1997 TMS Annual Meeting: Wednesday Session

ALUMINIUM REDUCTION TECHNOLOGY: Session VI: Cell Studies

Session Chairperson: Elmar Sturm, Hamburger Aluminiumwerk GmbH, P.O. Box 950165, D-21129 Hamburg, Germany

2:00 pm

ON THE BATH FLOW, ALUMINA DISTRIBUTION AND ANODE GAS RELEASE IN ALUMINIUM CELLS: Ove Kobbeltvedt, Department of Electrochemistry, Norwegian University of Science and Technology, N-7034 Trondheim, Norway; Bjørn P. Moxnes, Hydro Aluminium, Technology Centre Årdal, N-5870 Øvre Årdal, Norway

The bath flow was studied in prebake cells. The measurements were performed by recording the drop in bath temperature subsequent to alumina feeding as well as by measuring the wear of quartz rods immersed in the bath. The horizontal flow rate varied between 3 and 20 cm/s. It was found that the horizontal bath flow and consequently the alumina distribution pattern are determined mainly by a combined effect of the magnetic fields and the width of the channels above the working surface of the anodes. The quantity of anode gas which was drained into the centre channel was measured at different locations in a cell. The result showed that the distribution of the anode gas in the bath is strongly associated with the magnetic fields.

2:25 pm

ALUMINA DISTRIBUTION IN POINT-FED HALL-HEROULT CELLS: K. Tørklep, K. Kalgraf, T. Nordbø, Elkem a/s Research, P.O. Box 8040 Vaagsbygd, N-4602 Kristiansand, Norway

We describe the alumina concentration distribution in point-fed Hall cells in terms of finite physical elements. The need arose in the development of the New Søderberg technology at Elkem Aluminium ANS, but the technique is applied to prebake pots as well. The concentration at any point in the bath is computed from the dynamical equilibrium between consumption and transport of alumina (directly or indirectly dissolved) and presented as contour plots. Necessary inputs are the velocity field in the bath and the turbulent diffusivity, either measured or calculated. We prefer to measure both by injecting a small amount of a molten radioactive tracer and follow the distribution of this tracer in situ. Measured velocity fields have been found to vary significantly between presumed identical pots and over time. This is mainly ascribed to convective coupling with the metal, where the boundary conditions and hence the driving forces may well differ between pots and change with time. Contrary to extant models, our approach permits experimental verification through comparison of predicted and measured oxide concentrations at various locations in the pot under investigation.

2:50 pm

METAL PAD WAVE ANALYSIS USING FAST ANODE LOWERING METHOD: H. Q. Tang and N. Urata, Kaiser Aluminum & Chemical Corporation, Center for Technology, P.O. Box 877, Pleasanton, CA 94566

In order to study the metal pad wave in aluminum reduction cells, the two step fast anode lowering method was developed and applied to the operating cells. The metal pad waves after anode lowering were measured and analyzed. The period and shapes of the metal pad waves were determined by performing Fourier transform analysis of the measured data. The MHD wave equations were solved for cells of different magnetic fields to simulate the metal pad waves caused by this fast anode lowering. The period and shapes of the predicted steady state waves were in good agreement with the measured waves, although the measured wave had more transient nature. The cell stability was evaluated based on the metal pad wave analysis.

3:15 pm BREAK

3:35 pm

IMPURITY TRANSPORT MECHANISMS IN ALUMINIUM REDUCTION CELLS: M. Webster, Xiaoling Liu, Comalco Research Centre, Thomastown, Victoria, Australia; J. Metson, Department of Chemistry, University of Auckland, New Zealand

The purity of aluminum from pre-bake reduction cells is affected by the quantity of impurities introduced with the raw materials and the fraction of these reporting to the metal. The transport of Ti, V, Ga, Si, Fe and Ni from the reduction cell to the duct emissions stream has been studied for point feeder and bar break cells. Samples of material from bath, cell cover and duct emissions have been examined and analysed for a range of impurities. For elements with volatile, bath generated fluoride compounds, the impurity content of the cover and duct samples is proportional to the carbon content. Carbon grains with thick (>10 micron) impurity rich surface coatings are observed in both the loose cover samples and particles from bath skimmings. Thus a possible mechanism for the transport of these impurities may be postulated.

4:00 pm

PSEUDO RESISTANCE CURVES FOR ALUMINIUM CELL CONTROL--ALUMINA DISSOLUTION AND CELL DYNAMICS: Halvor Kvande, Hydro Aluminium a.s, P.O. Box 80, N-1321 Stabekk, Norway; Bjørn P. Moxnes, Jørn Skaar, Per A. Solli, Hydro AIuminium a.s, Technology Centre Årdal, P.O. Box 303, N-5870 Øvre Årdal, Norway

The pseudo resistance was measured as a function of the alumina concentration in the bath in five different types of cells. The pseudo resistance showed a minimum value of 5.0 to 5.5 mass% Al2O3, when the bath samples were analyzed by the LECO method. To the left of the minimum point on the curve the slope increased gradually until the anode effect occurred at 1.6 to 2.2 mass% Al2O3, while a nearly linear curve was found in some of the measurements. The difference in pseudo resistance determined just prior to the anode effect and at the minimum point, corresponded to a voltage difference between 100 and 300 mV. Immediately after the alumina feeding rate was reduced from overfeeding to underfeeding, a so-called "hysteresis effect" could be observed. This was characterized by a sudden decrease in cell voltage of about 100 mV in less than 30 minutes, in spite of practically constant bath composition and temperature in this time period. This effect may be caused by dissolution of alumina sludge in the bath phase above the metal pad, accumulated during the long overfeeding period of several hours, needed to reach concentrations to the right of the minimum point on the curve.

4:25 pm

THE INFLUENCE OF SODIUM ON THE ALUMINIUM REDUCTION CELLS: Mohamed O. Ibrahiem, Mohamed M. Ali, R&D Department, Aluminium Company of Egypt, Naga Hamadi, Egypt

Sodium always present in Hall-Heroult cells, has a decisive influence on cell performance and pot failure. In this paper, sodium was studied from three points of view. The first is the sodium mass balance. Alumina and cryolite are the major sources of sodium inputs to the electrolytic process, where 56.8% and 40.7% of sodium come from them, respectively. The second point is the sodium content in aluminium metal. Measured sodium content in aluminium of 203 kA prebaked cells is lower than that of 155 kA Søderberg due to lower bath ratio, higher excess aluminium fluoride, and higher magnetic fields. The third point is the study of sodium concentrations in failed carbon cathodes at different ages. This concentration at 3194 days is 8.5%, 7%, and 1% higher than that at 40, 60, and 1369 days, respectively.

4:50 pm

STUDY ON CATHODIC PROCESS OF Na+ IN THREE-LAYER ELECTROLYTIC REFINING OF ALUMINIUM: Li Guohua, Li Dianfeng, Zhao Xiangguo, Wang Qingna, Department of Nonferrous Metallurgy, Northeastern University, Shenyang, 110006, China

The Current-Voltage Method is adopted to study the cathodic process of Na+ in three-layer electrolytic refining of aluminium. The electrolyte studied is NaF-AlF3-BaCl2-NaCl molten system, whose composition is: NaF/AlF3 mole-ratio is 1.5-3.0, adding amount of NaCl is 0-8%, BaCl2 content is 60%. It is demonstrated that Na+ is not able to deposit on the aluminium cathode at the temperature range of 740-800°C while the cathodic current density is 0.4-0.8 A/cm2. This conclusion provides an important basis for modifying the electrolyte in three-layer electrolytic refining of aluminium.