(531b) A New Integrated Scheduling and Optimization Framework for Holistic Refinery Supply Chain Management

The holistic refinery supply chain management is critically important over the whole oil industry supply chain. It covers three highly correlated sub problems: the front-end sub problem, the refinery manufacturing sub problem and the multi oil-product pipeline distribution sub problem. According to Chima (2007), there is a need to ensure that each sector along the supply-chain can respond quickly to the exact demand of its downstream customers, protecting itself from problems with suppliers and buffering its operations from the demand and supply uncertainty it faces. Since the oil companies are aiming at maximizing the profit margin by weighing benefits versus costs throughout the entire refinery supply-chain. Thus, by coordinating the management and operations of the three sub problems along the holistic refinery supply-chain, it can greatly maximize the potential benefit margin. Although there exists a lot of works explicitly and deeply exploited and studied these sectors separately, only a few works addressed these sub systems along the refinery supply-chain simultaneously. Guyonnet et al. (2009) integrated the crude oil unloading, production planning, and distribution sub models on a tactical decision level by solving each part in a sequential push or pull manner, where the planning horizon is discretized into time periods, typically a day or a week. Siwi et al. (2018) proposed a strategical planning of the petroleum and petrochemical supply chain. However, systematic studies for the integrated scheduling and optimization of the holistic refinery supply chain are still lacking.

To minimize the total operating cost along the holistic petroleum refinery supply-chain, this work developed a new integrated scheduling and optimization framework covering the crude oil unloading, transferring, manufacturing, and pipeline distribution subsystems. Note that the specifications requirement for crude oil blending and oil product blending (where nonlinearity may occur due to the blending operations) must be satisfied. Meanwhile, operating rules such as tank can either receive or discharge oil product at a time has been considered, which further closed the gap between the modeling and the realistic problem (Yu et al., 2020). Also, the inventory/capacity limits, delivery feasibility constraints and oil product demands must be satisfied, which may result in a multi-stage and large-scale MINLP model. The efficacy of the developed integrated scheduling and optimization framework is demonstrated by an industrial-scale case study.

Keywords: Integrated scheduling and optimization, MINLP, Holistic refinery supply chain, Crude unloading and transferring, Refinery manufacturing, Pipeline distribution