Kriging As A Tool To Avoid Overestimation Of Grade In Sulphide Orebodies

This case study compares the actual production figures for the No. I and No. 2 copper orebodies of the Société Miniére Louvem, Val d'Or, Quebec, with the preproduction estimates and a postmortem geostatistical estimation using kriging techniques. The deposits considered are small subvertical sulphide orebodies which were mined from August 1970 to April 1973, producing 648,000 tons at 2.04% Cu versus the 1968 computed reserves of 671,000 tons at 2.51% Cu (with 20% dilution), thus showing a 25% overestimation of the copper contained. The postmortem geostatistical estimation based on the same 44 D.D.H. intersections available in 1968, shows 2.24% Cu (no dilution), compared to 3.23% Cu (no dilution) given by the conventional section method. Block by block kriging techniques were used, but within the ore contours as determined by the geologist. In this case, Characterized like many sulphide deposits by highly dispersed sample values, geostatistical kriging techniques permit a more realistic estimation of the metal contained within a given volume, than the conventional methods which overestimate it. This study has been conducted by the Mineral Exploration Research institute (IREM-MERI) for SOQUEM, and is one of the few case studies released with all pertinent data. More postmortem of this kind are required for geologists and geostatisticians to improve their methods and their "batting average".