A validation methodology aid for improving a thermal building model: Case of diffuse radiation accounting in a tropical climate

TL;DR

This paper presents a validation methodology for thermal building models, emphasizing the importance of sensitivity analysis and diffuse radiation accounting in tropical climates, demonstrated through improvements to the CODYRUN software.

Contribution

It introduces a validation approach that enhances model accuracy by addressing diffuse radiation calculation errors in tropical climate conditions.

Findings

Improved prediction accuracy of solar radiation in the model.

Identified and corrected diffuse radiation calculation errors.

Highlighted the importance of sensitivity analysis in building simulation validation.

Abstract

As part of our efforts to complete the software CODYRUN validation, we chose as test building a block of flats constructed in Reunion Island, which has a humid tropical climate. The sensitivity analysis allowed us to study the effects of both diffuse and direct solar radiation on our model of this building. With regard to the choice and location of sensors, this stage of the study also led us to measure the solar radiation falling on the windows. The comparison of measured and predicted radiation clearly showed that our predictions over-estimated the incoming solar radiation, and we were able to trace the problem to the algorithm which calculates diffuse solar radiation. By calculating view factors between the windows and the associated shading devices, changes to the original program allowed us to improve the predictions, and so this article shows the importance of sensitivity analysis in this area of research.