Shahid Pervez and Arifusalam for their support. We are also grateful to Mr. John R. Macleod for his assistance with English editing. We are indebted to Gary Folven, Publishing Editor OR/MS areas at Kluwer for his interest in this book, full cooperation and assistance since the initiation of this project. Carolyn Ford and other Kluwer staff members have been very helpful. We would like also to express our gratitude to our families for their patience and encouragements. Preparation of this book has been sometimes at the expense of their time. Finally, the Editors would like to acknowledge King Fahd University of Petroleum and Minerals for excellent facilities and for funding this project under the number SE/MODEL/199.

18 Contributing Authors M. Ben-Daya & S. O. Duffuaa Dept. of Systems Engineering, King Fahd Univerity of Petroleum and Minerals, Saudi Arabia V. Makis, X. Jiang, E. Tse & A.K.S. Jardine Department of Mechanical and Industrial Engineering, University of Toronto, Canada Dhananjay Kumar Nokia Svenska AB, Sweden Christian N. Madu Lubin School of Business, Pace University, New York, USA. D. N. P. Murthy The University of Queensland Australia Ibrahim EI-Amin Dept. of Electrical Engineering King Fahd Univerity of Petroleum and Minerals, Saudi Arabia A. K. Sheikh & M. Younas Dept. of Mechanical Engineering King Fahd Univerity of Petroleum and Minerals, Saudi Arabia Abdul Raouf Adjunct Professor King Fahd Univerity of Petroleum and Minerals, Saudi Arabia T. Dohi & S.Osaki Dept. of Industrial and Systems Engineering, Hiroshima University. Japan N. Kaio Dept. of Economic Informatics, Hiroshima Shudo University, Japan K. Cheng Institute of Applied Mathematics, Academia Sinica, China. Toshio Nakagawa Dept. of Industrial Engineering, Aichi Institute of Technology Japan

19 XXIV MAINTENANCE, MODELING AND OPTIMIZATION Moncer Hariga Industrial Engineering program King Saud University Saudi Arabia. Mohammad A. AI-Fawzan King Abdulaziz City for Science and Technology Saudi Arabia. R. Dekker, R. Wildeman, J. Frenk, and R. Egmond Econometric Institute, Erasmus University Rotterdam, The Netherlands. G. van Dijkhuizen University of Twente The Netherlands. Hai S. Jeong Dept. of Applied Statistics Seowon University, Korea E. K. Boukas Mechanical Engineering Dept, Ecole Poly technique de Montreal, Montreal, Quebec, Canada Q. Zhang Department of Mathematics, University of Georgia, Georgia, USA G. Waeyenbergh, L. Pintelon, & L. Gelders Center for Industrial Management Belgium.