Multi-objective Optimization and Uncertainty Analysis Reveal Resilient Water-Energy System Configurations on Small Islands

Multi-energy systems (MESs) integrate a variety of technologies and energy carriers into a unified framework, offering flexible and efficient solutions to the challenges of modern energy systems. These include the urgent need for decarbonization, greater integration of renewable energy sources, and a push for decentralization and energy market independence. However, long-term planning for MESs is becoming increasingly complex in a rapidly changing world, where socio-economic, technological, and climatic shifts can quickly render obsolete solution previously considered optimal. These challenges are particularly acute for vulnerable systems, such as those in small, isolated island communities.

This study focuses on identifying robust future configurations for the energy-water systems of Italian minor islands, which face significant challenges in energy and water supply. The aim is to increase energy independence and promote decarbonization. The analysis explores the integration of desalination plants and rooftop photovoltaic systems as replacements for fossil fuel generators and water transport via tanker ships, together with the expansion of the islands’ water storage capacity. The study incorporates various sources of uncertainty—technological, climatic, and economic—into a multi-objective analysis to evaluate their individual and combined impacts on optimized configurations.

A novel optimization framework is presented, which combines Multi-Objective Evolutionary Algorithms (MOEAs) with the multi-energy system planning model CALLIOPE. This approach identifies trade-off solutions for energy-water system configurations under conflicting objectives. The process is iterated across several scenarios, each defined by a unique combination of uncertainties. Sensitivity indexes and probabilistic analyses are employed to assess variations in performance metrics and the robustness of optimized configurations to each uncertainty source.

The results reveal that the optimal configurations include substantial integration of photovoltaic systems and desalination plants, effectively reducing CO₂ emissions and energy costs. The analysis also highlights the significant role of uncertainty in influencing system performance, particularly the impact of technology-specific parameters like ship tanker emission factors. and desalination plants efficiency. In most scenarios, especially for the island further away from the coast, the replacement installation of desalination plants coupled with renewable technologies proves to be both cost-effective and environmentally sustainable, demonstrating the robustness of the proposed configurations.

This work provides a new decision-making tool for multi-energy system planning, which emphasise the critical role of uncertainty analysis in ensuring resilient planning under fluctuating resource availability and demand. It explores multiple options for the implementation of renewable energy solutions in isolated and vulnerable regions, contributing to a sustainable and resilient energy transition.