LBNL: Evaluating The Performance Of HVAC Optimal Control Based On Real-Time Floor-By-Floor Occupancy Data
By Guanjing Lin, Armando Casillas, Maggie Sheng, Jessica Granderson @ Lawrence Berkeley National Laboratory
Meeting aggressive decarbonization goals requires radical advancements that reduce existing buildings’ carbon footprint. New smart building technologies that offer continuous dynamic optimization of commercial Heating, Ventilation, and Air Conditioning (HVAC) control hold promise to advance building operations for decarbonization, efficiency, and flexible
control.
Typical HVAC control sequences are designed to condition spaces over a fixed schedule to meet space temperature setpoints. By incorporating occupancy information into HVAC control, space conditioning can be delivered more efficiently and adjust to changes in occupancy.
This paper illustrates the results from a field evaluation of a cloud-based building operation platform installed in an office building. In the study, 22 occupancy counters were installed (two on each floor) as a part of the platform to measure floor-by-floor building occupancy in real-time. The platform used occupancy data, thermal modeling, and machine learning algorithms to implement optimal start-up, shutdown, and intra-day fan speed adjustment to the air handling units (AHUs) for each floor during the pandemic.
This paper presents how the technology is implemented, the energy savings performance, and how occupancy information can be used to support executing dynamic operations. The technology reduced weekday AHU run times by over three hours and reduced fan speed by more than 10% during lunchtime. The simulated savings results are presented and compared with another similar study. The assessment results provide valuable insights to help end-users and industry partners understand the real-world performance of occupancy-based control technologies and reduce technical risks.
Access The Case Study
More Case Studies
