The design of new gold plants with flotation and/or cyanidation processes tends to include Semi Autogenous Grinding (SAG). Basically, SAG milling eliminates secondary and tertiary crushing that leads to simplicity of plant design and construction, lower operating and maintenance staff requirements and ease of control. Also, throughput can be manipulated within limits by adding more balls. Then, design of the grinding circuit has an impact on the treatment plant economics. For example, when planning a new gold processing plant, Harmony Gold Mines estimated a 25% capital advantage for SAG milling over crushing and milling and operating costs were estimated to be about 28-32% less. SAG milling is not suitable for all ores and a special comminution testing program and some throughput simulation is necessary before taken the final decision.
Basically, a SAG circuit is effective when the ore feed has a suitable combination of hardness and size range. Variations in these parameters control build-up of critical size material in the mill and have a major effect on throughput. The future will see greater use of variable speed drives on SAG mills to compensate for some of these variations. Then, the additional capital cost can be recouped in a few years through increased gold recoveries and reduced media and liner costs. A number of options have been developed for dealing with critical size material. In South Africa where the gold-bearing reef is more friable than foot and hanging wall quartzites, some operations has tested sorting of pebbles extracted from the mill. Up to 80% of the extracted pebbles proved to be barren and could reject resulting in increased milling capacity and specific power usage. For the tests, pebbles were hand sorted, but for commercial applications a photometric or radiometric sorter may be employed. A third alternative is to crush the critical size material and return it either to the SAG mill feed or to secondary mills. The last approach is the most widely adopted and typically increases throughput by 20-30%.
A very important cost is related to the liners. Basically there are two options, one is to consider a SAG mill with reversible rotation and the other is to adopt a special liner design. In the first case, when the SAG mill has been operating in clockwise direction for one month, it is possible to change the change the rotation direction in order to extend the life of liners. In the second case, the plant must experiment the change by using new liner designs. For example, steel plates bonded to the leading edges of rubber lifter bars, thus combining the best properties of steel and rubber. Tests indicate an increase in liner life of about 40-50% reducing liner costs and improving mill availability. Grinding and classification are interrelated. Although there has been more experimentation with screens in grinding circuits, their capital and maintenance costs in comparison with hydrocyclones seem likely to restrict them to specific applications. Hydrocyclones separating efficiencies have improved markedly, especially in the finer size fractions (less than 38 microns) and these advances are now being extended to the coarser size ranges.