Paper Conference

Proceedings of uSim Conference 2020: 2nd uSim Conference of IBPSA-Scotland


15 Computational efficiencies and predication accuracy of a physics-based building energy simulation tool for urban energy analysis

Richard Charles Johnson, Mohammad Royapoor, Sara Walker

Abstract: The decarbonisation of residential building stock in the UK requires accessible tools that can reliably and rapidly model residential building power demands as a function of multiple low carbon technologies and building control schemes. Whilst a variety of modelling tools exists, these platforms are either intended for expert analysts, are not suited to rapid simulation (and therefore cumbersome at stock modelling scale) or are not flexible enough to allow analysis of detailed active control schemes. This work builds on a previously developed dynamic domestic building modelling tool developed in MATLAB/Simulink environment and intended for rapid generation of electrified heat demand profiles in buildings. The number of parameter inputs and timeresource required to prepare EWASP tool is several order of magnitude smaller than an equivalent EnergyPlus model, computational efficiency of this tool as well as its prediction accuracy are benchmarked against an equivalent E+ model. The EWASP model required 13 times less parameter input, reducing analyst time requirement and human effort. Both models produced similar trends of loads against external climatic changes for a Passivhaus case-study fabric while overall EWASP generated smaller ASHP electrical loads (4.4 kWh·m²·yr) than EnergyPlus model (5.8 kWh·m²·yr) which will be examined in future works. EWASP tool can assist assessment of the impact of fabric or HVAC retrofit and design and control scenario in buildings on the local distribution network and wider power grid.
Pages: 36 - 43