A new paradigm based on the concept of virtual coupling of train sets is being elaborated to provide innovative solutions so as to increase efficiency, operational flexibility, line capacity, competitiveness among market players and quality of consumer experience. Train-to-Train (T2T) and Train-to-Ground (T2G) wireless communications will be the backbone for the implementation of the virtual coupling functionalities. In the present paper, we perform a model-based dependability analysis of the wireless communication system under virtual coupling operation. Namely, Stochastic Colored Petri Net (SCPN) models are developed to depict the exchanges of the various information needed under virtual coupling operation. Dependability evaluation is then performed by means of simulation. In particular, the impact of various communication parameters is scrutinized while taking into account different operational scenarios. The obtained results allow the identification of the most impacting aspects on dependability analysis and provide valuable inputs to support the technological choices in terms of communication to implement virtual coupling.