Analysis of earth-moving systems using discrete-event simulation

This paper outlines experiments performed on a simulation model of excavator-dump-trucktype earth-moving operations with the intention of indicating the parts of an earth-moving system to which the output is most sensitive. Response-surface methodology was used initially to indicate the relationship between two factors: The truck travel time, from loader to dump area and back; and the truck spot (or maneuver) time at the loader. This method indicated that the form of the relationship between these factors varied at different values but did not indicate how often factors affected the output. Full factorial designs were then done that not only were more economical in terms of the number of experimental runs required but also indicated that the most important factors are number of trucks, the haul and return (travel) time, the number of passes per load, and the loading rate. The simulation has highlighted a number of sensitivities. If load pass time is reduced then production will not increase if the operation is already under resourced. Conversely, production will not automatically increase by adding more trucks to the operation.