A New Method for Handling Lockout Constraints on Controlled TCL Aggregations

Thermal loads are recognized as a valuable source of flexibility in face of the increasing variability caused by the large shares of renewable production. Lockout constraints can significantly reduce the flexibility of thermostatically controlled loads (TCLs). We propose a novel way of modifying the loads' lockout durations to achieve non-intrusive centralized control without relying on local computations and estimations. We derive analytical expressions for the flexibility reduction and validate them via simulations, which show that the proposed method describes the TCLs flexibility accurately. We further show that a simple stochastic centralized controller, which does not rely on local temperature measurements, outperforms the commonly used priority-stack controller in terms of system robustness against infeasible trajectories.