Real-time corrective control of active distribution networks: validation in future scenarios of a real system

Distribution grids are hosting more and more dispersed generation units exploiting renewable energy sources. Their proliferation, and the intermittent nature of their production creates new operation problems such as overvoltage and/or thermal overload. Hence, the need for realtime
corrective control goes increasing. This paper reports on comprehensive simulation tests of a centralized, coordinated control algorithm aimed at correcting abnormal voltages and/or branch overloads in an active distribution network. It relies on the model predictive control concept, for robustness with respect to model inaccuracies, noise, etc. The set-points of the dispersed generation units and the load tap changer are smoothly varied until the violation is cleared. The tests involve a real distribution network and plausible scenarios of future penetration of renewable energy sources. The dynamic evolution of the system has been simulated, with the controller in action, over full days where limit violations were taking place in its absence. Its capability to resolve the voltage and/or thermal issues is demonstrated. This corrective control allows postponing expensive network reinforcements and avoids to the largest possible extent curtailment of dispersed generation using renewable energy sources.