The objective of this activity is to validate an advanced real-time control framework for power distribution networks in order to control grid-connected battery energy storage system (BESSs) to satisfy the control’s objectives such as tracking a day-ahead dispatch plan of a distribution network hosting controllable and stochastic heterogenous resources. The control framework accounts for the grid constraints on the nodal voltages and lines and transformer capacities. The control algorithms rely on the availability of both short and day-ahead forecasting of the demand and PV generation developed in the context of feeder dispatching. In the scheduling phase on the day before operations, a stochastic optimization problem computes an aggregated dispatch plan at the grid connection point (GCP), accounting for the uncertainties of demand and PV generation via scenarios, and constraints of the grid and the controllable resource. For the day-ahead scenarios, we develop a Markov’s chain based forecasting method where we cluster the historical measurements for each day type and fit multi-variate Gaussian distribution for each cluster. In the real-time phase, a grid-aware model predictive control (MPC) computes the active and reactive power set-points of the battery so that it tracks the dispatch plan at the GCP while obeying to its constraints and those of the grid. The MPC problem leverages short-term forecasting of the demand and PV generation. The proposed control framework is validated by dispatching the operation of a 12kV/20MVA MV distribution network in Aigle, Switzerland (i.e. the REeL demonstrator) using a 1.5 MW/2.5 MWh BESS, which is controlled in real-time given the online grid state estimation enabled by the deployed distributed PMU-based sensing infrastructure.