Paper Conference

Proceedings of uSim Conference 2018: 1st uSim Conference of IBPSA-Scotland

     

OPEN SCIENCE BUILDING STOCK MODELLING: AN EXAMPLE USING GBXML, OPENBUILDING AND ENERGYPLUS

Steven K Firth, Bianca Howard, Jonathan A Wright

Abstract: Open Science is the process of scientific study using open, reproducible and reusable datasets, models and methods. This paper explores the potential for Open Science in the field of Building Stock Modelling and investigates the benefits of this approach for the Building Simulation research community. An Open Science workflow is described for running EnergyPlus simulations on multiple buildings that could be used as the calculation engine of a future building stock model. The stages of the workflow are: 1) the original building descriptions are held as gbXML files; 2) simulation runs using EnergyPlus are carried out on the buildings; 3) the simulation results, combined with the gbXML data, are stored in a graph data structure for analysis; and 4) the results are analysed to create research findings. This is achieved using gbXML files, Python scripts, Jupyter notebooks and the ‘openbuilding’ Python package. The entire process is made available on GitHub. In this paper, the workflow is demonstrated using an example stock of 72 typical UK detached houses. The paper concludes with a discussion of the benefits, challenges and limitations of Open Science approaches for building stock modelling. The choice of input datasets and simulation methods is discussed, as well as the options for making both the results and methods of stock modelling studies available for the wider research community. Techniques for improving the reproducibility and the reusability of datasets and methods are also presented.
Pages: 97 - 104
Paper:
usim2018_029