Stochastic Optimization of Energy Prosumption Systems : A case study on energy balancing for self-sustainable societies

Abstract: To achieve a sustainable future fossil electricity is being replaced with renewable, leading to higher uncertainties in electricity production. This has resulted in an incentive for consumers to produce, sell, and store their own electricity, hence becoming prosumers. Austerland Skags is a Swedish project that explores the possibility to convert a small society into a prosumption system. The system includes solar and wind power as electricity producers and hydrogen-fueled vehicles for commodity transport. To capitalize the most on their produced electricity they want to store excess electricity. This master thesis uses Austerland Skags as a case study to develop a stochastic linear optimization model to determine the optimal energy storage solution for an energy prosumption system with both electricity and hydrogen demand. The method used in this thesis was the sample average approximation (SAA) algorithm. The results from the SAA were compared to the expected results from the expected value problem (EEV) to show the difference between a stochastic and deterministic solution. The results from the SAA turned out to consistently outperform the EEV for the samples created. Since hydrogen demand could only be sourced in-house, the model was forced to use an electrolyzer and hydrogen tank. The final result from the SAA showed that both a battery and fuel cell was used in addition to the electrolyzer and hydrogen tank in the optimal solution. All capacities stayed within reasonable levels showing the possibility of realizing a cost-effective prosumption system.