Comparative assessment of multi-objective optimization of hybrid energy storage system considering grid balancing

Abstract

With the accelerated transition towards affordable and clean energy sources, the energy sector is undergoing a structural transformation that has resulted in a further increase in the complexity of energy system planning with rapid changes in techno-economic, environmental, reliability and social constraints. This signifies the consideration of purpose-driven multi-objective functions depending upon the functionality and applicability of the model. However, most of the studies adopt conventional bi-objective optimization either involving techno-economic, reliability and grid balancing parameters but there is a lack of comparative assessment of multi-objective optimization sizing for grid-interactive hybrid renewable energy system (HRES) consisting of short and long-term, battery and pumped hydro storage (PHS), energy storage systems (ESS). This study presents a comparative multi-objective framework to assess bi- and tri-objective function sizing techniques under grid balancing and non-balancing modes, to understand the scope and adaptivity of the modeling process for large-scale grid-interactive HRES. The analysis of results shows that the non-balancing mode underestimates the cost of energy (COE) by 18–30% compared to the grid balancing mode due to smaller decision variable space while long-term ESS dominance is vital for the reduction of grid burden compared to short-term ESS. In terms of configuration, a hybrid ESS system, 0.22 MWh battery, 18.1 MWh PHS, and 5.4 MW PV capacity, is the best optimal configuration in grid balancing mode with the COE, EEI and EII equal to 0.09 $/kWh, 7.5% and 10.5% respectively, whereas higher grid energy mismatch is induced by non-balancing mode with the overestimation of EEI and EII indexes up to 30% and 33% respectively. The environmental analysis shows that the carbon emissions avoided (CEA) are underestimated by 59.1% with the non-consideration of grid balancing. This signifies that the adaptive optimization model improves the design and planning process of grid-interactive HRES by capturing larger uncertainties related to COE, grid balancing, and CEA with changes in the system and ESS sizing. Overall, this analysis provides a purpose-driven perspective to energy modelers and policymakers for the energy system modeling process of grid-interactive HRES.