PITA FY2014 Projects

Modeling and Tracking Aggregations of Thermostatically Controlled Loads in Buildings for Demand Response Services

PI:Mario Berg├ęs, Civil and Environmental Engineering

University:Carnegie Mellon University


Industry Affiliate(s):kWantera

Summary:Existing electrical power reserve assets and regulation methodologies fail to provide the shortterm responses required to keep the load and generation balanced as the amount of renewable generation increases. Researchers have proposed to increase the information exchange within the power network and to introduce direct load control (DLC) to ensure robustness while accommodating the intermittent nature of these new generation resources. One particular type of load that has received much attention as of late in the research community, due to their large penetration rates and other desirable properties, are Thermostatically Controlled Loads (TCLs) such as refrigerators and electrical water heaters.

In this project, we propose to use TCLs in buildings to provide regulation services for the power grid. We would like to shed light on the system-level benefits that a heterogeneous TCL population can provide through centralized control. In particular, our goal is to create a hybrid test bed where a portion of the TCLs available in Carnegie Mellon University (CMU), along with simulated TCLs, are controlled simultaneously close to real-time to provide regulation services. The motivation for this effort is the limitations of using traditional control methods to maintain system stability and quality of service. We believe that real-time demand control through populations of TCLs provide a more agile alternative that compensates for system-level variations, decreasing the need for more costly generation alternatives, such as gas turbines. In addition, the proposed hybrid demonstration will provide significant information on real-world implementation of real-time demand control algorithms previously developed by our research team.

As a local industrial partner who has expertise in assisting commercial and/or industry customers in better managing their electricity demand, kWantera is providing support by allowing the researchers to share hardware, and provide some of the backend infrastructure for the proposed centralized control platform. Similarly, Sensor Andrew, a CMU project previously funded by PITA, is leveraged as a sensor middleware platform.