Designing integrated and resilient multi-energy systems via multi-objective optimization and scenario analysis

Decarbonizing local energy systems is imperative to achieve climate neutrality by 2050 and limit global temperature rise to 1.5 °C, aligning with the European Green Deal. Effective planning relies on accurate energy modeling tools, with a shift towards multi-energy systems which integrate diverse energy vectors and technologies for enhanced flexibility and renewable energy hosting capacity. Traditional single-metric cost-focused approaches in multi-energy systems planning are evolving towards multi-objective optimization strategies, balancing costs with sustainability objectives like global warming potential, environmental impacts, and reliability. The complexity of renewable energy sources, market volatility, and climate change necessitates robust solutions resilient to multiple uncertainty sources. This study introduces a decision analytic framework for optimal planning of sustainable, resilient multi-energy systems, integrating the single-objective CALLIOPE simulation-optimization model with multi-objective evolutionary algorithms. Multiple algorithms are tested, and the best-performing algorithm is used to extract optimal configurations under alternating scenarios of renewable energy generation potential and energy prices. The new framework is tested on a synthetic case study based on the Sulcis Iglesiente (SI) Province in Sardinia, Italy, and considers opportunities for investing in renewable resources, expanding the local renewable power capacity, installing energy storage batteries, and transitioning from gas and oil boilers to heat pumps and biomass generators. Identification of optimal solutions is achieved via multi-objective optimizations, considering contrasting objectives such as air quality, energy independence, economic considerations, and emission targets. Results demonstrate that the new methodology allows for the extraction of multiple optimal configurations of the multi-energy system, incorporating different technology combinations based on the relative importance of objectives. Among the tested algorithms, EpsMOEA and OMOPSO perform the best, thoroughly exploring the decision space and returning unique optimal configurations. Scenario analysis reveals that the attractiveness of certain technologies, especially for heat generation, is highly sensitive to different objectives and scenarios. In contrast, others, such as onshore wind plants, remain favorable regardless of circumstances. • MOEAs allow multiobjective optimization using well-established single-objective models. • We optimized the planning and management of a MES under four competing objectives. • We compared seven algorithms to benchmark their performance. • Energy mix varies greatly among optimal solutions according to objective importance. • Uncertainties affect optimal MES planning significantly for specific cost classes.