Demand response plays an important role in the development of the smart grid, which can effectively manage society's energy consumption. Cooling devices, such as refrigerators and freezers, are ideal devices for demand-response programs because their energy states can be controlled without reducing the lifestyle and comfort of the residents. Conversely, managing air conditioning and space heating would affect a resident's comfort level. Direct compressor control and thermostat control methods have been proposed in the past for controlling cooling devices but they are never studied concurrently. This paper proposes a new control mechanism and compares the effectiveness of the three control mechanisms for cooling devices in demand response. In addition, this paper illustrates the need for a damping strategy to mitigate demand oscillations that occur from synchronous fleet control.

No abstract

With the help of smart meters, power companies can remotely collect customers' power consumption and setup demand response programs accordingly. While there is no standard for the metering interval, many power companies collect data at hourly basis. Hourly measured data is sufficient for providing billing information and feedback on energy use. However, it does not reflect the true inter-hour dynamics of power and energy usage of residential or commercial consumers. Higher resolution electricity consumption information is important in setting up real-time demand response programs. In order to increase the information value of the collected data, power companies may simply increase the metering resolution or to adapt alternative metering methods. Threshold metering, which acquires data only when the change of measurements exceeds a certain level, is an alternative metering option. This paper compares interval and threshold metering methods with different settings for measuring domestic power consumptions. Instantaneous power consumption data are collected from four different houses in Calgary, Canada and these data are used to generate test data for interval and threshold metering study. Numerical results of statistical and accuracy measures and data size tradeoffs are provided to support our discussion.Index Terms-residential load modeling, smart grids, energy resolution, demand response.Anthony Schellenberg (M'02) received his B.Sc. and Ph.D. in Electrical Engineering from the University of Calgary in 2002 and 2006, respectively. His research interests include power system optimization, optimization under uncertainty, and stochastic optimization.