This paper proposes a methodology to determine both the technology of Energy Storage System (ESS) and its optimal sizing in order to provide secondary frequency control (SFC) to power systems with high penetration of wind generation. The objective is to determine the optimal investment in an ESS, considering the impact of the energy storage device on the costs of the electrical system and on the quality of frequency. The methodology allows calculating probabilistically the variable investment and operation costs of the ESS, considering the uncertainties associated with the stochastic behavior of the wind generation, conventional generation availability, network topology and the demand for energy. To this aim, a hybrid optimization using a meta-heuristic algorithm called Mean-Variance Mapping Optimization (MVMO) is utilized, whose control variable is the size vector (maximum power and energy capacity of the storage device), and an optimization model to compute the optimal power flow (OPF).