Benchmarking Estimation Methods for Coal Resource Estimation

The coal industry in Australia has been actively working in recent years towards the (re)integration of geostatistical techniques to the process of coal resource estimation. The benefits of using geostatistical techniques, over the current interpolation algorithms in use in coal modelling packages used to generate the grid estimates, need to be illustrated for the methodology to gain a broader acceptance in the coal industry. This paper focuses on the benefits gained at BHP Billiton Mitsubishi AllianceÆs Saraji mine for the estimation of the total product yield value at a 9.7 per cent product ash cut-off of a particular seam of the Saraji deposit. In order to perform this comparison, the results of the alternative estimation methodologies need to be benchmarked, which in most situations is achieved through the reconciliation of the estimates against production and/or process plant reconciliation data. Unfortunately, this benchmarking was not possible as no reconciliation data for the area of interest in the chosen seam was available, as it has not been mined at the time of the study. Where no reconciliation data is available, an alternative method is to use conditional simulations as the benchmark for comparison. The simulated values can be regarded as ærealityÆ, and the alternative estimation methodologies, based on a sampling of the underlying ærealityÆ, can then be compared using the simulated ærealityÆ as the benchmark for comparison.