Paper Conference

Proceedings of Building Simulation 2021: 17th Conference of IBPSA


Calibration of a multi-residential building energy model - Part I: Cluster-based sensitivity analysis

Charlotte Marguerite 1, Van Long Lê 1, Charlotte Beauthier 1, Olivier Fontaine de Ghélin 1, Cécile Goffaux 1, Loïc , de Moffarts 2
1 Cenaero, Belgium
2 Thomas&Piron, Belgium

Abstract: In the field of building models, sensitivity analyses are used to identify parameters that most influence buildings energy consumption. Most sensitivity analyses are applied to tertiary or industrial buildings. Few publications deal with sensitivity analyses applied to multi-residential buildings. The specific challenges of multi-residential buildings, such as the many apartments involved, each one having its physical characteristics (orientation, area, etc.), its specific equipment and energy use habits, are tackled with the new cluster-based method described here and applied at various levels (multi-scale analyses), to the case of a white-box model of a multiresidential building, having from 1 to 3 bedrooms, equipped with a centralised gas boiler. Results show that while the boiler characteristics and the building envelop parameters have the most influence on the energy consumption at building scale, the parameters related to customers’ behaviours have the most influence on the energy consumption at apartment scale and building scale. Those important parameters are then calibrated based on measured data and surrogate models (the calibration work is presented in a separate paper: Part II - Calibration Using Surrogate-Based Optimisation).
Keywords: multi-residential building, clustering methods, sensitivity analysis, calibration
Pages: 3156 - 3163