Optimization Strategies: Petroleum Refinery Planningunder Uncertainty

This work proposes a hybrid of stochastic programming(SP) approaches for an optimal midterm refineryplanning that addresses three forms of uncertainties:prices of crude oil and products, demands, andyields. An SP technique that utilizes compensatingslack variables is employed to explicitly account forconstraint violations to increase model tractability.Four approaches are considered to achieve solutionand model robustness: (1) the Markowitz'smean-variance (MV) model to handle randomness in theobjective coefficients by minimizing the variance(economic risk) of the expected value of the randomcoefficients; (2) the two-stage SP with fixedrecourse to deal with randomness in the RHS and LHScoefficients of the constraints by minimizing theexpected recourse costs; (3) incorporation of the MVmodel within the framework in (2) to formulate amean?risk model that minimizes the expectation andthe operational risk measure of variance of therecourse costs; and (4) reformulation of the model in(3) by adopting mean-absolute deviation (MAD) as theoperational risk measure imposed by the recoursecosts for a novel refinery planning application. Anumerical example is illustrated.