Iron Ore Grade Control for Flitch Mining – An Alternative Method for the Hope Downs Operation, Hamersley Province, Western Australia

Grade control systems within Rio Tinto Iron Ore have proven to be robust, when mining full benches. The Hope Downs 1 operation hosting two Marra Mamba Iron Formation deposits (North and South Deposits) presents a new challenge, where increased mining selectivity due to moderately plunging N-verging folds has made flitch mining necessary in the South Deposit. Poor flitch blasthole sample duplicate results in South Deposit over a 12 month period, compared to North Deposit, have highlighted concerns with grade predictions. In contrast a new grade control methodology (grade splitting) was proposed to alleviate blasthole sampling bias, utilising geological mapping and mineralogical logging, currently practised throughout Rio Tinto Iron Ore Pilbara operations, to estimate theoretical grades for each flitch from a single blast cone assay. To assess this method, reverse circulation holes (RC) were twinned to blastholes establishing an ‘actual’ grade. These blastholes were assayed using current flitch sampling techniques where: flitch positions were estimated, flitch positions were marked with paint, and a full cone sample was taken for ‘grade splitting’. Using these data sets multiple grade predictions were produced, then analysed using Half Absolute Relative Difference (HARD) statistical analysis. In addition to testing these grade control methodologies two other findings were ascertained: an inherent bias in blasthole assays exists particularly in the top flitch, where an average absolute difference of 0.6 per cent in SiO2 and 0.42 per cent Al2O3 was measured; and assays derived where kaolin rich soft shales are intersected were less representative. This is likely to be due to these materials not being properly represented in blast cones. Grade splitting accurately estimated block chemistry, except for a slight under-prediction in Al2O3 and P compared to current grade control methodologies. This method is a viable alternative where grade distribution is stratigraphically controlled.