(313a) Data-Driven Decarbonization of an Integrated Gas-Oil Separation Network Under Uncertainty

Nowadays, the energy sector is considered a major contributor to global emissions. To achieve their committed sustainability targets, conventional energy companies are exploring various opportunities throughout the supply chain to decarbonize their operations while maintaining financial competitiveness. This can be achieved by employing cutting-edge process systems engineering tools to find the optimal balance between sustainability and profitability.

This work focuses on the upstream sector of a typical oil and gas company, where a cluster of producing wells supplies a network of interconnected gas-oil separation (GOSP) facilities. The primary function of these facilities is to separate the three phases and remove contaminants. Additionally, they play a crucial role in eliminating flaring and greenhouse gas emissions by recovering associated gases.

Previously, efforts by scientists and practitioners to optimize network operations often prioritized profitability over sustainability objectives, relying on either simplified or computationally expensive models. We have recently proposed a novel deterministic approach that incorporates a machine learning-based surrogate model within a multi-objective mixed integer linear programming (MILP) framework to enhance network operations and reduce emissions. This work builds on the deterministic model to account for operational and market-related uncertainties, thus effectively enabling decision-making processes over an extended time horizon.

The proposed methodology has been validated using real-world industrial production scenarios, demonstrating its ability to significantly reduce greenhouse gas emissions while maximizing profitability.