A co-simulation based framework for the analysis of integrated urban energy systems
Lessons from a Swedish case study
Time: Fri 2021-04-16 08.00
Subject area: Energy Technology
Doctoral student: Monica Arnaudo , Kraft- och värmeteknologi, KTH, Integrated smart energy systems
Opponent: Professor Louise Ödlund, Linköping University
Supervisor: Professor Björn Laumert, Kraft- och värmeteknologi
As major responsible for CO2 emissions, the energy sector is urgently called to take action against climatechange. The integration of renewable energy resources is a solution that, however, comes with a challenge.In fact, renewables are often variable, unpredictable and distributed. These characteristics add an extremecomplexity to the design and control of energy systems. Sector-coupling is nowadays strongly supported asa promising approach to increase the flexibility of these systems. For example, wind power curtailment canbe reduced by using the power surplus to operate heat pumps. When the wind does not blow, the heat storedin the thermal mass of the buildings and waste heat recovery can be used instead. These solutions are largelyavailable at district-to-city level. However, a suitable framework to design these integrated urban energysystems is missing.This thesis work proposes such a framework, as a set of methodological steps and integrated modellingtools. Among them, the modelling and simulation approach is a fundamental aspect. Given theheterogeneity of integrated energy systems, dedicated technology-specific models are developed and usedto achieve the required level of detail. A co-simulation method is implemented when time step coordinationand data exchange are necessary. Scenarios are developed to compare the techno-economic andenvironmental performance of alternative solutions, based on sector-coupling. Levelized cost of energy andCO2 emissions are used as main performance indicators for this purpose. In order to show the applicabilityof this methodology, Hammarby Sjöstad (Stockholm, Sweden) is selected as a case study. This also allowsto tackle a real local open issue, which is the definition of the best solution between district heating anddomestic heat pumps for multi-apartment buildings.The proposed framework was successfully applied to the case study. Case specific results allowed toformulate more general conclusions applicable to similar multi-apartment residential districts, in a Swedishcontext. It could be shown that co-simulation is a useful approach to capture sector-coupling bottlenecksand opportunities. Respective examples are electricity grid overloadings caused by installations of heatpumps and the control of thermal mass in buildings to replace the use of heat peak boilers. However, cosimulationshould be strictly limited to cases where control feedback loops need to be taken into account,such as in the previous examples. This is because it involves a higher implementation complexity and ahigher computational time. Thus, for example, the models of a heat network and of an electricity grid withno coupling technologies, such as heat pumps and electric boilers, should be preferably analyzedsequentially. The levelized cost of heat was found to be a game-changer parameter when comparing energyinfrastructures, beyond the specific business aspects. For example, the replacement of a district heatingtariff with its levelized cost of heat clearly showed the economic advantage of heat networks againstdomestic heat pumps. The CO2 emissions factors of different energy resources (waste, biomass, electricitymix) were shown to be highly critical for two main reasons. Firstly, different assumptions for these factorsled to opposite findings regarding the carbon footprint of specific technologies. For example, heat pumpscould be estimated as both more and less polluting than district heating, depending on the assumedemission factors. Secondly, control strategies based on the CO2 emission factors of the electricity supplymix (power-to-heat) were found to be a promising sector-coupling solution. By analyzing integrated energysystems, it was possible to assess uncovered bottlenecks and suggest new options. In particular, it wasshown that the installation of a large number of distributed heat pumps can overload the electricitydistribution grid in a district. Demand side management, through the thermal mass in buildings andvehicle-to-grid, could help alleviating this problem. On the other hand, district heating was found to be aneven more promising alternative, by integrating demand side management and heat recovery. Heat pumpswere shown to be a suitable partner technology for supporting heat recovery and enabling power-to-heat.