Reliability, Availability and Maintainability (RAM) Analysis for the Evaluation of Process Facilities

摘要

Plant reliability has become a key economic consideration for operating plants, as well as those new plants in concept development or in early design. Owners of operating plants use reliability modeling as a tool to identify availability bottlenecks, maintenance requirements, and sparing levels. Plants in the design phase use reliability modeling to verify a new design (frequently in conjunction with a conventional process design analysis) and to test the limits of a design in availability terms, before equipment commitments are made. Reliability, availability, and maintainability (RAM) analyses utilize system engineering and discrete event simulation techniques to predict the characteristics of a production, process, or manufacturing facility. Equipment and event data is input into simulation software, which then generates statistical distribution of availability results using a Monte Carlo simulation engine. RAM analysis results provide owners with reasonable statistical bases for business decisions by defining a facility's total performance over the facility life. Typical RAM results define the probability of achieving desired Availability or production targets, and provide criticality rankings of events which contribute to overall process losses. This paper discusses the above methodologies using actual case studies of reliability analysis either for operating plants or new plant designs. As economics of ammonia plants, both older and new plants designs, become more competitive, the authors deem the above tools to be important more than ever.