TMS Logo

About the 1996 TMS Annual Meeting: Thursday Morning Sessions (February 8)

February 4-8 · 1996 TMS ANNUAL MEETING ·  Anaheim, California


Proceedings Info

Sponsored by: LMD Aluminum Committee

Program Organizer: Ms Fiona J Stevens, Comalco Research and Technology, Comalco Research Centre, PO Box 316, Thomastown, Victoria 3074, Australia

Thursday, AM Room: A9

February 8, 1996 Location: Anaheim Convention Center

Session Chairperson: Jennifer Purdie, Boyne Smelters Ltd, PO Box 524, Boyne Island, Gladstone, Queensland 4680, Australia

8:30 am

ON THE MECHANISMS OF ALUMINA DISSOLUTION WITH RELEVANS TO POINT FEEDING ALUMINIUM CELL: Ove Kobbeltvedt, Jomar Thonstad Department of Electrochemistry, Norwegian Institute of Technology, N-7034 Trondheim, Norway; Sverre Rolseth, SINTEF Materials Technology, N7036 Trondheim, Norway

The dissolution of alumina in cryolite melts was studied in a laboratory apparatus intended to simulate the conditions in industrial cells with point feeding. Rapid dissolution is achieved if (i) a large proportion of the batch of alumina added is dispersed in the bath upon addition, and (ii) the agglomerates which form, disintegrate quickly. Bubbling of gas through the melt increased the dissolution rate by promoting the dispersion of alumina grains as well as breaking up the agglomerates. Preheating of the alumina also resulted in faster dissolution, mainly due to effective dispersion of the alumina grains. High superheat of the bath enhanced the dissolution rate of the agglomerates. Measurements on a point fed industrial cell showed that the dissolution rate of the alumina decreased when the feeding hole was plugged compared to the situation when the feeding hole stayed open.

9:00 am

TRANSPORT OF DISSOLVED ALUMINA IN POINT-FED SODERBERG REDUCTION CELLS: K. Torklep, K. Kalgraf, T. Nordbo, Elkem a/s Research, PO Box 40, Vagsbygd, N-4602 Kristiansand, Norway

Knowledge on how dissolved alumina is distributed in the interpolar region is required to determine the optimum number of point feeders and their location, and the particular demands of Soderbergs make a straightforward application of experiences obtained with prebake technology suspect. We have studied the transport of a radioactive, isotopic tracer (24NaC1) injected molten in the bath in gram amounts at existing and proposed point feeder locations. A system of air-cooled Geiger-Muller detectors placed in the pot permitted continuous monitoring of the relative tracer concentration at six different locations. Convection velocities, patterns, turbulent diffusivities, and mixing times were obtained. Results are used in two-dimensional modelling of the oxide concentration in the bath.

9:30 am

SOME PITFALLS IN USING CELL RESISTANCE TO CONTROL ALUMINA CONCENTRATION: Richard Brown, Collin MacPherson Alcan Australia Ltd - Kurri Kurri Smelter, PO Box 1, Kurri Kurri, NSW 2327, Australia

Modern aluminium cells use sophisticated computer control algorithms to control the alumina concentration of the cell. The alumina concentration is inferred from the bath resistance as it cannot be measured directly. Stable and efficient cell feeding is dependent on the accurate interpretation of the cell resistance changes over time. This paper describes how, in certain conditions, mobile sludge can mask the true bath alumina concentrations. Under these conditions it is possible for the alumina feed control algorithm to "overfeed" the cell. This situation can have a negative effect on cell stability and cell operation making good alumina control more difficult than is often portrayed.

10:00 am

THERMAL EFFECTS BY ANODE CHANGING IN PREBAKE REDUCTION CELLS: Frank Anune, Institute of Inorganic Chemistry, University of Trondheim, N-7034 Trondheim-NTH, Norway; Marvin Bugge, Norsk Hydro a.s, Research Centre, N-3901 Porsgrunn, Norway; Halvor Kvande, Trygve Ringstad, Hydro Aluminium a.s, KarmØy Plant, N-4265 Havik, Norway; Sverre Rolseth, SINTEF Materials Technology, N-7034 Trondheim-NTH, Norway

Thermal effects caused by anode changing in prebake alumina reduction cells were studied by temperature measurements using thermocouples located at different positions inside the anodes. Anode changing caused a temperature reduction in the nearest neighbour anode of typically 20 to 40[[ring]]C, as measured close to its working surface. The minimum temperature was recorded about 5 hours after the anode changing, and it could take up to 24 hours before the old anode again showed a stable temperature reading. The cooling effect is due to heat radiation losses to the new anode, and thus the heat balance of the neighbour anodes is altered significantly after an anode changing. Temperatures were also measured inside the newly set anode. If the new anode voluntarily was set too low by 20 mm to provoke a formation of a spike on the working surface, the measured temperatures inside this anode could reach as high as 1080deg.C.

10:25 am BREAK

10:35 am

BATH AND LIQUIDUS TEMPERATURE SENSOR: Paul Verstreken, Heraeus, Electronite Int. N.V. Grote Baan 27a, 3530 Houthalen, Belguim; Siegfried Benninghoff, Hoogovens Aluminium Huttenwerk GmbH, ScheusensstraBe, 46562 Voerde, Germany

A newly developed sensor which measures both bath and liquidus temperature is described. The difference between both is known as superheat. Determination of the liquidus temperature is based on interpretation of the time-temperature trace during cooling down of a sample taken from the bath. Supercooling is avoided by vibrating the sample. Accuracy and reproducibility of the sensor is discussed. Examples of measurements and series of measurements in point-feed and center-worked cells are given. Evolution of bath and liquidus temperature is shown on short time bases, with respect to feed strategies and on longer time bases. To get better information out of temperature measurements, samples taken, or liquidus measurements performed, it is necessary to relate to timing of the feed strategy in use. Whether analyzing bath samples can be replaced by superheat measurements for cell control purposes will follow from ongoing experiments.

11:00 am

FURTHER DEVELOPMENT OF THE TEMPERATURE MODEL: Peter M. Entner, Alusuisse Lonza Services Ltd, Research and Process Development, 3965 Chippis, Switzerland; Gunnar A. Gudmundsson, Icelandic Aluminum Company Ltd., Straumsvik, PO Box 244, 222 Hafnarfordur, Iceland

At TMS in 1995 a model was presented to control the bath temperature of electrolytic pots. Regression methods calculate the parameters of model equations using historical data. Optimal set values of the pot voltage and aluminum fluoride feed are then determined. A pot may be in the reactive or inactive pot state. In the reactive state it responds readily to changes, e.g. an increase AlF3 feed causing an increased AlF3 concentration. In the inactive state however the bath temperature or the AlF3 concentration for instance may stay constant even when pot operation makes strong efforts to change the pot state. Especially interesting is the transition from one state to the other. To predict the moment of transition the temperature model uses indicators, derived from the time behaviour of pot parameters like metal height, bath height or alumina feeding. With the knowledge of a coming transition the model adjusts the optimal pot parameters. A mechanism is presented to explain the reactive and inactive pot states and the transitions.

11:30 am

PROCESS SIMULATION OF ALUMINUM REDUCTION CELLS: Imad Tabsh, CompuSIM Inc. 1003D 55 Avenue N.E. Calgary, Alberta, Canada T2E 6W1; Marc Dupuis, GéniSim, 3111 Alger, Jonquière, Québec, Canada G7S 2M9; Alexander Gomes, Alcan Aluminio do Brasil

A program was developed to model the dynamic behavior of an aluminum reduction cells. The program simulates the physical process by solving the heat and mass balance equations that characterize the behavior of eleven chemical species in the system. It also models operational events (such as metal tapping, anode change, etc.) and the process control logic including various alumina feeding policies and anode effect quenching. The program is a PC based Windows applications that takes full advantage of the Windows user interface. This paper describes the implementation of the process model and the control logic. Various results using the simulation are also presented.

The information on this page is maintained by TMS Customer Service Center (

Search TMS Annual Meetings TMS Meetings Page About TMS TMS OnLine