Geostatistical simulation of RQD variable using investigation of spatial continuity between Quaternary alluvial layer and hard rock of Gohar-Zamin mine to the determination of permeable zones

In this research, a sequential Gaussian simulation method has been used to determine the permeable zones in the hard-rock aquifer of the Gohr-Zamin open pit mine. For this purpose, 4946 RQD data from eighty-seven exploratory boreholes was used and exploratory-spatial data analysis of these data was performed using the preliminary statistics, location maps, histograms and variograms. Results of preliminary statistics show that this variable has not a normal distribution and that the statistical parameters are not significantly affected considering the declustering weights. The RQD data was transformed to a normal distribution with a Gaussian transform. In the next step, the omnidirectional/directional experimental variograms of RQD data were calculated and modeled using the standard models. Three directions of (1) Azimuth 0-Dip 0, (2) Azimuth 90- Dip 0 and (3) Dip 90 show the best spatial continuity. These variograms were combined together in a one single model. The highest spatial continuity is in the east-west direction compared to the other two directions. Finally, the sequential Gaussian simulation of RQD variable was performed using the 3D modeled variogram and normalized data. A probability map was generated using the results of 100 simultaions. Results show that the southeastern, southwest, and western parts of the mine pit are the best places for drilling new pumping wells. Introduction Gohar-Zamin iron ore mine is located in 50 km southwest of Sirjan city in Kerman province. Due to the tectonic condition, fracture and fissure in the hard-rock aquifer of area, the groundwater has been concentrated on the pit mine. Therefore, identifying the permeable zones is essential to mine dewatering process. Materials and methods In this research, ISATIS software has been used to sequential Gaussian geostatistical simulation. The exploratory-spatial data analysis was performed including preliminary statistics, location map, histogram and variograms of RQD values. Then the cell declustring method is used to decluster the RQD variable. Moreover, Gaussian transform method is applied to data normalization. In order to investigate the spatial continuity of the RQD variable, the omnidirectional and directional variograms were generated. Finally, sequential Gaussian simulation as well as simulation uncertainty analysis have been performed using three-dimensional variograms of the RQD normal values. Results and discussion It is clear from the indicator maps that high RQD values are concentrated in the south of pit and deeper levels. It is influenced by the shape of the iron ore deposit. According to the value of skewness and kurtosis coefficients of RQD, it seems that this variable do not has the normal distribution. The 3D variogram on the XY, XZ and YZ planes indicate (a) lower spatial continuity in the north-south direction relative to the east-west direction, (b) spatial continuity in the eastern-western axis is greater than the Z axis and (c) spatial continuity in the Y direction is greater than the Z direction. The final floor level in the pit is very important for designing and determining the position of pumping wells to dewatering process. Therefore, results of 100 simulations were overplayed. The mean values of simulations calculated from the final floor level in the pit to the top of hard-rock aquifer. It was observed that at each elevation, minimum of RQD were concentrated in certain regions. These maps can be used to determine the high permeable zone. Also, two main strike-slip faults with an almost NW-SE and E-W trend, pass through the southern part of pit. These fults and crush zones are located almost in the same direction. Conclusion According to the results of study, the RQD variable does not follow the standard distribution and declustering weight on the data did not have much effect on the mean value. It indicats a fairly uniform distribution of RQD data in the study area. Omindirectional and directional variograms of the RQD variable (normal values) showed that this variable has good spatial continuity. Moreover, the three-dimensional variogram of the RQD indicated that the E-W direction has the highest spatial continuity relative to the two N-S and Z directions. While, the spatial continuity in the N-S direction is more than the spatial continuity in Z direction. Results showed two straight-slip faults with NW-SE and E-W trends are located in the south of pit and These have important role in greating the crushed and permeable zones. Sequential Gaussian simulation and uncertainty maps in various parts of the pit showed that regions of southeast, southwest, and west of the pit have the highest probability is the low RQD and the high permeability.