Aplicação de estrutura de detecção e diagnóstico de falhas a um sistema de produção de petróleo com bombeio centrífugo submerso: uma abordagem baseada em observadores de estado com entrada desconhecida

Abstract

This article presents an application of a system for detection, isolation and identification of abnormalities and faults in an oil well simulation environment with Electric Submersible Pump (ESP) installations. The proposed method considers a linear reduced-order Unknown Input Observers (UIO) system configured in a Generalized Observer Scheme (GOS). The results were obtained from the simulation of abnormal events with component failure associated with the reservoir flow model parameter and the production choke valve that controls the column production. In both cases, plant-model mismatch scenarios were considered so as to evaluate the robustness of the proposed method, namely the response against a non-linear model and a model with noise introduction on system measurements. The proposed system proved to be capable of detecting the occurrence of abnormalities and was able to identify the abnormal modeled events.