Energy-agile laptops: Demand response of mobile plug loads using sensor/actuator networks

This paper explores demand response techniques for managing mobile, distributed loads with on-board electrochemical energy storage over a plug-level sensing/actuating wireless mesh network. We target laptops and construct a power consumption and battery charging model from measurements obtained across a variety of such devices. We then build simulations of charging patterns using the most general cases we observed. Our first simulation study explores a classic demand response scenario in which a large number of loads participate in a typical pre-scheduled demand response (DR) event. We show that we can achieve load curtailments in the range of 30-90% of aggregate baseline load as a function of the duration of a DR event by managing the charging schedules of laptops that randomly enter and leave the control jurisdiction of a DR event participant. In a second simulation study, we investigate a continuous demand response scenario in which charging schedules respond to a fluctuating renewable electricity supply (e.g. wind or solar) and show that we can reduce grid dependence by 26.8-33.8% compared to oblivious charging.