Abstract
This thesis develops and assesses the impact of a new comprehensive, integrated multi-scale optimisation framework for Recirculating Aquaculture Systems (RAS) that incorporates renewable energy sources to enhance their techno-economic viability and environmental sustainability, focusing on the minimisation of the Levelised Life Cycle Cost (LLCC) over the system lifespan. The framework is specifically designed to support all key decision-making phases of the RAS system, from the design of the energy system to the model-based real-time and long-term operational decisions. While RAS offers significant potential for sustainable aquaculture, its widespread adoption is often hindered by high energy costs, operational complexities, and inherent market and environmental uncertainties. Existing optimisation models frequently address these challenges in isolation, lacking a holistic approach that co-optimises critical, interdependent subsystems across different timescales. To address this gap, this research presents a set of interconnected mathematical models. The core of the framework is a Mixed Integer Linear Programming (MILP) model for the long-term optimal design of the RAS facility, which simultaneously determines the type and capacity of integrated renewable energy systems by synergising aquacultural energy demands, fish bio economic models, and geographical climate conditions. Building on this, a two-stage stochastic programming model is introduced to ensure the resilience of these design decisions under uncertainty, explicitly accounting for volatility in prices and resources. The framework is further extended to address medium-term production planning by optimising multi-species cultivation strategies, and a hierarchical, short-term operational control strategy based on Model Predictive Control (MPC) is conceptualised to translate strategic plans into efficient real-time operations. The primary contribution is a comprehensive decision-support framework that unifies biological, economic, and energy objectives, with case studies demonstrating its value across all timescales. At the design level, the stochastic models yield more profitable (up to 24% revenue increase) and resilient designs, with notable cost reductions from leveraging the water’s thermal mass as energy storage. At the planning level, the framework enhances profitability by optimising multi-species production schedules based on market and thermal dynamics. At the operational level, the hierarchical MPC strategy is shown to reduce daily costs and improve energy self-sufficiency. This research provides systematic, data-driven guidelines for designing next-generation aquaculture systems. Future work will focus on enhancing the system's circularity by incorporating waste-to-energy pathways, such as modelling the anaerobic digestion of sludge to produce biogas for on-site heat and power generation.