DR Control with Variable Capacity Commercial HVAC System

Demand Response (DR) programs play an important role in allowing SCE to provide reliable, affordable, and environmentally-responsible electricity to its customers. As California transitions to renewable energy sources as part of statewide goals to reduce Greenhouse Gas (GHG) emissions (as described in SCE’s Pathway 2045), the flexibility and resilience of customer resources will be increasingly valuable for managing the inherent variability of these non-dispatchable resources. Due to its contribution to system peak, Heating, Ventilation, and Air-Conditioning (HVAC) equipment is key in helping utility programs manage electricity demand. Variable-Capacity (VC) HVAC equipment provides enhanced Energy Efficiency (EE) and customer comfort benefits over conventional fixed-speed equipment. For light commercial applications, Variable Refrigerant Flow (VRF) systems and packaged Rooftop Units (RTUs) leverage variable-speed components and complex controls to achieve superior part-load efficiency over conventional equipment. Numerous studies have been completed in recent years to assess the benefits of variable HVAC systems, which have become a valuable resource in utility EE programs. Yet the variable-speed capabilities of these systems have not been fully leveraged for DR. With their on-board instrumentation and communications capabilities, VC systems are prime candidates for implementing both EE and DR functionality, potentially offering dual program participation. The implementation of DR control strategies that leverage the superior part-load efficiency of these systems could enable greater demand reduction or reduced impact on occupant comfort over DR with conventional fixed-speed HVAC systems. This project sought to address this market gap by demonstrating advanced DR control strategies for VC air-to-air HVAC systems in light commercial buildings in SCE territory. After outreach to several companies to partner in this project, a producer of variable-speed RTUs (Manufacturer A) and a major manufacturer of VRF systems (Manufacturer B) agreed to participate. Manufacturer B, the VRF system manufacturer, implemented a conventional DR strategy and supported OpenADR 2.0b, as well as three advanced DR strategies: Change in thermostat setpoint temperature (conventional strategy). Limit equipment thermal capacity (and therefore electric power) subject to a maximum allowable deviation in indoor temperature. Change in temperature setpoint to increase capacity delivery (“load up” to enable pre-cooling). Targeted capacity reduction, where cooling or heating is turned off for specific zones to meet the target reduction based on zone priority. SCE helped identify field sites for each technology in light commercial buildings that represent typical applications in their service territory, and one was selected for each technology (two total). Each site was instrumented with power monitoring and temperature sensors so HVAC system response and the impact on indoor temperatures could be analyzed in detail. For the VRF system, the findings of this study indicate that Capacity Limit control can reduce electrical demand with minimal impact on indoor temperatures. During tests with outdoor temperatures around 95°F, Capacity Limit control consistently provided 15-25% reduction in average power during 1- or 2-hour DR events compared to baseline. The strategy of Pre-cooling followed by Capacity Limit control also provided a 15% reduction. However, the other two major control strategies testing—Setpoint Offset and Targeted Capacity Reduction, where VRF indoor conditioned zones were disabled in a specific order—did not consistently reduce demand on peak days. Test results indicated that Setpoint Offset—the strategy most similar to conventional HVAC DR control using thermostats—could also significantly reduce VRF demand at mild outdoor temperatures. The VRF system was found to respond as expected to commands sent via OpenADR 2.0b, and an OpenADR Virtual Top Node (VTN) was used to schedule and initiate test events. As a result of participation in this study, Manufacturer B has integrated this advanced DR control strategy into a controller designed for commercial buildings. For the variable-speed packaged RTU, Manufacturer A, while understanding the benefits of DR, was unable to implement advanced DR controls or OpenADR support during this project. For this RTU, remote control of conventional setpoint offset events could only be initiated manually via a web portal. Several tests were completed using the web portal method, with mixed results. A significant time delay (15 to 60 minutes) was experienced between when the DR signal was sent, when the unit reacted, and when acknowledgement was received. On-site power and temperature measurements indicated the RTU did not take full advantage of its variable-speed capabilities for DR and instead shut off in response to setpoint offset commands in several test events, much like a conventional single-speed system. Project findings suggest several follow-on activities of interest. An industry standard should be developed to harmonize DR responses from variable-speed HVAC systems in this class. Unless a standard is available, manufacturers will continue to ask for a unified list of specific responses they can implement without creating custom solutions for each program. Recently, Air-Conditioning, Heating, and Refrigeration Institute (AHRI) Standard 1380 was published for VC HVAC equipment of less than 65,000 Btu/h (5.4 tons), which could be used as a starting point to define similar or enhanced DR responses for larger equipment (for example, 5.4 – 60 tons). Additional field testing of larger Manufacturer B VRF systems would be beneficial for several reasons. First, Manufacturer B recently integrated the DR controls tested in this study into a lower-cost controller, which has not been independently verified. Second, testing with a VRF system larger than approximately 14 tons would allow a greater level of demand reduction control, due to the additional capacity stages available. Furthermore, this project tested the equipment serving zones that had very low occupancy. Additional larger-scale field testing will provide a better understanding of how these controls perform under various loading conditions and building types, allowing a more detailed evaluation of the DR potential and possible impact on occupant comfort at scale, before full program rollout. Lastly, it would be advantageous to work with additional equipment manufacturers to expand the EE DR opportunity for this class of equipment. Several manufacturers have told EPRI they are beginning to assess and implement these types of controls in their light commercial variable-speed products. These manufacturers are eager to provide value to electric utilities with enhanced DR capabilities, but need assistance in implementing and testing the new controls.