Unidirectional probabilistic direct control for deferrable loads

The idea of harnessing the inherent flexibility in demand of many types of electric loads has been largely discussed in the last years for coping with the need to maintain the energy demand-supply balance. In particular, the fine tuning of the operation conditions of different thermostatic loads (such as air-conditioning, refrigerators, etc.) has appeared as the most natural solution for load control with minimal user discomfort. In this paper we focus on an alternative approach: deploying simple open-loop control strategies for deferrable loads with minimal communication overhead. The idea is to send a multicast control message to a group of users, on the basis of the expected and desired load profiles, for probabilistically enabling or deferring the activation requests of a specific load type. The control law and the most important performance metrics can be easily derived analytically. Despite the simplicity of the approach, which requires minimal or null investments, our results show that significant load shifts can be achieved.