Supermarket defrost cycles as flexible reserve

This work analyses how supermarket defrost cycles can be used as flexible reserve in a smart grid context. The consumption flexibility originates from being able to shift defrost cycles in time, while adhering to the underlying refrigeration systems constraints. It is shown how this time constrained behavior can be modeled and used by a defrost planner to optimally schedule defrost start times for several defrost cycles, according to electricity price. Because the defrost planning problem is a boolean programming problem a method of relaxing it is proposed. While the relaxed problem requires less computational effort, convexity is dependent on properties of the price signal. Simulation study show that the optimal planning strategy can reduce cost of defrost cycles with 31.7%, compared to a current strategy of executing all defrost cycles at the same time each day.