Incorporating Cycle Time Dependency In Truck-Shovel Modeling

The need to optimize surface mining operations has led to the use of discrete event simulation (DES) modeling of truck-shovel systems. Often, these models assume truck cycle times are independent and identically distributed (iid) random variables, although, with significant bunching on the haul routes, this may not be valid. The objective of this paper is to present a methodology to (i) account for truck bunching due to slow trucks; and (ii) test for cycle time dependence (i.e. whether truck cycle time data is iid or not). We built a DES model that accounts for bunching due to a slow truck(s) and used the cycle times as data for further analysis. We then tested for cycle time dependence using Pearson correlation coefficients. The results show significant correlation of truck cycle times for different number slow trucks, with variable speeds. Assuming iid in modeling results in over-estimation of productivity and the uncertainty surrounding it. The over-estimation of uncertainty is pronounced when the relative difference in truck speed increases. The confidence interval of the means of truck loads/shift decreases with increasing number of slow operators. This work extends the usefulness of DES in truck-shovel applications.