Optimized placement of thermo-electric energy systems in city districts under uncertainty
Aachen / E.ON Energy Research Center (2019) [Book, Dissertation / PhD Thesis]
Page(s): 1 Online-Ressource (xxv, 149 Seiten) : Illustrationen, Diagramme
Cities, as large energy consumers, offer great potential for greenhouse gas emission reduction. Urban energy system planning is important to identify low-emission, cost-efficient solutions to contribute to global emission reduction aims. However, the planning process is challenging due to uncertainty. Conventional approaches to deal with uncertainty mainly use robust optimization or stochastic programming, which are computationally intensive on urban scale and might not support the identification of energy systems with reduced sensitivity to uncertainty. The subject of this thesis is to identify urban energy systems, which are less sensitive to uncertainty. An optimization framework is developed, which combines a genetic algorithm for optimized energy system placement with an economic and ecological uncertainty calculator. Demand-related, technical, and economic uncertainties are taken into account. The effect of different risk attitudes, from risk neutral to high risk-aversity, is analyzed. Results show that a reduction of economic and ecological uncertainty is possible but at the drawback of worse mean cost and emission values. The uncertainty reduction within the given case study is achieved by increasing the nominal power of the combined heat and power (CHP) units as well as the capacity of the thermal storage systems and decreasing photovoltaic (PV) area sizes. Boiler-PV combinations define cost-efficient results. CHP usage with local heating networks generates further emission reduction. Reference optimizations without uncertainty and the optimization with uncertainty lead to comparable results. However, not all reference optimization results perform well under uncertainty. While many small-scale CHP configurations have a pareto-optimal behavior for the reference scenario, they lead to suboptimal performance under uncertainty. The presented framework enables the identification of energy systems, which perform well under uncertainty. Moreover, it enables a reduction of economic and ecological uncertainty for different risk-attitudes. However, the optimization process turns out to be computationally intensive, too. Thus, a reference optimization without uncertainty should be preferred for applications with low availability of computational resources. Still, the presented method provides deeper insights into the economic and ecological uncertainty of urban energy systems.