All over the world, there is a tremendous increase in disaster occurrences such as landslide, toxic gas pollutions and wild fires. Least developed countries like Uganda have taken minimal efforts in management and containment of such disasters on behalf of the local populace. The dangers of fires are as a result of lack of proper information about the location, intensity and rate of spread of fire. However, the use of WSN may be one of the alternative ways of reducing risks associated with spreading fires resulting into destruction of lives and property worth millions of shillings. Our research looks at fire monitoring using sensors deployed in an event field to detect possible precursors of a fire occurrence using a simulated approach of OPNET Modeler (Ver. 14.0). Mobile nodes were deployed in proximity of (500X500) m2 flat space moving at a uniform speed of 10ms-1 for 30 minutes. Our simulations are based on ftp high priority traffic to reflect real time information transfer for analysis. This paper evaluates and discusses how sensed information can be transmitted through a network with minimal delay and proposes a scalable WSN architectural design based on protocol perspective i.e. AODV and DSR. Two QoS parameters have been considered i.e. delay and throughput. Our results show that AODV scheme has got a minimum delay of 0.2ms-1and a maximum throughput of 1.7Mbps. Hence it’s highly recommended for monitoring fires in large open area settings compared to DSR that is suitable for smaller areas. Results show that DSR exhibits higher delays due to nodal congestion and high control overheads. However, introducing the proposed heterogeneous routing mechanism i.e. (AODV+DSR) into the network significantly improves the performance of DSR.