Closing the loop - Using actual concentrator performance to determine the true value of ore sources

CIM Bulletin, Vol. 98, No. 1086, 2005

A. Dance

The Antamina orebody is a geologically complex, polymetallic skarn located in the Peruvian Andes. Different copper and copper-zinc ores are campaigned separately through the concentrator producing copper and zinc concentrates as well as lead and molybdenum by-products. Ore type classification at Antamina is based on chemical assays; geological or lithological variations are not considered. Consequently, each ore type demonstrates a wide range of concentrator performance. Antamina is currently implementing a system to track ore movements from the shovel face to the concentrator. Sophisticated models of the run-of-mine and coarse ore stockpiles ahead of the mill will be used to account for the variable delay between mining and milling. One of the numerous benefits of ore tracking is the ability to directly associate the measured concentrator performance with a physical location in the orebody. Using existing mine planning software, three-dimensional models can be created of such performance measures as metal recovery, concentrate quality, operating cost, and reagent consumption. Such a database of information allows the true value of different ore sources to be determined. The value of an ore source is conventionally measured as metal content (individual or combined ‘equivalent’) with a cut-off grade applied to indicate some base level operating cost. However, the true value of an ore source should include dynamic factors like haulage cost, reagent use, smelter penalties, and even prevailing metal prices. If an ore source is considered from all perspectives rather than simply its potential to deliver value, the concept of ore becomes more flexible. Unprofitable ore sources can be eliminated from the mine plan or blended to improve their value. The ability to predict mill performance, and thereby profitability, is a powerful tool. In 2003, ore sources could only be resolved as polygons (with average grades) in the dispatching system. This was dramatically improved in 2004 with the high-precision GPS location of the shovel bucket being superimposed upon the 4 m by 4 m geological block model and individual blasthole assays used to calculate the grade of each bucket load. Similarly, block model x, y, and z coordinates are now tracked with every load. Therefore, not only is the ore source polygon identified, but the exact location within the polygon (with ± metre accuracy) being recorded and tracked with the material properties. Through the use of ore tracking, Antamina has recently identified areas within the copper-zinc ores that have resulted in lower mill throughput. In fact, mill throughput is now routinely predicted based on production drill performance parameters: an equation developed by tracking mill performance back to the ore polygons and its individual blastholes. These zones of different throughput correlate very well with mineralogy and are quite predictable with depth from the bench above. For a longer term prediction of mill throughput, point load index values from resource definition drill core have been related to production drill parameters, and hence mill tonnage. In an operation such as Antamina, where managing ore sources, campaigns, and the delivery of a variety of concentrates in a narrow timeline to customers is a daily challenge, prediction is the key to success and improved efficiency.