Today's batteries are commercially developed technologies and are commonly used for energy storage in active distribution grids. In spite of their numerous advantages, batteries are expensive; consequentially, they should be installed and managed in an optimal manner. Furthermore, every battery technology has particular technical and economic characteristics that may be particularly well- or poorly-suited to certain applications. This paper represents a comparison based optimal planning of several battery technologies to find the best choice in distribution grid applications. The proposed planning methodology is a novel four-layer procedure that considers the uncertainty of battery characteristics as well as load and wind power. The long-term planning layer optimizes the location, capacity and power rating of batteries. The short-term scheduling layer includes the probabilistic optimal power flow with respect to technical constraints. In the uncertainty modeling layer, the technical and economic characteristics of battery technologies and load demands are modeled using fuzzy values. Moreover, in order to consider the correlations and independencies of the wind power profiles, they are classified to several categories using factor analysis technique in the classifying layer. The numerical results show that Zn-Br technology is the most suitable option in deterministic studies, however, the Na-S technology can be an alternative in uncertain conditions. Several sensitivity analyses are carried out to generalize and extend the results. Finally, it is found that more storage capacity is required in realistic conditions thanks to using factor analysis technique. (C) 2016 Elsevier Ltd. All rights reserved.