NSB 2014 edition:10 location:Lund date:15-19 June 2014
A more precise knowledge of the as-built thermal performance of our buildings’ fabric is of prime importance for the ongoing tendency to more stringent building performance demands. The methods that are commonly used for on-site thermal characterisation, such as the average method and linear regression technique, are based on stationary boundary conditions. As the latter are never encountered on site in practice, the methods’ validity depends on outdoor weather conditions. This paper examines the practical applicability of different in-situ thermal characterisation methods based on simulated data for an insulated cavity wall. Common semi-stationary methods are compared with a more advanced dynamical data analysis method, giving special attention to the reliability of the methods’ estimation results when confronted with data sets of limited measurement time spans and different measurement periods throughout the year. From this research, it can be stated that the use of semi-stationary methods for the characterisation of an insulated south-faced cavity wall can lead to accurate results in realistic measurement time spans when applied during winter months. The methods become less reliable when the temperature difference across the wall decreases. The dynamic method showed to be less sensible to the measurement period, provides more accurate results and needs shorter measurement time spans. The analysis itself however, showed to be more time consuming.