Unlocking demand response in commercial buildings: Empirical response of commercial buildings to daily cooling set point adjustments

[Display omitted] •Experiments with zone-level temperature set points in commercial buildings.•385 daily samples are used to estimate the cyber-physical response of six buildings.•The uncertainty in building flexibility is quantified by statistical models.•Our stress test approach is a scalable method to estimate real life flexibility. Buildings represent over half of global electricity demand. Cooling buildings already accounts for over 9% of global electricity demand and is expected to grow rapidly due to climate-change induced hot-spells and increasing prosperity in developing economies. In the US, commercial buildings represent 35% of nationwide electricity consumption. Increased electricity demand for cooling services will challenge already stressed power grids, particularly during times of peak demand. This work explores the flexibility and demand response potential of large Heating, Ventilation and Air Conditioning (HVAC) systems based on an extensive set of measurements from six commercial buildings in a Warm-summer Mediterranean climate. Over a three-month summer period, zone-level temperature set points were adjusted daily in six commercial buildings to determine the effect on chilled water and electricity loads, as well as on zone-level temperatures. External weather conditions were measured continuously during the testing period. The experimental data that were collected are published with this article. These experiments confirmed the potential to provide flexibility by reducing energy demands based on modest zone-level temperature set point adjustments. A two-degree Fahrenheit increase of the cooling set point resulted in a 13–28% reduction in daily building-level cooling loads on average for four office buildings and 3–4% for two laboratory buildings. The impact on electric loads was less than 2% (excluding for cooling water but including for ventilation). Zone-level temperature increases were measurable but temperatures remained within the target ranges. By collecting 385 experiment-days of experiment data, we were able to parameterize statistical models for the response of the buildings. These models provide statistical guarantees on the reliability of thermal demand response. This work provides a blueprint for constructing building and zone-level energy-response functions and highlights the value of testing buildings repeatedly and across a range of weather conditions. Providing statistical performance guarantees will be critical for wides