This paper describes an integrated approach to the evaluation of ore loss and ore dilution via geostatistical simulation of in situ geology and grade characteristics, simulation of blasting mechanisms and simulation of loading from a blast muck pile. The approach is demonstrated on a case study._x000D_
The sequential Gaussian method is used to simulate in situ bench grades on the scale of the smallest selectable volume within a blast muck pile. The simulated grades of these component volumes are interpreted as the reality that would be available with exhaustive sampling. The simulated bench is subjected to a blast simulator, based on the SCRAMBLE model, which effectively moves each of the component model volumes to its final resting place in the blast muck pile. Ore/waste delineation and selective loading processes are applied to the simulated muck pile to determine the degree of selectivity that can be achieved by various sizes of loader and types of loading and to quantify ore dilution and ore loss._x000D_
An in situ model representing the reality of knowing only the data and information that are available from specific grade control drilling and sampling grids is obtained by sampling the geostatistically simulated model on a specified grid. The volumes comprising the in situ model are then populated by estimates based only on the data and from the specified grade control drilling and sampling grids. Different drilling and sampling grids are used to generate different models, each reflecting the levels of data and information available._x000D_
Selectivity is assessed as a function of the drilling and sampling grids as well as the size and type of loader. Performance is assessed against the ideal selectivity that could be achieved on the perfect knowledge model comprising the simulated values of each component volume. Costs, prices and financial criteria enable an optimal selection of the grade control drilling grid, size of loader, type of loading and blast design.