The Reliability Data Handbook is exceptional in both its approach and coverage, giving a uniquely comprehensive account of the subject. Component failure rate data are a vital part of any reliability or safety study and highly relevant to the engineering community across many disciplines. The Reliability Data Handbook focuses on the complete process of data collection, analysis and quality control. The subject of reliability data is covered in geat depth, reflecting the author's considerable experience and expertise in this field. Rarely is reliability data 'clean'. There are problems in its collection due to poor recording of the exact cause of failure. The time to failure of many items in the data collection scheme may not be known due to censoring and truncation issues. For new equipment or very reliable equipment there may be no data available at all. All of these practical facets of data used in real--world reliability studies are contained in this book. Analysis methods are not presented in a clinical way -- they are put into context, considering the difficulties that can arise when performing assessments of actual systems.
A unique feature is that the text is always fully illustrated with worked examples, many from the author's own experience in industry. CONTENTS INCLUDE Introduction -- objectives and scope Component, equipment and system reliability, reliability models for non--repairable and repairable equipment Failure patterns, failure rate calculations Discreet and continuous distributions, scatter diagrams and histograms The Weibull distribution and Weibull analysis The Duane reliability growth model Reliability forecasting, basic models, data requirements Reliability statistics, sources of generic data, tables of equipment and component failure rates Data collection and quality control Reliability testing, inspection and maintenance decision analysis.
Bob Moss is a consulting engineer with over 30 years' experience in safety, reliability, and risk analysis. Prior to forming his own company in 1979 he was a Principal Engineer with the UKAEA Safety and Reliability Directorate with responsibility for the group that developed methods for the availability assessment of large, potentially hazardous chemical plant. Since 1979 he has been responsible for a wide range of projects concerned with nuclear installations, petrochemical plant, and offshore and onshore oil and gas platforms. He was one of the principal architects of the OREDA (Offshore Reliability Data) project and supervised the collection and analysis of reliability data from the UK sector of the North Sea. Bob is a past Vice Chairman of the Institution of Mechanical Engineers Process Industries Division, past Chairman of its Mechanical Reliability Committee and an Honorary Member of the European Safety Reliability and Data Association. For 8 years he was a Visiting Research Fellow in the Department of Mathematical Sciences at Loughborough University and was awarded a PhD for research into rotating machinery reliability. He has lectured extensively worldwide and has published over 50 papers on reliability and risk assessment. His current interests focus on the application of reliability and risk assessment methods to maintenance and asset management.
Acknowledgements xiii; Preface xv; Notation xvii; Chapter 1 Fundamentals 1; 1.1 Introduction 1; 1.2 Quantified reliability data 2; 1.3 Collecting the data 2; 1.4 Preliminary studies 3; 1.5 Presenting the results 5; 1.6 Summary 7; Chapter 2 Basic Concepts 9; 2.1 Reliability performance 9; 2.2 Summary 12; Chapter 3 Component, Equipment and System Reliability 13; 3.1 Component reliability 13; 3.2 Equipment reliability 15; 3.3 System reliability 17; 3.4 Summary 21; Chapter 4 Failure Data and Failure Models 23; 4.1 Introduction 23; 4.2 Preliminary statistical analyses 24; 4.3 Bar charts 24; 4.4 Histograms 25; 4.5 Frequency curves 27; 4.6 Cumulative frequency diagrams 27; 4.7 Summary statistics 28; 4.8 Measures of location 28; 4.9 The median 29; 4.10 The mode 30; 4.11 Measures of spread 31; 4.12 Variance and standard deviation 32; 4.13 Coefficient of variation 35; 4.14 Summary 36; Chapter 5 Discrete and Continuous Distributions 37; 5.1 Introduction 37; 5.2 Discrete distributions 37; 5.3 The binomial distribution 38; 5.4 Mean and variance of the binomial distribution 40; 5.5 Variance of a binomial distribution 42; 5.6 The Poisson distribution 43; 5.7 Continuous distributions 44; 5.8 Continuous failure time models 47; 5.9 Summary 57; Chapter 6 Patterns of Failure 59; 6.1 Introduction 59; 6.2 Failure patterns 60; 6.3 Quantile plots 61; 6.4 Symmetry 63; 6.5 One-dimensional scatter plots 65; 6.6 Box plots 65; 6.7 Stem and leaf diagrams 66; 6.8 Symmetry plots and transformations 68; 6.9 Density plots 71; 6.10 Review of component failure patterns 74; 6.11 Repairable system failure patterns 74; 6.12 Trend analysis 79; 6.13 Reliability growth 84; 6.14 Duane model 86; 6.15 Duane/AMSAA model 88; 6.16 Summary 91; Chapter 7 Estimating Distribution Parameters 93; 7.1 Introduction 93; 7.2 Theoretical quantile-quantile plots 94; 7.3 Probability papers 96; 7.4 Weibull probability paper 99; 7.5 Summary 102; Chapter 8 Weibull Analysis 105; 8.1 Introduction 105; 8.2 Graphical Weibull analysis 108; 8.3 Complete samples 110; 8.4 Example of a 'complete' graphical Weibull analysis 112; 8.5 Example of Weibull analysis: left-truncated dataset 116; 8.6 Hazard plotting 117; 8.7 Time-terminated right-censored example 119; 8.8 Failure-terminated right-censored Weibull analysis 120; 8.9 Weibull multiple-censored example 121; 8.10 Curved and bent plots 123; 8.11 Analytical point estimates 127; 8.12 Outliers 129; 8.13 Goodness-of-fit tests 130; 8.14 Confidence limits 134; 8.15 Summary 137; Chapter 9 Repairable Systems 139; 9.1 Introduction 139; 9.2 Equipment failure characteristics 140; 9.3 Monitoring reliability performance 142; 9.4 Reliability growth 145; 9.5 Analysis of repair times 148; 9.6 Summary 153; Chapter 10 Generic Reliability Data Requirements 155; 10.1 Introduction 155; 10.2 System definition 156; 10.3 Failure modes and failure causes 171; 10.4 Summary 173; Chapter 11 Generic Reliability Data Sources 175; 11.1 Introduction 175; 11.2 Generic data analysis 176; 11.3 Common mode failures and human error 191; 11.4 Tabulated generic data 193; 11.5 Bayesian updating 196; 11.6 Summary 200; Chapter 12 The No-Data Problem 203; 12.1 Introduction 203; 12.2 Paired comparisons 203; 12.3 The DELPHI method 205; 12.4 Morphological analysis 210; 12.5 Failure mode and effects analysis 216; 12.6 Summary 220; Chapter 13 Data Collection and its Quality Control 223; 13.1 Introduction 223; 13.2 In-service reliability data 224; 13.3 Data quality assurance 232; 13.4 Corrective actions 240; 13.5 Summary 241; Appendices 243; Glossary 275; Index 283 A availability; As system availability; A(t) availability function; CMTBF cumulative mean time between failures; C coefficient of variation; dN(t) cumulative failure to time t; dt infinitesimal small interval; du infinitesimal small interval; Dcrit KS test statistic; Dmax largest difference (KS test); F cumulative failures; probability that an item has failed; Fs probability that a system is in a failed state; FE(t) probability of equipment failure before time t; Fl theoretical frequency in interval i; F(xl) observed frequency in interval i; f/Mh failures per million hours; f/year failures per year; FDT fractional dead time; f(t) failure probability density function; f(u) probability density function; H(t) cumulative hazard function; h(t) hazard rate function (non-repairable items); i failure number; kl stress factor for stress i; MOl mean order number; MR median rank; MTBF mean time between failures; MTTF mean time to failure; MTTR mean time to repair; m expected number of failures; m(t) conditional repair rate (repair process only); maintainability PDF; M(t) maintainability function; n number of failed items in a sample; N number of items in a population; N(t) cumulative failures before t; N(T) total number of failures; P probability of an event; PES fail to start probability; PF failure probability; Pw(t) probability that a component or system is working at time t; P[mu] estimate of percentage failed at mean life; R reliability; R(t) reliability function; probability of survival to time t; RE(t) equipment reliability function; RS(t) system reliability function; S number of survivors or sample space; Sl survival time; Sn CUSUM; t time; aggregate system life; tl ith ordered age at failure; interval between (i - 1)th failure and ith failure; tk largest ordered age at failure; t0 initial time; start time of surveillance; t1 smallest ordered age at failure; t2 second smallest age at failure; t* transformed age at failure; t mean life; T total elapsed time; CUSUM target; U unavailability; U(t) unavailability function; x number of failures; X mean weight; yl system age at ith failure; z normal (0, 1) random variable; z(t) failure rate function (repairable items); Z(t) expected number of failures in time t; [alpha] proportion of failures in specific failure mode (FMEA); Duane model slope; [alpha]A proportion of failures in mode A; [beta] ratio of common-cause failure rate to the total failure rate; Weibull shape parameter; Duane AMSAA model slope; [gamma] Weibull location constant; [GAMMA] gamma function; [eta] Weibull scale parameter (characteristic life); [theta] test or inspection interval; [theta]c cumulative MTBF; [kappa] number of intervals; [lambda] failure rate or hazard rate; [lambda]E equipment failure rate; [lambda](t) hazard rate or failure rate function; [mu] true population mean; [mu](t) repair rate function; II product symbol; [rho] number of parameters; [rho](t) DA model failure intensity function; [sigma] mean deviation of a random variable or of a population; [SIGMA] summation symbol; [tau] inspection interval; [PHI](t) distribution function