Data-driven control and optimization of oil and gas production systems

Post Doctoral fellow Thiago Lima

Main Supervisor Alexey Pavlov

Sponsor: NTNU

This project deals with data-driven control and optimization methodologies for oil- and gas production systems in different time scales. On a long-term scale, the proposed approach combines simulation models with data-driven optimization to deal with unavailable gradient information and parametric uncertainties. On a shorter time scale, the adopted approach is a data-driven automatic optimization method referred to as Extremum-Seeking Control (ESC),
which allows one to achieve automatic optimization of steady-state behavior of an unknown plant. The methods developed in this project support a number of other BRU21 projects.

Project result: An improved method for optimal gas-lift allocation using automatic well testing 
Ensuring a smooth transition, maximizing total oil rate and guaranteeing constraints are met at all times  

BRU21 Conference 2022