Optimized Multivariable Control Of An Industrial Run-Of-Mine Milling Circuit

This paper describes the application of a throughput-optimization strategy to a run-of-mine (ROM) mining circuit that is controlled by a multivariable controller. The objective of the optimization strategy is to maximize the ROM throughput at a specific product size. Results from long-term tests, which indicated how 'steady-state' optimization can be achieved, were used for the throughput optimization. This strategy increased the ROM throughput by 11,4 per cent. A mathematical relationship relating residue to grind can be used to calculate the economic benefits of stabilizing control. It was estimated that the throughput optimizer could result in an increase in gold production worth Rl,4 million per annum when waste rock is treated, and that stabilizing control could increase the treatment plants annual revenue by R150 000. If the extra milling capacity brought about by the optimizer is interpreted as a capital cost saving of 10 per cent on the price of a new mil, then a capital cost saving of RI million could result.