Abstract

In this second part, we evaluate the performances of our control framework by applying it to a case study that contains a minimum set of elements allowing to show its applicability and potentials. We show how the computation of the PQt profiles, belief functions, and virtual costs can be synthesized for generic resources (i.e., dispatchable and stochastic generation systems, storage units, loads). The metrics of interest are: quality-of-service of the network represented by voltages magnitudes and lines current magnitudes in comparison with their operational boundaries; state-of-charge of electric and thermal storage devices; proportion of curtailed renewables; and propensity of microgrid collapse in the case of renewables overproduction. We compare our method to two classic ones relying on droop control: the first one with only primary control on both frequency and voltage and the second one with an additional secondary frequency control operated by the slack device. We find that our method is able to indirectly control the reserve of the storage systems connected to the microgrid, thus maximizing the autonomy in the islanded operation and, at the same time, reducing renewables curtailment. Moreover, the proposed control framework keeps the system in feasible operation conditions, better explores the various degrees of freedom of the whole system and connected devices, and prevents its collapse in case of extreme operation of stochastic resources. All of these properties are obtained with a simple and generic control framework that supports aggregation and composability. (C) 2015 Elsevier B.V. All rights reserved.