Towards an in-situ evaluation methodology of thermal resistance of basement walls in buildings

Abstract

One of the essential aspects in building energy analysis is determining correctly the thermal characteristics of envelopes. Characterizing envelope elements accurately allows energy consumption to be predicted and energy saving measures to be established. Basement walls are among those common envelope elements in buildings. The estimation of their thermal resistance is a challenge due to the difficulties and the lack of methodologies to know the wall stratigraphy. This study establishes a methodology to characterize the thermal resistance of this kind of walls. Such methodology consists in monitoring easily measurable variables (internal air temperature, external air temperature, ground temperature, internal surface temperature, and heat flux) which, together with the use of regression models, estimate thermal resistance. The methodology is validated in 2 phases: (i) in a calibrated hot-box adapted to elements in contact with the ground, and (ii) by analysing actual case studies. With the data obtained, various regression models are created based on time windows. The M5 Prime algorithm with a time window of 2 days of observation made accurate estimations in the case studies. An efficient methodology is therefore developed for estimating the thermal resistance in actual case studies. Also, the calibrated hot-box adapted could be used to analyse samples under controlled conditions in laboratory.