| Author: | Luo, Xi |
| Title: | Data-driven optimization for sustainable shipping : addressing weather uncertainty, environmental dynamism, and hybrid propulsion complexity |
| Advisors: | Wang, Shuaian (LMS) |
| Degree: | Ph.D. |
| Year: | 2026 |
| Department: | Department of Logistics and Maritime Studies |
| Pages: | xv, 165 pages : color illustrations |
| Language: | English |
| Abstract: | The global maritime industry confronts mounting pressures from stringent emission reduction regulations and escalating fuel costs that constitute a substantial portion of operational expenses. Compared to capital-intensive technological retrofits, operational optimization through speed planning offers a cost-effective and readily implementable pathway for achieving energy efficiency improvements and emission reductions. Ship speed optimization aims to minimize a ship’s fuel consumption by determining optimal speed profiles along planned routes, typically following a predict-then-optimize paradigm where fuel consumption is first predicted and then used to optimize speed decisions. Since fuel consumption prediction relies on weather forecasts as critical input variables, and sea and weather conditions exhibit considerable variability during voyages, accurate weather forecasts become essential for effective speed optimization. However, existing speed optimization methods encounter two challenges that constrain their practical effectiveness. First, current models predominantly rely on deterministic weather forecasts while neglecting inherent forecast uncertainties. These approaches lack scientific frameworks to quantify the operational value of probabilistic forecast information. Second, static pre-voyage planning fails to capitalize on continuously updated weather forecasts available throughout voyages, resulting in suboptimal performance under evolving environmental conditions. Furthermore, while operational optimization offers immediate efficiency gains, achieving deeper decarbonization targets requires technological innovation. Emerging hybrid electric propulsion systems, which integrate battery storage with conventional engines, introduce greater operational complexity encompassing multi-dimensional coordination challenges in routing, speed selection, and energy management decisions that transcend conventional optimization paradigms. This thesis first establishes an innovative dual-model evaluation framework that scientifically simulates the planning-execution gap in maritime operations. This framework demonstrates that ensemble weather forecast-based speed optimization achieves approximately 1% additional fuel savings compared to deterministic forecast-based approaches, validated through computational experiments on two nine-day voyages under both ballast and laden conditions. The superior rhumb line-based data fusion method developed for this framework ensures high-quality inputs for the ship’s fuel consumption rate predictions. Building upon this foundation, this thesis introduces a rolling horizon optimization approach that enables ship speed replanning based on continuously updated weather forecasts throughout voyages. Computational results on two laden voyages demonstrate that this framework achieves 10.86% fuel savings compared to fixed speed operations and 7.53% additional savings compared to static optimization approaches, unlocking operational efficiency gains beyond conventional pre-voyage planning. While ship speed optimization offers immediate energy efficiency improvements, achieving deeper decarbonization targets fundamentally requires technological innovation. This thesis further addresses the operational complexity of hybrid propulsion systems by formulating the integrated model for hybrid electric platform supply vessel (PSV) operations. This model simultaneously coordinates routing, speed, charging, and energy management decisions under fuel quota constraints that enforce predetermined emission reduction targets. Using realistic instances from North Sea operations, we benchmark hybrid PSV performance against two diesel-only references representing current industry practices: fixed design-speed operations and speed-optimized operations. Results demonstrate that operators currently using fixed speeds can achieve substantial emission reductions while simultaneously reducing total costs through combined hybrid propulsion and speed optimization. For operators already employing speed optimization, hybrid PSV operations can achieve moderate emission targets with stable abatement costs of approximately 0.77 NOK/kg CO₂. Our analysis further quantifies the critical role of operational factors in determining the cost-effectiveness of hybrid operations: offshore wind farms prove essential for achieving ambitious emission targets by eliminating costly detours to remote charging stations, while voyage time budget flexibility creates favorable trade-offs between operational efficiency and emission reduction costs. These findings offer actionable insights for maritime operators navigating the transition to hybrid propulsion systems under increasingly stringent environmental regulations. This thesis systematically addresses weather forecast uncertainty in ship speed optimization and enables adaptive speed replanning with updated weather forecasts throughout voyages. By further extending the optimization framework to integrated hybrid system management, these methods provide practical tools for green shipping, offering viable pathways for the maritime industry to achieve sustainable development under mounting regulatory and economic pressures. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14396