A near-optimal solution method for year-round operational planning of energy supply-storage systems utilizing time-domain decomposition

Energy savings and cost reduction can be effectively achieved using photovoltaics. Seasonal storage of surplus photovoltaic electricity improves self-sufficiency and supports decarbonization in the energy supply sector. Feasibility studies on energy supply-storage systems, involving power-to-gas and hydrogen storage, must incorporate year-round operational planning in their design. To improve computational efficiency and searching ability for such problems, a near-optimal solution method utilizing time-domain decomposition is developed. First, the technique decomposes a mixed-integer linear programming operational planning problem into smaller subproblems with short planning horizons and a master problem addressing time-coupling constraints. A lower bound is calculated via column generation to decomposed problems. Moreover, a finalization approach is uniquely developed to assign binary variable values based on partial solutions, to search near-optimal solutions. This method is applied to a year-round operational planning problem for an energy supply-storage system involving a photovoltaic, a water electrolyzer, a metal hydride tank, and a pure hydrogen fuel cell combined heat and power. The results indicate that the developed method obtains feasible solutions 20.3 % faster than conventional methods and achieves a near-optimal solution with a small optimality gap. The analysis includes year-round storage planning for electricity, hot water, and hydrogen.