Sample bias and conditional probability ore reserve estimation at Valley

"At the time of this study, the Valley mine was producing 23 000 tonnes of ore per day from a large, low-grade porphyry copper orebody. Sulphide mineralization, principally bornite and chalcopyrite, occurs in a stock work of quartz veins in one phase of the Guichon Creek batholith. The property was explored from 1967 to 1982, and was placed in production in 1983.Using a three-parameter lognormal approximation developed by G.F. Raymond at Similkam een and Mount Isa Mines, a geo- statistical conditional probability ore reserve was estimated prior to mining. Comparisons with blasthole kriging and milled ore for the first 17 months of production showed close agreements at all cut-offs between conditional probability estimates and mineable ore. This was not true of either kriging or polygon estimates. An important part of this work was identifying and correcting for bias in both exploration and blasthole samples.IntroductionCominco's Valley lake deposit is a large, low-grade, porphyry copper orebody in the Highland Valley of British Columbia. Total indicated plus inferred ore reserves are 700 M tonnes at ore grade of 0.47070 Cu. At the time of writing, ore milling is at a rate of 23 000 tI d through the old Bethlehem mill. Mining is in 10m horizontal benches using 9.8 m3 shovels and 90 tonne trucks. Since this paper was written, Valley and nearby Lomex have been combined into Highland Valley Copper with a combined oreproduction of about 120 000 tI d.Copper mineralization, principally bornite and chalcopyrite, is regularly distributed throughout the deposit as disseminated low-grade within a granodiorite host rock, and as higher grade veinlets within a stock work of near-vertical quartz veins generally 10 cm or less in width. It is likely that problems with sample variability and sample bias are mainly related to mineralization within quartz veining. Valley geology has been described in detail by Allen and Richardson and by Osatenko and Jones (2). The exploration sampling program and identification of diamond-drill sample bias have been discussed by Armstrong and Newman'»."