Batch tests in a 0.6m diameter mill were used to simulate fully autogenous
primary and secondary grinding. Substantial improvements in primary mill ca
pacity were obtained by modifying the mill contents to simulate crushing of
small pebbles and recycle of the crushed material. Secondary grinding effi
ciency(as measured by kWh per ton finer than 75 microns or Work index), was
improved substantially by reducing the pebble size to 20/28mm. The process
as a whole was optimised when crushed rock coarser than 4 mm was recycled
to primary mill, as the small pebbles were most efficient with "sand". The
benefit of using small grinding media has been known for some time, but the
high cost and rapid consumption of small steel balls makes their use uneco
nomical. However a plentiful supply of small pebbles is available from a pr
imary mill. A further and unexpected improvement in power efficiency was ob
tained when pebble milling was performed at 92% of critical speed. (C) 2000
Elsevier Science Ltd. All rights reserved.