In a smart grid environment, a WiMAX network has to support a large number of machine-to-machine (M2M) applications with diverse bandwidth, latency and network entry requirements. While M2M devices like smart meters and time-triggered sensors will periodically perform network re-entry to send data, a large number of event driven sensors might initiate network entry procedures simultaneously after an event. Without mechanisms to handle such a large number of network access requests, the entire ranging channel could be congested degrading the performance of the whole network. In this paper a novel ranging scheme is proposed that can provide contention free network access to the periodic M2M applications in a smart grid. We first evaluate the performance of the existing WiMAX ranging channel in terms of random access success rate, access delay and access throughput using a discrete event simulation model based on OPNET Modeler v16.0. The results indicate that at high random access loads, the access success rate drops sharply and the mean access delay increases exponentially. We then simulate a smart grid event where the underlying WiMAX network faces a heavy random access load and show that the proposed scheme is able to provide better performance in comparison to that of the conventional ranging scheme.