On the robustness of ‘insurance type’ spares provisioning strategies

‘Insurance type’ spares are spares for unique parts which are essential to the operation of the equipment in which they serve, and which have an appreciable probability of not needing replacement during the lifetime of the equipment. They require a stocking strategy which is dependent on time and the actual occurrence of failures. The values of parameters in even the simplest of models of the dynamic spares stocking decision problem are usually known with great uncertainty, if at all. For one such model the sensitivity of (near-)optimal strategies to variations in the parameter values is investigated.