Mixed integer optimization of an LNG supply chain in the Baltic Sea region

Alice Bittante*, Raine Jokinen, Jan Krooks, Frank Pettersson, Henrik Saxén

*Corresponding author for this work

Research output: Contribution to journalEditorialScientificpeer-review

Abstract

A numerical tool for designing optimal small-scale supply chains of liquefied natural gas (LNG) is presented. The main problem is formulated as a supply task, where LNG is delivered from a set of supply terminals to a set of receiving (satellite) terminals by ship transportation and by land-based truck transports from the terminals to customers on or off the coast. The objective is to minimize the overall cost, considering the price of LNG, investment cost of the receiving terminals and LNG trucks, rental costs of the ships and delivering costs. The problem is written as a mixed integer linear programming (MILP) problem. The optimization results give information about the placement of the satellite terminals and their capacity, the optimal fleet (ship size and number), the number of trucks, travelling routes of the ships and trucks, and the amount of LNG to supply to the demand sites. The system developed is illustrated by a set of examples designed to shed light on the future LNG supply in the region around the Gulf of Bothnia. The supply chain is optimized under different price of the alternative fuel and the arising solutions are analysed. It is demonstrated that there is a general consensus on where to build the satellite terminals, even though the delivered quantities of LNG vary depending on the price difference to the alternative fuel. Given the short computational time required to solve the examples of the paper, the model can easily tackle more complicated supply chain problems in the future.

Keywords

  • Energy Systems
  • MILP
  • Optimization
  • Small Scale LNG
  • Supply Chain

Fingerprint Dive into the research topics of 'Mixed integer optimization of an LNG supply chain in the Baltic Sea region'. Together they form a unique fingerprint.

Cite this