A Robust Optimization Application for Distributed Energy System Planning Considering Multiple Uncertainties

The efficient design of distributed energy systems is affected by many uncertain factors. However, many current studies have not fully considered uncertainty, and the solutions obtained are not feasible in extreme situations. In this regard, the planning of a distributed energy system in a hospital in North China is modeled as a mixed integer linear programming, combining global sensitivity analysis and robust optimization to obtain robust optimization schemes with different degrees of conservatism. According to the findings, the deterministic case’s total yearly cost of distributed energy systems is between 4.42 and 6.84×106 EUR, while the corresponding total annual carbon emissions are between 54.4 and 37.6 thousand tons. The annual total carbon emissions of the system will be reduced by up to 30.9%, and the corresponding annual total cost will be increased by 54.7%. Considering the uncertainty of the system, through the global sensitivity analysis, it is found that the price of natural gas has the greatest impact on the system. Only when the price of solid oxide fuel cells falls to 712 EUR/kW, it will replace other traditional technologies. Finally, robust optimization of natural gas price changes is carried out to explore solutions with different degrees of conservatism.