Modelling of cement raw material compositional indices with direct sequential cosimulation

Spatial distribution of cement raw material compositional indices is a precondition of conducting quarry stope layouts. These indices are calculated from chemical components, such as SiO2, Al2O3, Fe2O3, CaO and MgO, and the most important are LSF (lime saturation factor), SIM (silica modulus), ALM (alumina modulus) and CS (lime silica ratio). Geostatistical images of the grades are more useful in this respect, as they reflect different representations of the reality and show the spatial distribution more accurately. This paper presents the construction of sets of simulated images of the chemical components SiO2, Al2O3, Fe2O3, CaO and MgO as primary variables, using the direct sequential simulation (DSS) and cosimulation (CoDSS) algorithms. For the joint simulation, a novel methodology uses principal components (PC) as secondary variables. Finally, the simulated images are combined algebraically in order to estimate local distribution functions of the compositional indices (LSF, ALM, SIM and CS), showing whether the local probability of each index is within the optimal interval.