The growing demand for flexibility services to control the electricity grid is sparking a great interest in the aggregation of distributed energy resources. This report deals with methods and tools for managing virtual aggregates consisting of distributed generation, especially from renewable sources, storage and loads. In particular, a prototype of an Energy Management System was implemented based on a stochastic optimization algorithm in which the uncertain variables are the next day load profiles, photovoltaic production, and energy price. Optimized stochastic management allows you to define a device usage profile such as to well satisfy (i.e. independently of the trend of the uncertain variables) the load and technical constraints, and guarantee a minimum profit with a certain probability. We used a scenario-based approach, which is a technique for solving probability-constrained problems.

At a later time, stochastic optimization techniques were applied to a case study consisting of an aggregate of residential photovoltaic storage and generation systems, in order to evaluate the flexibility margins that could be used to participate in energy Day-Ahead Market (DAM) and Dispatching Services Market (MDS). The approach used to solve the problem was formulated with strong constraints and an expected value objective function. Furthermore, the document reports the results of two pilot projects in which two BSPs (Balance Service Providers) participate with two aggregates of residential users equipped with home storage systems. In 2020, the first tests of user remote control were carried out, the first residential users were added to a UVAM (Virtually Aggregated Mixed Units ) already existing in the MDS and other contracts are being consolidated to form a UVAM with a greater number of residential users.