Optimal sizing of an HRES with probabilistic modeling of uncertainties − a framework for techno-economic analysis

•Study of PV/Biomass/Battery system for a remote village in Bangladesh.•Probabilistic modeling to address the uncertainties of resources and load demand.•Optimization with three different algorithms for effective global minima finding.•Performance superiority of Dandelion Optimizer (DO).•Obtained results provide technical and economic margins for stakeholders. Hybrid Renewable Energy System (HRES) has become a popular alternative for locations restricted to national grid connection due to geographical limitations. The study investigates the available renewable resources of a remote village in Mymensingh district of Bangladesh to propose and evaluate optimal sizing and cost of a grid independent HRES. The intermittency of solar irradiance and uncertain variation in load demand are taken into account by adopting probabilistic scenario-based analysis (SBA). Beta distribution and Gaussian distribution are considered for solar irradiance and load uncertainties, respectively, and Roulette Wheel mechanism generates multiple probabilistic scenarios. Objective function is formulated to minimize the total system cost (TSC) under defined constraints. Dandelion optimizer (DO), a relatively recent and unexplored metaheuristic algorithm along with two other popular optimization algorithms, Slime Mould Algorithm (SMA) and Real Coded Genetic Algorithm (RCGA) are applied to size the components for different probabilistic scenarios generated by Roulette Wheel. DO outperformed SMA and RCGA providing the solution set with 18.4 % and 3.2 % reduction in simulation time and system cost, respectively, in comparison with RCGA, whereas, 11.3 % and 16.7 % reduction in simulation time and system cost, respectively, in contrast to SMA. DO’s computation of cost margin (1.17 million dollars to 1.46 million dollars), PV allocation (minimum 7 modules to maximum 218 modules), biomass power allocation (maximum of 70.34215 kW) and optimized battery allocation computed by the algorithms (24 units) provide authorities a margin of cost-efficient estimation with conservative preparedness for developing HRES in the selected region. The study also provides a scientific framework to account for the uncertain parameters while proposing and analyzing HRES for off-grid localities, where inadequate literary exploration is evident in the context of Bangladesh.