Ore Sorting helps the economics of mining when ore becomes diluted by country rock because of narrow slopes or because of weak wall rock, or when it naturally contains barren material which is distinguishable from the ore, sorting and rejecting the worthless rock may be economical. This is done in a casual manner at some small mines and on a large scale at others, as was the case at Alaska-Juneau (40 per cent rejected), or covering whole districts, as at Kolar, India (10 per cent), and on the Rand (9 per cent). Sorting at a gold or silver mine may mean its existence, as at the Alaska-Juneau, or it may result in lower costs because less ore is crushed and treated to recover the same amount of gold, as at Cripple Creek, at Kolar, and on the Rand.
SORTING ORE IN THE FLOWSHEET
The place at which ore sorting will be done is mainly dependent upon the size of the mine-run ore. If the ore is not in too big chunks, the sorting belt may be placed below the grizzly or trommel to receive the oversize after it has been sprayed with water. The grizzly undersize is transported direct to the storage bin. If mine-run ore is in too large pieces, it should be broken to 4- to 7-in. size, sprayed, and then fed to the sorting belt. The wash water containing the fine material sometimes carries enough gold to be worth saving; therefore provision should be made for sampling and assaying it also for its proper disposal.
Ore Sorting on the Rand
The sorting of waste (low-grade) rock has been accepted practice for many years on the Witwatersrand. Low labor cost and the peculiar structure of the “banket” ore are both contributing factors. In some plants tube-mill pebbles as well as reject rock are removed from the belts and delivered to separate bins, while other plants merely pass the rock after sorting to grizzlies or trommel screens, where 6-in. oversize is separated out as grinding media and the undersize passes to the secondary crushers.
In his presidential address before the Chemical, Metallurgical and Mining Society of South Africa, A. Clemes states:
There has been little change in the last 10 to 15 years in the established method of removing waste rock from slow-moving conveyor belts feeding the primary (and possibly secondary) fine crushers. On most plants ore passing through jaw breakers set at, say, 5 in. and retained on grizzlies (or screens) set at between 2 and 3½ in. is subjected to waste sorting. Where intensive sorting is deemed necessary (say 15 per cent of the ore delivered), it is usual to arrange for two sets of sorting belts, one carrying minus 5 in. plus 3 in. material and the other minus 3 in. plus 2 in. The admixture of these sizes is, obviously, detrimental to intensive sorting, but where sorting of waste rock is not important, the capital cost of installing additional screening and belt conveyor equipment may be uncalled for.
“Time studies” of the efficiencies of native sorters have been carried out on most mines with some measure of individual improvement, but efforts in this direction are largely nullified by the fact that often only poor quality or transient labor is allocated to this work, whilst the sortability of ore varies widely. In general, it may be accepted that under normal conditions natives sorting coarse waste can average about 10 tons per 8-hr. shift, falling to well below 5 tons per shift under conditions of intensive sorting of the smaller sizes.
Purely from the angle presented to the metallurgist, benefication of the ore to the mill still seems profitable. Assuming a total cost of metallurgical treatment of 2s 9d per ton with a residue value 1s 9d (0.20 dwt.), it appears worth while to discard waste rock valued at 3s (0.35 dwt.) at a cost for sorting and dumping of about 1s 2d per ton; equivalent to a gain of 4d per ton.
Such a simple arithmetical justification of waste sorting is not, however, completely acceptable today. Native labor is in short supply, and a number of reduction plants are working below capacity; possibly, more mines could now profitably take in waste rock yielding, say, 80 per cent extraction of 0.35 dwt. or 2s 4d per ton. Thus, under present and forecast conditions of shortage of man power, one is inclined to wonder if the continuance of waste sorting on some producing mines can be justified, also if the capital cost of providing for waste sorting—conveyor, belts, waste bins, waste-disposal equipment, etc.—should be incurred in new plants.
Ore Sorting at Randfontein
The ore is first screened on seven 8- by 3- ft. Tyrock double-deck screens, with 3-in. round hole openings on the top deck and 1½-in. square mesh screen on the bottom deck. The undersize of these screens goes directly to the mill bins, and the oversize to washing and sorting.
The +3-in. oversize of the top deck, which ranges in size up to 14 in., is washed and sorted on five 36-in. wide by 118-ft. long belts, from which waste and primary tube-mill pebbles are sorted. The oversize of the lower deck passes to two similar belts from which waste and secondary tube- mill pebbles are sorted after washing.
The washing is done by sprays on the lower end of the sorting belts, using 525 gal. per min., and drainage from the belts or washing fines are dewatered in two simplex Dorr classifiers with rake product going to the mill bins and overflow 84 per cent minus 200 thickened in two intermittent settling tanks, from which the thickened pulp is pumped to the secondary grinding circuit classifiers.
The sorting belts run at 25 ft. per min., and on an average 40 native boys pick about 60 tons per hour of waste and tube-mill pebbles. In 1946 the waste amounted to 2.05 per cent of the crude ore and had an average value of 0.152 dwt. per ton. The cost was 13.06d (21.8 cents) per ton of waste.
Ore Sorting at McKenzie Red Lake, Canada.
The crushing plant will handle the mill tonnage (225 tons per day) in 8 hr. The ore from the 200-ton mine ore bin is fed by a swing-hammer feed regulator to a 30-in. link-belt feeder delivering to a 2-in. bar grizzly at an angle of 35 deg. The fines drop to an 18-in. conveyor belt, and the oversize to a 36-in. sorting belt, where it is water washed by sprays prior to sorting, which is done under fluorescent light. The average production per day in 1945 was 209 tons, of which 35 tons was sorted to waste. The total cost per ton of ore mined and milled was $7,656.
THE ECONOMICS OF Ore SORTING & Pre-Concentration
The question as to whether in any given instance sorting is justified is strictly a matter of economics. The cost of installation and operation of a sorting plant and the inevitable loss of some gold value, however small, in the rock discarded must be balanced against the saving in milling cost resulting from the elimination of low-grade rock. In this connection, R. D. Lord in “Milling at Preston East Dome,” C.M.J., August, 1941, gives the following formula:
Tons milled = A
Tons sorted out = B
Tons mined = A + B
Cost of mining per ton mined — C dollars
Cost of milling per ton milled = D dollars
Cost of sorting per ton sorted — E dollars
Value of waste as backfill = F dollars
Grade mined = m
Grade sorted = p
Grade milled = m(A + B) – pB/A
Recovery from mined ore = x (as a decimal fraction)
Recovery possible from sorted low grade = y (as a decimal fraction)
Recovery from milled ore² = xm(A + B) – ypB/m(A + B) – pB
To obtain the above formula we have that the gold recoverable from the total ore hoisted is mx(A + B) and that the gold recoverable from the material sorted out is ypB.
Then for the ore reaching the mill the gold recoverable is xm(A + B) — ypB out of the total gold reaching the mill which is m(A + B) — pB. This makes the fraction recovered expressible as
mx(A + B) – ypB/m(A + B) – pB
Operating cost = A(C + D)
Recovery = xmA dollars
Cost in dollars per dollar in gold recovered = A(C + D)/xmA = C + D/xm
Operating cost of sorting = BE in dollars
Cost of picking per ton mined = BE/A + B
Cost of milling = AD
Total cost = C(A + D) + BE + AD + (ypB – BD)
Recovery = xm(A + B) – ypB/m(A + B) – pB x m(A + B) – pB/A x A = xm(A + B) – ypB
Cost in dollars per dollar gold recovered
= C(A + B) + BE + AD + (ypB – BD)/xm(A + B) – ypB
Then the maximum value of material that can be discarded without increasing the cost per ounce of production occurs when
C + D/xm = CA + CB + BE + AD – (ypB – BD)/xm(A + B) – ypB
This equation reduces to
p = xm(2D – E)/y(C + D + xm) X xn 2D – (E – F)/y(C + D + xm)
when introducing F, the value of the waste as backfill.
ORE SORTING BY SINK-FLOAT for Pre-Concentration
As yet not applied to the sorting of gold ores, this method is extensively used for the elimination of low-grade material in the treatment of coal and various “metallic” and “nonmetallic” ores. Based simply upon the principle of “floating” the waste in a heavy media suspension having a density intermediate between that of the desired “sink” and rejected “float,” the method presupposes that an appreciable difference exists (at least 0.2) between the specific gravity of the ore constituents. Ordinarily as applied today, the method will handle minus 2-in. plus 3/16-in. feed (occasionally as fine as 10 mesh), but improved methods are under development which may soon make it economically possible to handle material as fine as 100 mesh.
The principal reason why gold ores have not thus far been handled by this method lies no doubt in the fact that it is rare to find coarsely mineralized gold ores in which the values are highly concentrated in a heavy fraction after crushing to the size range above indicated.