Robust Model Predictive Control Applied to an ESP-lifted Oil Well System

  • Bruno A. Santana Mechatronics Program, Polytechnic School, Federal University of Bahia, BA
  • Victor S. Matos Mechatronics Program, Polytechnic School, Federal University of Bahia, BA
  • Daniel D. Santana Department of Chemical Engineering, Polytechnic School, Federal University of Bahia, BA
  • Márcio A. F. Martins Mechatronics Program, Polytechnic School, Federal University of Bahia, BA; Department of Chemical Engineering, Polytechnic School, Federal University of Bahia, BA
Keywords: Artificial Lift Methods, Electric Submersible Pump (ESP), Robust Model Predictive Control, Robust Control

Abstract

This work presents the first robust model predictive control approach for controlling an Electric Submersible Pump (ESP) lifted oil well system considering its benefits and operational envelope constraints. The proposed scheme is based on a robust infinite horizon model predictive controller (RIHMPC) with multi-model formulation as the uncertainty description and zone control scheme to explicitly consider the time-varying ESP (downthrust and upthrust) envelope constraints. The proposed control strategy is tested through simulation for the disturbances commonly found in the oil wells with ESP installations and (nonlinear) plant/model mismatch scenarios. Results show an applicable formulation capable of accommodating nonlinearities in the form of uncertainties and a low computational cost, with characteristics suitable for real- time applications.
Published
2022-10-19
Section
Articles