(0c) Using Combined Failure Data Sources to Assess Equipment Failure Rates

Understanding asset reliability is of the utmost importance to manage and maintain your assets. This is even more crucial when the equipment is critical to operations and/or related to safety systems. The challenge in estimating failure rates of equipment is to identify the proper and most relevant data to support the assessment. Examples of available data sources include generic data, and operator and vendor failure data. Generic data has the advantage of being proven and widely accepted, but is criticized for not being specific enough to the asset under evaluation or to the application. Alternatively, vendor and operator data is praised for being specific, but the cleanliness of the data and the sample size can be challenged. This presentation will discuss the pros and the cons of the different sources of failure data and approaches for assessing failure rates. The presentation will illustrate a methodology for combining different data sources, as a trade-off for failure rate estimations by using a Bayesian framework. The methodology may be used for estimating an asset’s failure rate, which will in turn allow the stakeholders to make informed decisions about asset selection and maintenance, and to manage equipment failures and related risks. Ultimately, the benefits of the proposed methodology are the decisions made based on the analysis that can lead to increased equipment reliability, system availability, and production uptime.