The Role of Power-to-X Technologies in Decentralised Multi-Energy Systems

In order to support energy self-reliance within countries, decrease greenhouse gas emissions, and reduce dependence on a declining fossil fuel supply, renewable energy sources are planned to replace a large percentage of fossil fuel electricity generation by 2050. With the replacement of centralised plants with decentralised solar photovoltaics or wind technologies, the future’s energy system may rely on partial shifts from centralised energy generation to distributed energy generation. In order to support an increasing penetration of renewable generation in Switzerland, both short and long-term storage technologies will be required to mitigate the temporal mismatch of fluctuating renewable production and end-user demand. Power-to-X is one group of storage pathways that are capable of both being installed in decentralised settings and storing energy long-term. Some of the relevant pathways for decentralised cases include Power-to-Hydrogen, Power-to-Methane, Power-to-CHP (combined heat and power), Power-to-Heat, and Power-to-Mobility. In this dissertation, the economic feasibility and emission reduction potential of several different Power-to-X pathways are investigated from 2015 until 2050. In order to test these pathways, this thesis has main three objectives: (1) to develop a multi-objective optimisation model that is capable of investigating Power-to-Gas pathways in multi-energy systems including long-term storage, (2) to incorporate personal transport energy demands and Power-to-Mobility pathways into the optimisation framework, and (3) to assess the uncertainties and sensitivities of these pathways from 2015 until 2050. The model in this work uses multi-objective optimisation that minimises both total annual cost and total emissions of the system. According to these objectives, the optimisation model selects the capacity of conversion technologies (photovoltaics, fuel cells, electrolysers, methanation, heat pumps, and gas boilers) and storage technologies (batteries, compressed natural gas storage, hydrogen storage, and thermal storage). Reduced order approximations for the part-load efficiencies, ramping limitations, temporal reduction techniques for long-term storage, and export constraints are used to reflect the performance and operational limitations. This model is tested using several scenarios in two case studies: one rural and one urban. It is found that the rural case study has high renewable potential and uses both batteries and Power-to-Heat (e.g., heat pumps and thermal storage) for short-term storage needs. Long-term Power-to-CHP storage is also used in cases that required deep decarbonisation. For the urban case study, there is not enough renewable potential to meet the energy targets or to significantly utilise storage systems. After testing these two case studies, the model is then expanded to simulate personal transport demands of the building occupants. The model expansion includes the selection of different vehicle technologies. These include internal combustion engine vehicles with gasoline, internal combustion engine vehicles with natural gas, battery electric vehicles (BEV), plug-in hybrid vehicles, and fuel cell electric vehicles (FCEV). This model is then applied to a suburban case study. The results show that BEVs are the preferred vehicle technology and that FCEVs are not predicted to be optimal due to their high costs and lower efficiencies. In order to assess the future performance of the system, an uncertainty analysis is performed using a Monte Carlo simulation and Latin Hypercube Sampling. Finally, a sensitivity analysis is performed on the optimisation model using Monte Carlo Filtering. This method specifically looks at the most important parameters that influence the selection of Power-to-CHP, Power-to-Methane, Power-to-Mobility with FCEVs, and at solutions that are able to achieve lower costs while still meeting emissions targets. In the uncertainty analysis, the most popular pathways are Power-to-Mobility with BEVs and Power-to-Heat. These pathways are used in 100% of the Monte Carlo simulations that met the emissions targets. Batteries and Power-to-CHP are the next most popular storage methods and are used in 84% and 19.2% of Monte Carlo samples respectively. The least popular pathways are Power-to-Methane and Power-to-Mobility with FCEVs, both representing less than 5% of samples. From these results, we can conclude that heat pumps, solar PV, and BEVs are three of the most important technologies for reducing emissions in both our buildings and transport sectors. Long-term storage may also be a valuable asset in certain case studies with sufficient renewable deployment, particularly if deep decarbonisation targets are required