Statistical analysis of blasthole sampling data from the Paradise Peak mine, Nye County, Nevada

Abstract

Various production sampling tests were performed during 1987 and 1988, at the FMC Gold Paradise Peak precious metal open-pit mine. Testing of blasthole hammer and reverse circulation drill sampling was conducted to determine what degree of sampling error (bias) was being introduced by drilling methods and sample acquisition techniques. Statistical analysis of sample data obtained from run-of-mine and high grade ore, using a below- the-deck frame mounted bag, has indicated that on average a sampling error of 24 to 29 percent can result from production bench drilling employing air hammer methods. Tests employing a below-the-deck wedge shaped pie pan have demonstrated that on average sampling error can be reduced to 17% due to an improved representation of the actual chip pile particle size distribution. However, problems remain with over and underfilled sample containers due to various drilling conditions. Comparison of rotary reverse circulation with down-the-hole hammer sampling of a mill feed grade ore test block (31,000 ton), has shown that on average reverse circulation upgraded over blasthole hammer sampling by as much as 6 to 7% for gold and 8 to 11% silver. Bulk sampling of the entire test block during mill feed transport to the fine ore bin, indicated that the bulk chip pile samples for 140 blastholes upgraded by 0.03 oz./ton gold and 2.2 oz./ton silver over the average mill bulk sample grade. Upgrading of both reverse circulation and blasthole hammer sample data suggests that significant sampling error is being introduced in the form of bias that produces non-representative (upgraded and downgraded) assays caused by contamination or loss of particles into the actual container (bag or pan) employed. The main recommendations obtained from these tests and their statistical results are that 1: to improve acquisition of the chip/dust particle size distribution that is delivered from the blasthole, it is essential that larger sample volumes be taken (i.e. through-the-deck pie pan or larger bag(s)) in conjuction with the use of full wrap-around drill skirts, 2: consideration should be given to develope an automated sampling system incorporating down-the- hole sample acquisition, cyclone collection, and on-site splitting.