An Adaptive Infinite Horizon Model Predictive Control Strategy Applied to an ESP-lifted Oil Well System

Abstract

This work presents the benefits of using an adaptive model predictive control approach for controlling an ESP-lifted oil well system. The idea behind the proposed scheme is the successive linearization of the plant nonlinear model and incorporate it into an infinite-horizon MPC formulation, which handles feasibility issues by considering slacked terminal constraints. Also, suitable use of a zone control strategy to deal with time-varying ESP operating envelope constraints (downthrust and upthrust) becomes the approach implementable in practice. Nonlinear plant-model mismatch scenarios, including simulation of unmeasured disturbances and the tracking of ESP operation economic targets, show the effectiveness of the proposed controller concerning different operating conditions of the oil production process equipped with ESP.