Paper Conference

Proceedings of Building Simulation 2017: 15th Conference of IBPSA


A Functional Principal Components Model for Internal Loads in Building Energy Simulation

Rebecca Ward1, Ruchi Choudhary1, Yeonsook Heo2, John Aston3
1Energy Efficient Cities initiative, Department of Engineering, University of Cambridge, UK
2Department of Architecture, University of Cambridge, UK
3Statistical Laboratory, University of Cambridge, UK

Abstract: There is currently no established methodology for quantifying uncertainty in occupant-related building internal loads. In this paper, we propose that distinct spaces within a building may be assigned an operational signature comprising the daily base load, load range and diversity profile. A Functional Data Analysis (FDA) approach has been used to analyse monitored electricity consumption data for the derivation of such signatures. This approach enables simulation of the inherent stochasticity. It represents a step forward towards an ability to propagate uncertainty through a building energy simulation and to quantify the change in electricity consumption associated with a change in building operation.
Pages: 136 - 144