Discussion - Vacuum Filtration: Available Equipment and Recent Innovations – Technical Papers, MINING ENGINEERING, Vol. 31, No. 10, October 1979, pp. 1479-1486 – Moos, S. M. and Dugger, R. E.

In the paper the authors show [(Fig. 12)] in diagramatic form a "typical vacuum filtration system." This indicates a drum filter with two filtrate receivers-one for mother liquor and one for wash liquor. Have the authors ever found a drum filter that will adequately separate mother liquor from wash liquor? I recall in about 1957 being superintendent of a major copper leaching operation treating oxide flotation concentrates. The plan was being expanded, and filters installed to replace CCD following acid leaching and primary thickening of the pulp. The design of the expansion allowed the mother liquor to join the thickener overflow as feed for the electrolytic tankhouse. The water balance required that only pan of the wash liquor could be electrolysed direct, and the remainder had to be treated by lime precipitation. The design assumed the mother liquor would have the same copper content (about 50 gpl) as in the leach pulp solution, and the wash liquor about 50% of this (25 gpl from memory). You can imagine my horror when, on starting up, the solution from the mother liquor receiver contained somewhat under 40 gpl and, from the wash liquor receiver, somewhat over 40 gpl. The filters were provided by two of the most reputable manufacturers and, in the end I believe they agreed that never before had they supplied rotary drum filters that would segregate the filtrates. By changes to the valve design we made minor improvements, but eventually decided that insufficient segregation of the filtrates was possible to warrant trying to do it. On further consideration I concluded that, owing to the time delay between solution passing through the cake and entering the valve, segregation was impossible, though radical alteration of the internal piping of the filter might have improved it. I would be interested to know if anyone has found a solution to this problem using a drum filter. If not, why go on publishing information that infers that it is possible?