Projecting Laboratory and Test Data into Operating Reality

"Laboratory and pilot-scale tests are almost invariably required before metallurgical development can proceed. Difficulties with sampling, as well as with classical chemical and physical scaleups, are normally anticipated and relatively well-understood, but more subjective factors also come into play, including misunderstandings of the operating scheme chosen, wrong ""lessons"" learned from similar operations, failures to handle data appropriately, and even arbitrary intervention by management. The paper examines some of these subjective factors and suggests ways to guard against them. Several anecdotal case studies are used as examples, including when and where to heap leach or to mill specific types of gold ore.When the Romans first exploited the gold-rich gossans at Rio Tinto, they were guided by analyses and tests that delineated the ore deposits, identified the richer portions, and measured the gold recoveries achieved. There is little evidence, however, to suggest that test data were used to develop improved methods of metallurgical extraction. Indeed, it is likely that ores too difficult to treat were simply passed by, at least until improvements in full-scale operations could deal with them. For example, it is not difficult to imagine that an early metallurgist might test a tin ore by placing some in a fire; a smallscale emulation of the method actually used in practice."