Simulation results depend on many parameters such as: modeling assumptions (e.g. Gaussian versus Indicator simulation; point simulation versus block simulation); practical implementation (e.g. sequential simulation versus turning band); and case specific parameters (e.g. top-cut values). The user is generally aware of the impact of case specific decisions because they are very similar to the decisions taken during an estimation exercise. However, the user is often in the dark when it comes to the impact on the results of modeling assumptions and software implementation. Not much literature is available; checks are difficult to complete because of the random number generator imbedded in the simulation algorithm. This paper is a comparative study of several Gaussian related simulations for gold at Boddington Gold Mine, a large porphyry gold/copper deposit. Point and block Gaussian simulations were considered. The study showed that careful calibration/validation is necessary in all cases to avoid significant biases. Point simulation is easy to validate against primary data because the support does not change. Block simulation is more difficult to validate because of the change of support. All software/algorithms were able to provide what they have been designed for, i.e. simulated values that are conditional and that reproduce the required grade point or block distribution and variogram. Surprising differences, however, were noted between the re-blocked point simulation results and the direct block simulation results. Differences in application and results of the methods together with advantages and disadvantages are presented and discussed. |
