A TWO-STAGE STOCHASTIC PROGRAMMING MODEL FOR THE MID-TERM OIL REFINERY PLANNING UNDER UNCERTAIN DEMAND

This work proposes a two-stage stochastic mixed-integer programming model to deal with the mid-term refinery production planning problem. The model considers that decisions regarding the amount of refinery input to be processed in two production periods must be taken in a first stage, where complete and accurate knowledge of the market demands is not available. In a second stage, decisions regarding flow balancing among units, stock levels and production are taken. The challenge lies in defining the amounts of the different inputs that maximize the expected profit of the refinery.