Informative yet compact monitoring for building retrofits: Joint information entropy–based measurement system design applied to a real office building

Building energy retrofits are essential for enhancing the sustainability of the built environment. However, their effectiveness is impacted by the performance gap, the difference between expected and actual building energy performance, particularly when pre-retrofit building knowledge is inaccurate. Advances in monitoring technologies have improved the ability to measure energy performance and indoor environmental quality, helping to reduce this gap. Nevertheless, the systematic design of measurement systems remains insufficiently explored. Measurement-system design in building energy contexts, carried out prior to sensor installation, requires the simultaneous consideration of measurement type, location, and time interval. The integrated treatment of these design parameters has received little attention, especially in full-scale applications. This study addresses these gaps by implementing an approach that identifies the most informative measurements through the maximization of joint information entropy, demonstrated in a full-scale case study of an office building in Germany. Using building management system data from 2017, measurements were ranked with a hierarchical greedy algorithm that maximized joint entropy. The resulting ranking was validated through system identification. The findings indicate a trend between increased joint entropy and improved model parameter identification performance, suggesting that joint entropy is a meaningful proxy for information gain in building energy retrofit scenarios. Comparable system-identification performance was achieved using only 45% of the available measurement devices, demonstrating the potential for compact yet informative measurement systems. While the findings are specific to this study and future research is recommended to generalize the observations, this study provides empirical evidence that maximum-entropy-based measurement system design supports rational and efficient monitoring strategies for assessing building energy performance and guiding retrofit decision-making.