Application of different dynamic analysis approaches to the estimation of the building component U value

System identification is very useful for finding the thermal properties of building components from outdoors tests. Many factors, such as data analysis skills, quality of recorded data, performance of the analysis tool employed, approximations and hypotheses on the component and its boundary, etc., can influence the results. The work reported in this paper intended to estimate achievable agreement when different analysis approaches are applied to the same and different datasets to find the U value of a given building component. The spread in the results is analysed. To isolate and highlight the effects of the different analysis approaches and test conditions from the effects of other assumptions about the test component and its boundaries, a well-known, quite simple uniform opaque wall was analysed taking direct component level measurements into consideration.