| Title |
Examining the Energy Performance Gap After Window Replacement in an Aging University Building |
| DOI |
https://doi.org/10.5659/JAIK.2026.42.5.263 |
| Keywords |
Window Replacement; Performance Gap; Baseline Model; Masking Effect; Rebound Effect; University Building |
| Abstract |
This study examines the energy performance gap after window replacement in an aging university building using 48 months of measured
electricity-use data. A weather-normalized baseline model showed minimal mean bias, with NMBE of 0.00 percent, but relatively high
monthly variability, with CVRMSE of 21.05 percent. Therefore, the results should be interpreted with caution. Following the retrofit, annual
electricity use was 9.8 percent higher than the weather-normalized baseline and 20.2 percent higher than the pre-retrofit average. In particular,
the largest increases occurred in May and June. As a result, post-retrofit electricity use deviated from the baseline prediction, especially
during the early cooling season. However, because the dataset is aggregated monthly and the model contains some uncertainty, this gap is
better understood as a post-retrofit deviation pattern rather than direct evidence of the causal impact of window replacement. In addition,
several factors may have contributed to the observed changes, including shifts in operational conditions such as post-pandemic normalization
of campus activity, cooling-related behavioral responses, and variation in non-weather-sensitive base loads. Overall, this case suggests that the
whole-building electricity impact of a single envelope retrofit may be limited or partly masked in university buildings with highly variable
operational loads. Therefore, future evaluation would benefit from sub-metering and more detailed operational data. |