Development of voyage optimization algorithms for sustainable shipping and their impact to ship design

The environmental impacts from shipping and the societal challenges of human and property losses caused by ship accidents are pressuring the shipping industry to improve its energy efficiency and enhance ship safety. Voyage optimization is such an effective measure that has been widely adopted in today’s shipping market. The voyage optimization algorithm is the dominant part of any voyage optimization methods. The main objective of this thesis is to develop sophisticated voyage optimization algorithms, explore their applications to sustainable ship operations, and study its impact on ship fatigue design. In this thesis, five commonly used voyage optimization algorithms are first implemented and compared to provide a foundation for understanding optimization algorithms. A three-dimensional Dijkstra’s algorithm is then developed with further improvement based on the comparison. It can provide globally optimal solutions and conducting multi-objective voyage optimization. An engine-power based multi-objective optimization algorithm is proposed for the aid of ship operations with power-setting in their navigation system. Furthermore, the influence of the uncertainties from voyage optimization inputs, e.g., metocean forecast, implemented ship performance models and voyage optimization algorithms, on the optimization results is investigated. Moreover, the capabilities of the proposed voyage optimization algorithms to handle other optimization objectives, i.e., less fatigue damage accumulation and lower fatigue crack propagation rate, is also investigated. Meanwhile, two statistical wave models are compared to study the variation of a ship’s encountered wave environment for ship fatigue design. The impact of voyage optimization aided operations on a ship’s encountered wave environments and fatigue life assessment is also researched in this thesis. The three-dimensional Dijkstra’s algorithm addresses the limitations of conventional voyage optimization algorithms and allows for voluntary speed variation. It has a great potential of saving fuel up to about 12% in comparison with the case study ship’s actual sailing routes. The ship engine setting-based optimization algorithm provides a scheme based on a genetic algorithm and dynamic programming concept. It has the potential to save fuel up to approximately 14.5% compared to the actual sailing routes. This study also shows that metocean uncertainties in the voyage optimization process have great influence on the optimization results, i.e., 3-10% difference in fuel consumption for the same voyage optimization method. In addition, statistical wave models have been proven to capture ship-encountered wave statistics. It is also shown that the actual wave environments encountered by ships differ significantly from the wave scatter diagram provided by class guidelines. A good voyage optimization method can help to extend a ship’s fatigue life by at least 50%. Keywords: Dijkstra’s algorithm; Energy efficiency; Expected time of arrival (ETA); Genetic algorithm; Metocean forecast; Ship safety; Sustainable shipping; Voyage optimization algorithms.