Risk based maintenance strategy: a quantitative approach based on time-to-failure model

摘要

The appropriate maintenance strategy for a specific system requires a detailed analysis of the system’s reliability. Reliability analysis relies greatly on the historical failure and repair data. Lack of data, poor quality data, data censoring, and combining data from different systems working in different operating conditions are some of the challenges for maintenance analysts. Hence, based on the nature of historical failure and repair data (independence of data, constant failure rates, identically distributed variables, etc.), an appropriate statistical model should be selected to predict the reliability of the system. Thereafter, based on the defined reliability characteristics of the system, an appropriate maintenance strategy needs to be established. In this study, a systematic methodology is developed, based on the available guidelines for maintenance strategy selection, founded on the reliability analysis of historical failure data. The application of the methodology is shown by a case study from the crushing and mixing bed hall unit at a cement factory.