In this paper we apply PSim, the three phase discrete event simulation software on the Bejaia aggregates quarry and see the best way to compute the performance measures using the mean, median or mode. The behaviour of this case-study depends on input random variables with known probability density functions. A logical model for the quarry aggregates using an expanded Activity Cycle Diagram is built from which we have simulated the quarry on a computer. So, experiments are carried out on the model built and unknown parameters of the output random variables of interest are estimated. To do so, Refined Descriptive Sampling (RDS) was chosen between different Monte Carlo methods, to generate input random variables. The performance measures of the aggregates quarry system are established using different parameters of position, such as the mean, median and mode.  Different replications were carried out to study the variance of each performance measure using the cited parameters of position. Different curves of variances were plotted using the SCILAB programming language.  Generally, these curves have shown that the mode gives the best results. We conclude that it’s better using the mode to estimate output parameters when using RDS to generate inputs.