Energy balance and ancillary services provided by distributed storage systems to active distribution networks represent two aspects of a single problem that needs to be properly treated in view of the typical distribution networks parameters. In this context, the paper focuses on the problem of the optimal siting and sizing of distributed storage systems. In particular, the paper proposes the formulation of a problem that accounts: (i) the voltage support of storage systems to the grid, (ii) the network losses and (iii) the cost of the energy-flow towards the external grid. As the formulated problem is mixedinteger, non-convex and non-linear, its solution requires the adoption of heuristic techniques. In this respect, a two-stage iterative procedure is proposed. The first stage of the procedure utilizes a genetic algorithm to provide locations and sizes of the distributed storage systems; the second stage evaluates the fitness of the solution provided by the first part by solving a daily AC optimal power flow. An application example, referring to the IEEE 13 busses test feeder, is included in order to demonstrate and discuss the efficiency of the proposed method.