Electric Network Frequency is the frequency of power distribution networks in power grids that fluctuates about a nominal value with respect to the changing loads.Its ubiquitous nature has made notable contributions to forensic analysis that has substantiated its use as a significant tool in this area. In this paper we have proposed a technique to identify the power grid in which the ENF containing signal was recorded, without the assistance of concurrent power references. The recognition of the appropriate power grid is facilitated by certain statistical attributes extracted from the varying ENF signals that serve as an instrument of comparison. These features are distinctive to the grid-of-recording and are given as inputs to the training phase of a machine learning system. The learnt characteristics are then applied by this system to match the ENF signals to their respective power grids with utmost efficiency.