Community-level Geothermal Heat Pump system management via an aggregation-disaggregation framework

This paper presents an aggregation-disaggregation framework for the community-level management of widely-used Geothermal Heat Pump (GHP) systems in energy efficient buildings. In accordance with the layered operating architecture of the current power grid, this framework operates at two different time-scales. At a slow time-scale, each building predicts its GHP system future power consumption information consisting of both upper and lower bounds and the utility function, using aggregate thermal zone information and short-term disturbance forecasts. The information is then reported to a control center, i.e., an aggregator for all GHP systems in the community who determines and notifies their future nominal consumptions. At a fast time-scale, a distributed control scheme is designed based on a primal-dual gradient method, such that the nominal consumption can be tracked as closely as possible under real disturbances, using only local measurements and neighboring communications. In general, our framework provides an approach to scaling up the optimization and control for zone-building-aggregator interaction, which can be used to provide flexible ancillary services from a population of GHP systems to support the power distribution/transmission systems.