Stowage Planning with Optimal Ballast Water

Beizhen Jia, Kjetil Fagerholt, Line Blander Reinhardt, Niels Gorm Maly Rytter

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review


Stowage planning is at the essence of a maritime supply chain, especially for short sea Ro-Ro ships. This paper studies stowage optimisation of Ro-Ro ships with a focus on stability constraints and the applicability of models. The paper contributes to short sea Ro-Ro ship stowage in two ways. First, we propose an integrated approach of designing stowage models with the consideration of loading computers. Second, we present a mathematical formulation of the Ro-Ro Ship Stowage Problem with Ballast Water with a discretisation method, to generate an optimal stowage plan which meets stability requirements by means of the weight of cargoes instead of excess ballast water, i.e. excess fuel consumption. Computational tests based on empirical data indicate significant savings and potential of model application in the real world. Preliminary results show 57.69% ballast water reduction, equivalent to 6.7% fuel savings and CO2 reduction. Additional tests on instances with various cargo weight distribution and discretisation levels are conducted, and finally, improvements are suggested for further research considerations.
Original languageEnglish
Title of host publicationComputational Logistics. ICCL 2020
EditorsEduardo Lalla-Ruiz, Martijn Mes, Stefan Voß
Publication date2020
ISBN (Print)978-3-030-59746-7
ISBN (Electronic)978-3-030-59747-4
Publication statusPublished - 2020
Event11th International Conference on Computational Logistics - Enschede, Netherlands
Duration: 28. Sep 202030. Sep 2020


Conference11th International Conference on Computational Logistics
SeriesLecture Notes in Computer Science


  • Ballast water
  • Environmental impact
  • Maritime transportation
  • Stowage optimisation


Dive into the research topics of 'Stowage Planning with Optimal Ballast Water'. Together they form a unique fingerprint.

Cite this