Plant Reliability Prediction, Analysis and Modeling Engineering

High plant reliability is critical for every successful company, and it has never been more important than it is in the present economic climate. The costs associated with equipment downtime and reduced production can be significant, and engineers must ensure that you are using every possible means of maximizing plant reliability and performance. Of the five fundamental ways in which engineers can approach the maintenance of plant, one of the least commonly used (because it is least commonly understood) is Reliability Centred Maintenance (RCM).

The heart of an RCM approach is the creation and exploitation of reliability models that use previous failure data to predict future plant performance and hence permit the selection of a maintenance strategy and frequency optimization of planned maintenance activities. Reliability modeling as part of an integrated maintenance strategy is an approach that can no longer be sidelined or ignored by high performing companies.

This course is a combination of instructor-led topic areas and extensive computer-based analysis and modeling. You will learn in detail about, and practice using, best-of-breed approaches to statistical failure data analysis and reliability modeling. Furthermore, throughout the course, you will have the opportunity to analyze your own data and to ask lots of questions about how best to apply reliability analysis and modeling techniques in your organization.

Module 1: Maintenance strategies and the power of historical data

• Fundamental approaches to maintenance
• Formulating a maintenance strategy
• The importance of maintenance history records
• Understanding plant performance
• An introduction to the statistical analysis of failure data
• The principles of failure data analysis
• Industry-standard measures of reliability (Availability, MTBF, MTTR, etc)
• Extensive hands-on experience

Module 2: Statistical analysis of failure data

• Pareto analysis, rank order charts and standard deviation
• Linear regression models and determining model accuracy
• Failure mode analysis
• Interpreting failure mode shapes
• Extracting failure mode shapes from real data
• Optimizing PM activity using mode shape analysis
• Knowing when to use a breakdown maintenance approach
• Extensive hands-on experience

Module 3: Reliability models and approaches to modeling

• The principles of RCM and reliability modeling
• Developing a reliability model
• Weibull statistics and the range of Weibull models (2 parameters, 3 parameters, maximum likelihood, maximum accuracy)
• The Weibull curve and plotting data on a Weibull scale
• Defining parameters: shape, scale, mean life, minimum life, characteristic life, standard deviation
• Model accuracy assessment (observed model accuracy and hypothesis rejection)
• Interpreting model results
• Confidence levels and Weibull critical values
• Key graphical functions:
• The reliability function: survival probability
• The cumulative distribution function
• The failure probability density function
• The failure rate function
• Extensive hands-on experience

Module 4: Cost based maintenance and the basis of a reliability toolbox

• Converting reliability model data into cost-based maintenance decisions
• Optimizing PM activity based on cost and by using reliability predictions (note that the course will NOT cover the costing of maintenance activities, but will assume that this information is already known)
• Calculating the cheapest PM interval for age-based replacement policies
• Graphing costs versus PM interval
• Predicting future failures
• Predicting spares utilization
• Development of the key components of a reliability toolbox
• Extensive hands-on experience
• Open discussion

Module 5: The finalization of a comprehensive reliability toolbox in Excel

• The cost of maintenance convenience and making informed maintenance optimization decisions
• Incorporating real-world effects within reliability models
• Specifying the PM interval and understanding the implications of doing this
• Completing the reliability toolbox
• Graphing toolbox results
• Toolbox testing and comparison of results with best-of-breed modeling software
• Extensive hands-on experience
• Overall review of concepts learned and how they can be applied in practice