Regrind Mills: Scaleup Or Screwup

As ore deposits become finer grained the requirements for regrinding before cleaning or leaching have increased substantially. Despite this increasing need, there is no standard test to predict grinding energy requirement below 70 microns. The standard Bond Mill test applies for coarser ball milling, but is not appropriate for stirred milling with fine media grinding to 70 microns. With no industry -standard test, the energy requirement is often made on the basis of supplier estimates or benchmarking against similar applications. Yet suppliers use vastly different scale-up methods which result in widely different energy estimates. Estimates can differ by 100-300% even for similar mills. This must result in serious errors to install either too much or too little grinding power. This paper explores this by comparing the actual performance of full scale regrind mills against their original design estimate. It confirms that serious scale-up errors have been made, and have then been perpetuated by ?benchmarking?. The test conditions to achieve accurate power estimation are discussed. The essential test conditions are: continuous (not batch) tests, ensure steady state (no coarse fraction retained in test mill), correctly account for classification, measure energy directly (not inferred), and using the same media size as the full scale installation. Failure to correctly address even one of these factors can bias results by 40%; failure to address several multiplies the error.