A modelling methodology for the assessment of preventive maintenance on a compressor drive system

Huge rotary machines are commonly used in oil and gas processing plants for separation, compression and boosting. Their reliability is of high importance to avoid operation downtime and production loss. In this paper, we present a modelling methodology, based on the AltaRica 3.0 modeling language and stochastic simulation, to assess the average production level of a compressor drive system. This system consists of six trains, where each of them contributes to one sixth of the total production capacity. It runs under two operation modes (full and reduced capacity) corresponding to seasonal demand periods (winter and summer). The problem at stake is to design a model at system level that captures the various degradation processes, monitoring policies, and maintenance rules involved in the system under study. The aging of units is represented by means of multiple degradation levels. Given units information provided by monitoring and inspection, preventive and corrective maintenance interventions are decided locally to each unit. Performance indicators such as the cumulative production and production loss over a certain mission time can then be assessed. This paper contributes to the development of engineering models for maintenance assessment based on framework and patterns designed to architecting some typical oil and gas systems.