This paper investigates the usage of radio frequency (RF) signal as a source in energy harvesting systems. The open issue in the related literature is the characterization of battery recharging time of an energy harvesting receiver node. RF energy harvesting has challenges due to the wireless propagation environment and conversion efficiency. On the propagation side, the different channel models between source and harvesting node should be taken into account in order to obtain realistic results. The main goal of this paper is to propose statistical models for battery recharging time for the Nakagami-m and the generalized-K fading channels. We also include the effects of lognormal shadowing. We derive the associated closed form probability density function, cumulative distribution function, moment generation function, mean and variance expressions for battery recharging time. The simulations are used to verify the theoretical results.