A mine operator's implementation of geostatistics

CIM Bulletin, Vol. 80, No. 899, 1987

N.I. NORRISH and G.H. BLACKWELL, Brenda Mines Ltd. Peachland, British Columbia

At most operating mines, geostatistical ore reserve estimation methods have been implemented by head office or consulting groups. With recent advances in small computer technology it is now possible for a remotely-located mine department to independently implement a geostatistical package. This paper describes the evolution of geostatistical routines to calculate the mineral inventory for porphyry type copper/ molybdenum or similar ore bodies. The problems encountered in down sizing the geostatistical routines to run using limited minesite computing facilities are discussed. The high quality of the results, in spite of some compromises to geostatistical theory, is emphasized. The first stage entailed the development of a mineral inventory based on diamond drill data. The difficulties in fitting variogram models to two metals, copper and molybdenum, and the shortcomings of the mineral inventory due to the mixing of lesser quality exploration data with data from more recent diamond drilling are discussed. To improve the ore reserve estimates for short-term planning, a mineral inventory was produced using both the diamond drill and production blast hole drilling data. This inventory, when compared to actual grades, was found to be quite satisfactory for short-term planning. The inventory also provided a reliable data base for long range pit design and financial planning. By regularly updating the mineral inventory, many of the uncertainties of mining a low-grade orebody were minimized.
Mots Clés: Mine planning, Geostatistics, Ore reserves, Mineral inventory, Kriging, Grade control.