Smart Homes with Price-Responsive Thermostats

We develop a framework for a smart home's storable appliances to respond optimally to dynamic electricity price signals, and for assessing the resulting market price equilibrium in a large service region. Such appliances are characterized by their ability to manage a store of energy, of which air conditioning is our canonical example. Starting with rst principles of consumer and appliance behaviours, we develop two mathematical models for smart price-responsive storage appliances: a "price-only" model which is not aware of home occupancy, and an "occupant-aware" model. We show that under certain conditions it is socially optimal for the electricity utility to pass spot prices down to consumers. We present extensive numerical results on ComEd's residential customers' prospective responses to dynamic prices through air onditioners during a hot summer month, both in isolation and in equilibrium. Our results suggest that dynamic prices reduce power bills significantly and even more so with price-responsive appliances. On the other hand, they increase power bills significantly on peak days while price-responsive air conditioners mitigate these bill increases. Overall, the social welfare may increase up to 2.4% for the month and up to 6.2% on a peak day. However, our results indicate that air conditioning alone is not sucient to account for the peak load reductions observed in many real-world pilot tests, unless consumers are willing to experience substantial thermal discomfort.