Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE THIRD INTERNATIONAL CONFERENCE ON SOFT COMPUTING TECHNOLOGY IN CIVIL, STRUCTURAL AND ENVIRONMENTAL ENGINEERING
Edited by: Y. Tsompanakis
Influence of Data Density on Grade Estimation in a Limestone Quarry using Kriging
Y. Chen1, M. Bessho2 and T. Ito3
1Graduate School of Energy Science, Kyoto University, Kyoto, Japan
Y. Chen, M. Bessho, T. Ito, "Influence of Data Density on Grade Estimation in a Limestone Quarry using Kriging", in Y. Tsompanakis, (Editor), "Proceedings of the Third International Conference on Soft Computing Technology in Civil, Structural and Environmental Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 15, 2013. doi:10.4203/ccp.103.15
Keywords: kriging, limestone quarry, grade estimation, .
In order to construct an effective management of the product from a limestone quarry, the correlation between the precision of interpolation and data density when estimating the distribution of CaO grade in quarries by kriging was considered. The reference distribution map of which resolution is 1 meter was firstly generated for the square area of 160 x 160 meters using the grade measurement data of 1,472 cuttings from an operating limestone quarry. Then 8 testing distribution maps with the resolution of 2 to 9 meters (i.e., 25.6 to 1.4% data density compared with the reference distribution map) were prepared by thinning out the unnecessary data of the reference distribution map. These testing distribution maps were compared with the reference distribution map. As a result, reasonable CaO grade distribution was obtained for the testing distribution with 2 to 6 meter interval data (25.6 to 3.1% data density). It is consequently confirmed that the precision of interpolation decreases as the data density decreases, but the decrease tendency is not linear in the test conditions. In the point of error percentage, grade distribution estimated by the testing data set with 5 meter intervals (4.3% data density) shows good interpolation results with less than 1% error range for more than 75% of the analyzed area of the working face, while more than 50% of the area is beyond the 1% error range using the test data set with 9 meter intervals (1.4% data density).
purchase the full-text of this paper (price £20)