This paper investigates the possibility to use a new modeling of Energy Storage Systems based on zero integral functions. Such functions represent the course of the energy level stored in batteries during the solution of optimal management problems in smart-grids. Storage devices, such as all the other components that are required to meet an integral capacity constraint along the dispatch time, must show the same State of Charge at the start and at the end of the timeframe considered for operation. In this paper, a set of sinusoidal functions have been used for the synthesis of the charge and discharge course of energy Storage Systems. Such representation allows to eliminate the difficult constraint about the State of Charge at the start and at the end of the considered timeframe and in general allows to reduce the number of parameters involved in the optimization. A few tests of the proposed representation for an optimal dispatch problem solved with evolutionary computation in a small distribution system is considered in the application section.