A simulation analysis of the vari-metric repairable inventory optimization procedure for the U.S. Coast Guard

This paper documents a simulation study undertaken to gain insights into the vari-metric multi-echelon repairable inventory optimization method. The method was analyzed in the context of the Coast Guard's fleet of fixed and rotary wing aircraft, for which operational availability is the key performance metric. Failure rates of parts in this system exhibit variance-to-mean ratios higher than one, and we outline a procedure for generating a failure arrival process described by a negative binomial distribution. Previous studies of the vari-metric model examined a single repairable part; in this study we analyzed a system comprised of three repairable parts. This allowed us to gain insights into how the vari-metric procedure selects part stock levels to attain a desired level of system availability. Further analysis of the simulation model allowed us to examine the efficient frontier for this multi-criteria problem (maximizing spare part availability rates while minimizing the cost of part inventories).