Determining critical power equipment inventory using extreme value approach and an auxiliary Poisson model

This paper is devoted to the establishment of reasonable levels of critical power equipment inventory using available historical data and strict legal requirements for reliable electrical power supply. Several models, including an Extreme Value Theory and a Homogeneous Poisson model, are used with results compared to provide a probabilistic formulation of suggested inventory. Several procedures are provided to establish proper inventory levels of critical power equipment. Simplifications are suggested for automating the calculation process for large samples. The presented approaches have a wide scope of industrial applicability, and can be applied to different types of critical equipment. Mathematical background is provided to assist practitioners that do not have extensive knowledge of probabilistic tools such as Extreme Value Theory or Homogeneous Poisson Process models.