Scenario Based Two-Stage Stochastic Programming Approach for the Midterm Production Planning of Oil Refinery

Abstract

Recently, oil refining industry is facing with lower profit margin due to uncertainty. This causes oil refinery to include stochastic optimization in making a decision to maximize the profit. In the past, deterministic linear programming approach is widely used in oil refinery optimization problems. However, due to volatility and unpredictability of oil prices in the past ten years, deterministic model might not be able to predict the reality of the situation as it does not take into account the uncertainties thus, leads to non-optimal solution. Therefore, this study will develop two-stage stochastic linear programming for the midterm production planning of oil refinery to handle oil price volatility. Geometric Brownian motion (GBM) is used to describe uncertainties in crude oil price, petroleum product prices, and demand for petroleum products. This model generates the future realization of the price and demands with scenario tree based on the statistical specification of GBM using method of moment as input to the stochastic programming. The model developed in this paper was tested for Malaysia oil refinery data. The result of stochastic approach indicates that the model gives better prediction of profit margin.