Does a SAG Mill reduce steel consumption? Do SAG mills cost less to operate compared to conventional rod or ball mill grinding circuits? Mining companies are always searching for new ways to cut costs. In order to be competitive and justify the tremendous capital costs of new projects, there has been a strong trend toward mills with a capacity of 36,300 or more tonnes per day (40,000 stpd). Since autogenous or semiautogenous circuits have high unit capacity and could thus result in lower capital costs and in lower operating costs, many companies have installed these types of circuits. Indeed, it is obvious that large capacity units of any kind should be carefully investigated in the early studies preceding a decision for any mining venture.

Cyprus Pima constructed a 2700 tonne per day (3000 stpd) mill in 1956 and expanded it three times until by 1968 daily capacity exceeded 34,500 tonnes (38,000 tons). This No. 1 mill had conventional three stage crushing followed by rod mills and ball mills. When a fourth expansion was considered in 1970-71, extensive pilot plant tests indicated that primary crushing, followed by semiautogenous grinding and ball mills had advantages in both capital and operating costs. The No. 2 mill, put on stream late in 1971, had a design capacity of 12,700 tonnes (14,000 stpd).

Identical ore was fed to both concentrators so an excellent comparison can be made of the steel consumed by each reduction method. Sufficient cleaning and regrinding capacity was available in Mill I so that it was necessary to produce only a rougher flotation concentrate in Mill II. However, excellent comparisons can be made between all unit operations through rougher flotation.

Table No. 8 summarises the liner and media consumption for both mills. Total liner consumption in Mill I was 0.0953 vs 0.0898 kg/tonne (0.1906 vs 0.1796 lbs/ton) in Mill II. Total media consumption in Mill I was 0.8635 vs 0.7540 kg/tonne (1.727 vs 1.508 lbs/ton) in Mill II. Mill II thus shows a 6% lower liner consumption and a 15% lower media consumption.

Identical ore was treated in both mills. Mine trucks delivered ore to whichever primary crusher needed feed. No distinction was made in feeding ore to either crusher except when exceptionally wet or sticky ore was encountered – that was fed to No. II crusher whenever possible to avoid crushing plant problems associated with such ore.

Final grind was maintained at 20% plus 65 mesh in both mills and was generally within plus or minus 2% of the goal.

The 6% better liner consumption shown in Mill II is small but it is significant. As time goes on and improved alloys and designs are developed for specific semiautogenous liners, it is believed that liner consumption can be improved. It is unlikely that much more improvement can be made to liner consumption in Mill I because the extensive liner testing done in the late 1960’s and early 1970’s had already shown a significant improvement by 1974.

The 15% lower media consumption in Mill II is certainly significant because media costs usually are in the range of 15-20% of total direct mill costs.

The results at Cyprus Pima over the four year period from 1974 to 1977 show that total steel, consumption was 13% less using a semiautogenous-ball mill circuit versus a conventional crushing-rod mill-ball mill circuit.