development and application of an importance measures for predictive maintenance in oil and gas installations

A Prognostics and Health Management System (PHMS) provides information on the degradation state of an industrial component and the prediction of its Remaining Useful Life (RUL). In this thesis work, we consider a complex industrial plant made by several components which can be in multiple states. The main objectives are: 1) To propose a method to identify which components should be equipped with a Prognostics and Health Management System; 2) To obtain the best impact on the overall plant performance and to quantify the expected impact. To this aim, we proposed a modification of the risk achievement worth importance measure which allows prioritizing components with respect to the expected benefits of equipping them with a PHMS. The computation of the proposed importance measure has required: 1) The definition of a novel method based on the simulation of the Prognostics and Health Management System predicting Remaining Useful Life and characterized by a given performance for estimating the average component availability; 2) The embedding of the method within a Monte Carlo simulation approach for the computation of the overall plant performance An application to an oil and gas offshore installation is proposed. The components are prioritized according to the expected benefits of equipping them with a Prognostic and Health Management System taking into account variabilities in Prognostic and Health Management System performance. Furthermore, the benefits of equipping a component with a Prognostic and Health Management System in term of overall plant performance is discussed.