Integrating Smart Ceiling Fans and Smart Thermostats in a Demand Response ScenarioPosted for AESP
- Jul 25, 2019 2:27 am GMT
- 1029 views
By Christian Taber, Brittany Adam and George Myers
It is well known that increasing the setpoint of a thermostat that controls central air conditioning will decrease energy usage. The drawback is that occupant comfort is often sacrificed. Here we present a study that explores the potential of increasing thermostat setpoints and using automated ceiling fans to keep occupants comfortable, in order to offset electrical demand during peak utility hours.
As demand for energy rises, it is increasingly challenging for utility companies to supply electricity to all customers reliably during times of peak energy use. Demand Response (DR) technologies help accomplish this by shifting electrical use to times of lower demand. With total residential electrical consumption predicted to increase by 21 percent by 2040, utility companies must explore DR technologies to offset peak residential demands2. Studies have shown that residential air conditioning can account for 40 to 60 percent of that peak demand, so limiting or shifting air conditioner operation during these times is critical3.
MATERIALS AND METHODS
In the study, smart thermostats and smart ceiling fans were installed in seven homes of varying sizes in the Lexington, Ky., area, and monitored during a period from August 20 to October 12. The smart thermostats used Wi-Fi connectivity to communicate with the smart ceiling fans, which are themselves equipped with sensors that monitor temperature, humidity, and occupancy. The fans automatically elevate their speed to compensate for increasing temperature and humidity; and they are set in motion by someone entering or exiting the space.
Criteria for selection included:
- At least one occupant home during the day
- At least one room in the home with an existing ceiling fan and thermostat
- Willingness to complete comfort surveys
Homes were instrumented for indoor air dry bulb temperature, indoor air relative humidity, air conditioner power draw and runtime, ceiling fan operating speed, outdoor dry bulb temperature and occupancy. Each home’s occupants defined their original thermostat temperature setpoint, which is their preferred indoor air temperature with minimal air movement. The chosen setpoints were 68°F, 70°F (chosen by two homes), 72°F, 73°F, 74°F, and 75°F and were assumed to represent a comfortable condition for the respective occupant.
Case numbers were assigned to each home, and individual thermostats were programmed each week via the user interface. The homes were cooled to their original thermostat setpoints for over 12 hours before increasing setpoints by 2°F at noon. This two-degree increase simulated a smart system response to a demand call from the local utility. Throughout the afternoon, the homes were allowed to drift up to the increased setpoint, thereby “floating” through peak hours with minimal air conditioning use. Automated fans were used throughout this period to maintain comfort levels. The total operating time of the system during the simulated peak demand event and the length of time before the air conditioning system reactivated were recorded and analyzed. The homes remained at that temperature for one to two days before experiencing another simulated demand call; these days are referred to as control days.
During the next simulated demand call, thermostat setpoints were increased by an additional 2°F at noon, putting them four degrees above the original setpoints. If occupants felt uncomfortable or indicated they wished to change the thermostat setpoint on a given day, that day’s data was immediately marked as achieving “unsatisfactory thermal comfort,” and was not included in further analysis of load shifting within acceptable thermal comfort. Additionally, that home’s setpoint temperature was adjusted for the following days.
The data demonstrated the desired offset or delay in air conditioner operation from the beginning of peak hours (See Figure 2). In plots A and B, the indoor air temperature, air conditioner operating status (on/off), and setpoint were plotted over the peak hours for two similar homes on August 23. The top plot (A) is from a home that experienced no change in setpoint temperature. The home in plot A experienced only an eight-minute delay. The bottom plot (B) is from a home experiencing the simulated demand call from the utility. The home depicted in B experienced a 258-min. delay before the air conditioning turned on.
This delay did not fully characterize the potential demand reduction because of the 2°F deadband in which the thermostat does not activate the air conditioner (one degree above and below the setpoint). Since the home temperature was fluctuating in that deadband before the simulated demand call, the home temperature could be within a degree of the new setpoint temperature or as much as three degrees below it.
The depressed space temperature caused the effectiveness of the demand call to be dramatically increased, which created issues with analyzing the success of the test using this metric. The success of the simulated demand call from the utility to the thermostat is better gauged by the decrease in the total air conditioner operating time during peak hours, compared to control days.
The total air conditioner operation time after the simulated demand call and the air conditioner operation time during peak hours on the control days were unique for each home. The trend in mean air conditioner operation time (tAC) was highly correlated to the difference between outdoor temperature (TOUT) and setpoint temperature (TSPT), where ΔT = TOUT − TSPT (See Figure 3).
On the days of the simulated demand call, air conditioner operation time during peak hours was lower than on control days. The plot shape of the air conditioner operation time during demand calls shows the largest potential reduction in operating time at lower outdoor air temperatures. Savings were demonstrated at higher outside air temperatures, but the weather during this study did not include enough extremely high temperature days to accurately model the load shifting potential.
This suggests that the demand call increase of 2°F is even more effective on days where temperature difference between the setpoint and the outside air is smaller, unique to each home. This can be seen in the plot as the lower, yellow curve. However, in all the homes tested, the range of setpoint temperatures for which this method offers significant savings was larger than the range in which air movement can compensate for increased dry bulb temperature to maintain the occupants’ thermal comfort. The limiting factor is occupant comfort, not the range over which the simulated demand call effectively reduced air conditioning runtime during peak hours.
When a simulated demand call response increased the homes’ setpoints by 2°F, air conditioner operation times were reduced up to 70 percent (See Figure 4). The setpoint response to simulated demand calls offered significant reductions in air conditioner operation times when compared to the control days. Automated, intelligent ceiling fans compensated for the increase in thermostat setpoint, keeping occupants comfortable even in elevated air temperatures. Occupants indicated 100 percent comfort in the 2°F increase, and 71 percent indicated that they remained comfortable even at 4°F above their original thermostat setpoint.
This study included a limited number of hours at extremely high outdoor air temperatures, and further study of the capabilities of the system during periods of extremely high outdoor temperatures is recommended.
Since each home offered such consistent behavior, air conditioner operation time data might further serve as a unique input into more advanced demand response technology. Utilities could then offer performance based incentives for homes with longer delay times that historically have significantly lower total air conditioner operating time after a demand call. Under those circumstances, those homes would provide the same peak demand reduction as several homes with poorer performance.
Control logic for individualized demand response can be implemented to maximize the data already collected by the smart thermostat and the onboard sensors on the smart ceiling fans. The control logic is able to shift peak loads without short-cycling air conditioner units, and the fan maintains occupant thermal comfort.
Additionally, the proposed control would be compatible with each home through software integration, not hardware installation, allowing it to work with both legacy systems and modern, more complex air conditioning systems.
This study demonstrates the potential for adding smart ceiling fans to a utility’s portfolio of residential and commercial DSM programs, to achieve further peak demand reduction. Utilities looking for the next thing in load shifting should evaluate incorporating the technology used in this study to potentially improve program performance and increase customer satisfaction scores.
Christian Taber is Principal Engineer of Codes and Standards, and George Myers is Director of Tax Services at Big Ass Fans, which specializes in the design, manufacture, and sales of energyefficient products.
1 T. Hoyt, E. Arens and H. Zhang, "Extending air temperature setpoints: Simulated energy savings and design considerations for new and retrofit buildings," Building and Environment, vol. 88, pp. 89-96, 2015. http://escholarship.org/uc/item/28x9d7xj