Blasthole sampling (replicate and variographic experiments) in LKAB open pit iron ore mines – fit-for-purpose representativity?

• Main Authors: Engström, Karin, Esbensen, Kim Harry
• Format: Article in Journal/Newspaper
• Language: English
• Published: Australasian Institute of Mining and Metallurgy 2017
• Subjects: Kiruna
• Online Access: https://vbn.aau.dk/da/publications/54876355-ab5f-41c2-af5f-1a67bfc4c9d0; https://www.ausimm.com.au/publications/epublication.aspx?ID=17242

The conclusion from the debate over the last decade in the mining sector is that, for grade control, reverse circulation (RC) drill sampling gives more reliable results due to less sampling problems than blasthole samples. However, by conducting the latter with well-controlled procedures there still exist fair possibilities to counteract the major blasthole sampling problems and achieve fit-forpurpose representativeness, especially if based on minimal loss of fines. In the present PhD study, extensive replicate experiments and variographic characterisations are conducted to evaluate the performance of manual blasthole (BH) sampling and sectorial segment sampling. The experiments were performed in LKAB’s Leveäniemi open pit mine in the north of Sweden. The experimental area is an apatite-magnetite iron ore deposit of the Kiruna type, with the demand to exhibit a maximum of local heterogeneity to allow the experiment the largest possible conclusion powers. Gathering and splitting the complete BH drill cutting cones (fines cone, coarse cone) served as an authoritative reference on which to assess the representativity of the two BH sampling methods. Results show that manual BH drill sampling is able to produce samples with an acceptable accuracy and precision if performed properly and with insight. Based on variographic comparison, segment sampling shows similar precision for chemical grades as the manual sampling, while the particle size distributions were more deviating for segment sampling than for manual sampling at the high and low size ranges. Possible reasons for these partly surprising results are discussed. For further investigation of the performance of BH versus RC drill sampling, a second experiment has been planned on a parallel profile. This will result in an unprecedented opportunity for a full variographic comparison between RC and BH sampling, aimed at designing specific, fit-for-purpose sampling procedures as a function of ore quality, cut-off grades and important contaminant levels.