ABSTRACT
The toll booth on the Kaduna - Zaria highway has been modelled as a queue system by adapting the makino model. Data collected on several days at the site was used to calibrate the model. Field conditions wherein a police check-point is encountered just before the toll booth justified the idealization of the system as a twostage queue. Allowance for U-turning in-between the police checkpoint and the toll booth was made by incorporating a fixed probability, deduced from field observations into the model. A wide range of vehicle arrival rates was simulated for each of the vehicles categories served. Results from the simulation correlated highly with actual observed simulations and the trends obtained conformed with theoretical values. For the Car channels, the expected waiting time ranges from 5.76 to 29.16 seconds; number in queue ranges from 0.4 to 9 cars; customer loss ratio ranges from 0.8585 to 9.3899 while utilization factor ranges from 0.4612 to 0.9541 and blocking probability ranges from 0.0000 to 0.4100. A number of options for improving the performance of the system was then investigated. Of all the options available for cars, the option which has one policeman in series with two collectors is adopted. Here, the two police channels are replaced by the super service channel, the three toll booths normally used for cars are reduced to just two channels.