GAS DISPERSION MANAGEMENT IN A COPPER / MOLYBDENUM SEPARATION CIRCUIT

摘要

The effect of superficial gas rate distribution (J_g profile) on down-the-bank metallurgical response was tested in the rougher-scavenger bank (eight 10-m~3 tank cells, 4 roughers and 4 scavengers) at Highland Valley Copper Mine's Cu/Mo separation circuit. Three J_g profiles were tested: (a) as found (i.e., typical practice), (b) increasing-decreasing (i.e., increasing from cell 1 to 4 and decreasing from cell 5 to 8), and (c) decreasing-increasing (i.e., decreasing from cell 1 to 4 and increasing from cell 5 to 8). The J_g profile approach proved to be of considerable value. Profile (b) gave the best metallurgical results, i.e., the highest down-the-bank grade/recovery (increase in grade at constant recovery was approx. 3-4 percent compared to the as found case) and the most efficient Cu/Mo separation (cumulative Cu recovery was reduced by almost half -approx. from 20 percent to 12 percent, at 99 percent Mo recovery). The success of the evaluation prompted the implementation of profile (b). Detailed gas dispersion characterization tests (J_g and D_b (bubble size)) showed significant discrepancies between the J_g measured directly (using the McGill University sensor) and the J_g calculated from field instrument readings (volumetric gas flowmeter) and machine dimensions (cell cross-sectional area). Tests were designed to reveal the source of disagreement. It was observed that a significant number of bubbles were not able to float (i.e., they were entrained with the tailings), the evidence appearing to be a significant number of very small bubbles (approx. 200-300 mu m) which did not have sufficient buoyancy to rise. Consequently, rather than relying on the field instrument, the evaluation depended on the McGill University J_g sensor, which collects bubbles in a tube near the pulp-froth interface, thus providing significance to the data as the measurements were obtained from bubbles that almost certainly floated.