A Multi-objective Optimization Approach in Defining the Decarbonization Strategy of a Refinery

Nowadays, nearly one quarter of global carbon dioxide emissions are attributable to energy use in industry, making this an important target for emission reductions. The scope of this study is hence that to define a cost-optimized decarbonization strategy for an energy and carbon intensive industry using an Italian refinery as a case study. The methodology involves the coupling of EnergyPLAN with a Multi-Objective Evolutionary Algorithm (MOEA), considering the minimization of annual cost and CO2 emissions as two potentially conflicting objectives and the energy technologies’ capacities as decision variables. For the target year 2025, EnergyPLAN+MOEA has allowed to model a range of 0-100 % decarbonization solutions characterized by optimal penetration mix of 22 technologies in the electrical, thermal, hydrogen feedstock and transport demand. A set of nine scenarios, with different land use availabilities and implementable technologies, each consisting of 100 optimal systems out of 10000 simulated ones, has been evaluated. The results show, on the one hand the possibility of achieving medium-high decarbonization solutions at costs close to current ones, on the other, how the decarbonization pathways strongly depend on the available land for solar thermal, photovoltaic and wind, as well as the presence of a biomass supply chain in the region.