Abstract:
Alumina feeding and its subsequent dissolution in molten electrolyte plays a major role in the operation of aluminium reduction cells. In the ideal feeding situation, all the alumina powder added disperses immediately in the electrolyte with which it comes into contact and dissolves rapidly without clumping or excess agglomeration occurring. However in practice alumina clumping/agglomeration and sludge formation occurs, which causes slow and unpredictable alumina dissolution. In recent years several technological advances (ie. reduced flow velocities through magnetic compensation, smaller superheat and lower melting point electrolytes) in the aluminium process have made the formation of 'sludge' or 'muck' much easier. This is because these advances have either reduced the saturation solubility of the alumina and/or lowered the rate at which the alumina can be dissolved in the molten electrolyte. Sludge formation and accumulation is undesirable because it causes disturbances in the metal pad and the alumina present in the sludge is not readily available for electrolysis causing losses in process efficiency. Sludge also causes erosion of the cathode and an increase in cell resistance. In particular the type of feeding system used (breaker beam or point feeders) by a smelter has a significant influence on the amount of sludge formed, due to the size of alumina addition and also to the amount of crust broken into the cell. When large amounts of alumina and crust are introduced into a reduction cell, the electrolyte cannot supply sufficient heat (energy) to dissolve the alumina and therefore sludge formation occurs.
In an attempt to alleviate the thermal constraints to dissolution and therefore reduce sludge formation in cells, a new feeding system which continuously feeds alumina at adjustable flow rates was developed. The adjustable flow rates are achieved by controlling the cross-sectional area of an orifice using a tapered cone. Laboratory trials and off-line plant trials were conducted and these studies established design data for the effect of alumina head, height of the feed windows, alumina quality, orifice size, and cone/orifice clearance and also demonstrated the continuous feeder's mechanical reliability. Alumina quality was identified as the most significant factor influencing the feeder performance. However other factors which affect alumina flowability such as electrostatic interactions and moisture content were not observed to significantly affect the feeder flow rates in the off-line trials. A semi-empirical model was developed and the flow rates from the continuous feeder unit using smelter grade alumina with circular orifices was given by: N 1.4d02-di)t2 = 0.33 p Bgli2 ((/ - 1.44)2 - - 1.44 + d.
The equation was developed following the analyses of other workers who found the gravity flow of materials through horizontal circular orifices should be proportional to orifice diameter to the power of 2.5. The equation was compared to models developed by other workers and it was found to best fit the experimental data from the off-line trials. A mass flow control system was developed in conjunction with the continuous feeder unit. The mass flow concept was developed in an attempt to overcome the problems associated with volumetric alumina meters, which rely on uniform alumina properties for reproducible additions of alumina. However the mass flow calibration curve (voltage signal versus flow rate) was sensitive to tray incline angle and alumina quality. Therefore the mass flow system was not independent of alumina properties and suffered from similar problems as volumetric meters. A laboratory study of the maximum dissolution rates for continuously fed alumina under various operating conditions were conducted. Increasing the initial alumina concentration of the melt decreased the maximum dissolution rate. While increasing the stirrer speed was observed to significantly increase the maximum dissolution rate achievable. The experimental technique allowed a transient heat balance to be performed, allowing the heat of solution for alumina at different alumina concentrations to be determined. The values determined for the heat of solution for alumina were consistent with those found by other workers.