Unexpected, catastrophic events initially result in anxiety and confusion. The congestion of pedestrians and vehicles further exacerbates the already turbulent environment. This uncertainty results in a chaotic site of victims, curiosity seekers, emergency responders, and mere transient individuals. Initial public announcements are often delayed, inaccurate, and met with skepticism and disbelief. However, interpersonal communications, although potentially less accurate, may have a greater influence on behaviour. With this in mind, an agent-based model was developed that included vehicle movement on roadways with traffic lights and stop signs, a disaster site, evacuation sites, and communication systems consisting of both public broadcast and personal communication networks. As an extension of earlier work, we explored the effect of the public announcement control variables and contrasted the results with those from a personal communication network. In addition, the model is thoroughly explored, explained, and shown to demonstrate the potential for improved site evacuation via personal communication networks, i.e., cell phones.