Renewable electricity supply in Amazonas: Energy modeling to decarbonize isolated systems

Abstract

This study evaluates hybrid systems using local renewable energy resources to decarbonize isolated systems (SISOL) in the Brazilian Amazon. Using the Calliope energy optimization framework, a model for SISOL was built and applied to two case studies: Careiro da Várzea and Rio Preto da Eva. The systems were analyzed under scenarios combining solar photovoltaic (PV), batteries (Bt), and green hydrogen (H${}_2$), considering two land-use settings for PV deployment: no spatial restrictions (default) and the minimum required area to meet local electricity demand. The installed PV and storage capacities correspond to the cost-optimal solutions obtained through optimization. In Rio Preto da Eva, the PV–Bt scenario under the default condition reduces the levelized cost of electricity (LCOE) by 13% compared to the optimized fossil fuel (FF) baseline, and is characterized by 65 MW of PV capacity and 330 MWh of Bt storage. The PV–greenH${}_2$–Bt scenario increases the LCOE by 67%, with 117 MW of PV, 715 MWh of H${}_2$ storage, and 2.2 MWh of batteries. Under the strictest land-use constraint, which differs between scenarios due to their distinct conversion and storage needs, the LCOE of the PV–Bt scenario exceeds that of the PV–greenH${}_2$–Bt scenario. In this setting, the PV–Bt system deploys 53 MW of PV and 1.4 GWh of Bt storage, while the PV–greenH${}_2$–Bt configuration relies on 94 MW of PV, 3.2 GWh of H${}_2$ storage, and 36.6 MWh of batteries. Compared to FF, GHG emissions decreased by 92% in the PV–Bt scenario and by 85% in the PV–greenH${}_2$–Bt pathway. In contrast, Careiro da Várzea faces significant land constraints, making scenarios spatially infeasible under the evaluated conditions. These findings highlight the critical role of local context in designing possible strategies to deeply decarbonize isolated systems.