Low carbon multi-vector energy systems: a case study of the University of Edinburgh's 2040 'Net Zero' target
The ultimate goal of this research was to develop a methodology to support decision-making by large (public sector) organisations regarding future energy technology choices to reduce carbon emissions. This culminated in the development of a multi-vector campus energy systems modelling tool that was applied to the University of Edinburgh as a case study. To deliver this a series of objectives were addressed. Machine learning models were applied to model building heat and electrical energy use for extrapolation to campus level. This was applied to explore the scope to reduce campus level emissions through operational changes; this demonstrated that it is difficult to further reduce the carbon emissions without technological changes given the University’s heavy reliance on natural gas-fired combined heat and power and boilers. As part of the analysis of alternative energy sources, the scope for off-campus wind farms was considered; specifically this focussed on estimation of wind farm generation at the planning stage and employed a model transfer strategy to facilitate use of metered data from wind farms. One of the key issues in making decisions about future energy sources on campus is the simultaneous changes in the wider energy system and specifically the decarbonisation of electricity; to facilitate better choices about onsite production and imports from the grid, a fundamental electricity model was developed to translate the National Grid Future Energy Scenarios into plausible patterns of electricity prices. The learning from these activities were incorporated into a model able to develop possible configurations for campus-level multi-vector energy systems given a variety of future pathways and uncertainties. The optimal planning model is formulated as a mixed-integer linear programming model with the objective to minimize the overall cost including carbon emissions. A numerical case study for the planning of three real-world campuses is presented to demonstrate the effectiveness of the proposed method. The conclusion highlights the importance of energy storage and a remote wind farm in these energy systems. Also, it is noted that there is no single solution that works in all cases where there are differences in factors such as device cost and performance, the gap between gas and electricity prices, weather conditions and the use (or otherwise) of cross-campus local energy balancing.