Title: A robust global optimisation framework for stochastic integrated refinery planning with demand and price uncertainties

Authors: Mahmud R. Siamizade; Theodore B. Trafalis

Addresses: School of Chemical, Biological and Materials Engineering, University of Oklahoma, Norman, Oklahoma 73019, USA ' Department of Industrial and Systems Engineering, University of Oklahoma, Norman, Oklahoma 73019, USA

Abstract: This study presents a global optimisation framework for the strategic planning of an integrated oil refinery while accounting for the uncertainties in the crude oil and final product demand and market prices. To model the uncertainties, three modelling schemes have been utilised: 1) a robust optimisation framework; 2) a fuzzy possibilistic programming approach; 3) a risk aversion two-stage stochastic model with financial risk management. These methodologies are applied to an industrial case study, comprising an integrated oil refinery with tasks spanning from unloading crude oil at the refinery's front to distributing refined products to distribution centres by pipelines. The notable feature of this study is developing an optimisation approach with a multi-period mixed-integer nonlinear programming (MINLP) model and solving it through an aggregation/disaggregation linearisation algorithm and obtaining ε-global optimal solutions. The results indicate significant economical and operational advantage of the robust optimisation scheme depending on the decision maker's risk attitude.

Keywords: integrated refinery planning; global optimisation algorithms; uncertainty appraisal; robust programming; fuzzy possibilistic programming; stochastic programming; financial risk management; mixed-integer nonlinear programming; MINLP.

DOI: 10.1504/IJMOR.2022.126044

International Journal of Mathematics in Operational Research, 2022 Vol.22 No.4, pp.496 - 527

Received: 12 May 2021
Accepted: 04 Jun 2021
Published online: 10 Oct 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article