Slurry Rheology Influence on the Performanceof Mineral/Coal Grinding Circuts Part 2

Part 2 of this article continues the discussion of a 10-year, multimillion dollar research and plant testing program on slurry rheology and grinding circuits. The first part of the article (ME, Dec. 1982) presented the basic concepts identified by the research and some laboratory test results. This section Illustrates typical Industrial scale test results and Identifies some industrial operating implications of controlling rheology by different methods. At least four controllable factors decide the rheological character of a slurry-slurry density or percent solids, particle size distribution, chemical environment, and slurry temperature. The second factor has two interrelated facets: the shape of the particle size distribution which controls packing behavior of the solids, and the fineness of the distribution. Finer particles increase interparticle forces and viscosity. As indicated in Part 1, during a given grinding test it is possible for all four factors to change. However, regardless of the particular settings of the four factors in a given test, if the resulting rheological character is either dilatant, pseudoplastic, or pseudoplastic with yield, the associated breakage rate is correlated with the current rheological character. It is obvious, for example, in batch grinding tests run at constant percent solids, that the second and third factors, where appropriate, are changing during grinding because size distributions are changing and the production of fresh surface area takes up unadsorbed chemical. Thus the corresponding rheological character change in batch tests with increasing grind time would be dilatant to pseudoplastic to pseudoplastic with yield. The degree to which this transformation occurs depends on the changing setting of the four factors over grinding time. In continuous grinding tests, one might logically expect that dramatic changes in any one of the four factors will be less likely to occur. It will be shown that continuous mill operations offer some unique opportunities to take advantage of possible rheological transformation by more direct operational control of the settings for the four factors. One extra observation noted in the rheological studies was the variability in the location and extent of region B (pseudoplastic behavior) for the various coals and ores tested. The location of region B was usually in the region of 45-55% solids by volume and was of the extent of 0-8%, or 2-11% with chemical addition. The corresponding increase in net production ranged from 0-10% in region B and 0-21% in region B&apos;. When region B is small or zero (no pseudoplastic character is exhibited) no increase in production will be observed and the use of chemicals is often marginal. There are several reasons for some materials exhibiting this quick transformation from dilatancy to high yield values often at surprisingly low percent solids by volume such as 30%. One condition identified was for materials containing high levels of viscosity-producing elements such as- carbonates or clays. A second condition documented was for materials that exhibited unusually fine primary fragment distributions (fine B1,; curves). This corresponds to materials having small y values of < 0.5. Such a slurry developed a yield value quickly during grinding because of the rapid buildup of fines. A related problem can be presented by materials exhibiting excessively coarse primary fragment distributions, y values > 0.9. Breakage of this type of material produces size distributions that give poor packing efficiencies even at long grind times, thus hindering the normal rheological transformation presented earlier. A final condition that can cause the occurrance of region B to be small or zero is when the media void volume filling of slurry is < 100%. Figures 5a and 5b use previously published data of this study to demonstrate the influence of solids loading and weight percent solids on the net production of taconite ore in a laboratory batch ball mill. In particular, Fig. 5a illustrates several trends not generally recognized until this study. They include these two: • Increased slurry density allows for increased solids loading before passing through the maximum in the net production curve where the fall-off is due to non first order breakage. • The use of rheology control chemicals such as GA-4272 allows this trend to be extended to higher slurry loadings with an increase in net production over any previous condition by keeping grinding first order. Figure 5b shows the same data as Fig. 5a plotted for constant weight loadings. These various figures of net production versus percent solids show the location and extent of the regions A, B, and C presented in Part 1 as a function of solids loading. It is obvious that the rheological transformation pattern described earlier does not hold true for slurry loadings corresponding to less than the void volume of the media, which is also a region of