Dynamic integrated method applied to assessing the in-situ thermal performance of walls and whole buildings. Robustness analysis supported by a benchmark set-up

• Published in: Applied thermal engineering 2019-04, Vol.152, p.287-307
• Main Authors: Chávez, K., Ruiz, D.P., Jiménez, M.J.
• Format: Article
• Language: English
• Subjects: Benchmarks; Building envelope; Buildings; Differential equations; Dynamic analysis; In-situ tests; Mathematical analysis; Mathematical models; Outdoor testing; Performance indicators; Robust control; Robustness; Thermal parameters; Three dimensional analysis; Walls; Weather
• ISSN: 1359-4311; 1873-5606
• DOI: 10.1016/j.applthermaleng.2019.02.065

•Assessment method based on averages generalised by considering key physical issues.•Benchmark set up and data under different conditions support the result reliability.•Validity of results and conclusions reinforced by deep construction knowledge.•The accuracy of the results strongly depends on the applied integration period.•High uncertainty using 1 day as integration period (IP), accuracy demands longer IP. This article considers the application of an integrated dynamic method for assessing the in-situ thermal performance of walls (obtaining the U value by one-dimensional analysis), and whole buildings (obtaining the UA value and gA value by three-dimensional analysis). This method is based on the integration of differential energy-balance equations, considering an integration period long enough to make the accumulation term much lower than the other terms, and using averages to estimate integrals. Several aspects of this approach have been systematically analysed. On the one hand, the minimum integration period that allows this simplification has been identified and, on the other, several candidate models considering different approximations for the energy-balance equations have been evaluated. The robustness of the method has been analysed by comparing the results from an eight-month period including different test and weather conditions. A simplified building has been considered as a case study. Its detailed and accurate knowledge reinforces and complements the validation criteria. It is concluded that relatively simple models can give accurate results. The greater complexity provides no significant improvements. However, significant differences have been detected for the different integration periods: from insufficient daily averages to optimum six-day averages for the walls and four-day averages for the whole building.