A mathematical model for planning oil products distribution via pipeline

TL;DR

This paper introduces a new discrete-time mixed integer linear programming model for short-term multiproduct pipeline scheduling, optimizing product flow, injection, and delivery to reduce operational costs.

Contribution

It presents a novel discrete-time MILP model that handles multiple refineries and distribution centers, including simultaneous injections and deliveries, improving operational efficiency.

Findings

Significant cost reductions demonstrated on benchmark problems

Model effectively manages simultaneous injections and deliveries

Optimizes product sequencing and pipeline utilization

Abstract

Compared to other transportation modes, road, railroad and vessel, multiproduct pipelines are the safest and most economical way of conveying petroleum products over long distances day and night. During the last two decades, the operational multiproduct pipeline scheduling has gained increasing attention, where most of the contributions are based on continuous time representation. In this paper, we present a new discrete-time mixed integer linear programming model for the shortterm scheduling of multiproduct pipelines featuring multiple refineries and distribution centers. The model optimally determines the sequence and planning of product injections at input nodes, the sequence of product deliveries to output nodes, and traces the size and the products location along the pipeline at any time. The proposed model also allows to execute simultaneous injections at the input nodes and to tackle simultaneous injections and deliveries at an intermediate node that can act as both the input and output node. Solutions to two benchmark example problems illustrate that the proposed model presents significant reductions in the pipeline operational cost.