Optimal maintenance, repair, and replacement for Great Lakes marine diesels

Reliability-based models are developed and implemented to rationalize current winter layup replacement practices for one-engine and two-engine Great Lakes marine diesel ships. A semi-Markov competing-process approach is used in our models, where the age-dependent system failure behavior is treated as a race among engine components. Howard's one-set competing process model is implemented and extended to two sets of competing processes. An efficient enumerationprocedure is presented to select the replacement policy which produces the minimum expected cost for the operating season. Computer codes are developed using the above models, and several examples are considered. Sensitivity analyses are performed for several parameters for which we have insufficient or no information from the industry to see their influence on the minimal expected costs and corresponding replacement policies.