Stein Rørvik

Measurement of Cathode Surface Wear Profiles by Laser Scanning

The service life time for high amperage aluminium reduction cells with graphitized cathodes is limited by cathode wear. The wear is normally very non-uniform, and it is commonly documented by photography and/or manual point measurements. In an attempt to record the wear pattern in a much more detailed way, a laser scanning procedure was developed. A laser scanner with a single point accuracy of 10 mm has been used to produce a 3D model based on three overlapping scans with an average resolution of about 1 cm. The same cathodes were also measured manually for comparison. The method developed gives detailed information regarding the wear at different positions within the cell, and it may become a valuable tool for investigating the influence of different parameters on the cathode wear.

INVESTIGATION OF THE CATHODE WEAR MECHANISM IN A LABORATORY TEST CELL

Cathode wear has become one of the major challenges for the life time of high amperage aluminum reduction cells due to the use of graphitized cathodes. The fundamentals of the cathode wear are still a matter of debate, and a laboratory procedure for testing of cathode materials is desired. Here, we present a laboratory electrolysis cell, which has been designed for cathode wear tests of industrial cathode materials. The formation and transport of aluminum carbide have been considered to be an important factor for cathode wear, and the laboratory test cell was designed in such a way that the cathode is exposed directly to the electrolyte. Aluminum carbide formed at the cathode may dissolve directly into the electrolyte. Here we present the study of the cathode wear of a commercial high density graphitic material, where the influence of the cathode surface morphology, diffusion and hydrodynamics in the electrolyte, have been in focus. The cathode wear and the penetration of electrolyte into the cathode were investigated by optical and electron microscopy. The influence of current density, hydrodynamics and transport of carbide in the electrode are discussed in relation to the experimental results.

Characterization of Prebake Anodes by Micro X-ray Computed Tomography

As part of the continuous work in improving anode quality at Hydro Aluminium, series of pilot scale anodes have been manufactured with systematic changes in coke type and green paste production including recipe. The anodes also support M.Sc., Ph.D. and PostDoc work in programs supported by Hydro and the Norwegian Research Council. In addition to regular analysis, the pore, void and grain distribution has been investigated using Micro X-ray Computed Tomography (µCT). This non-destructive 3D imaging is now implemented at a cost to allow larger numbers of samples, and a methodology has been developed by SINTEF that yields surprising sharp detail, suited to interpret important structural factors for relatively large anode volumes of 10–130 mm diameter. Given a better cost to information yield than image analysis and mercury porosimetry, Hydro will continue to support academic work with CT analysis. Examples will be shown of baked anodes before and after electrolysis testing plus crack patterns after mechanical testing.

Determination of Coke Calcination Level and Anode Baking Level ‐ Application and Reproducibility of L‐sub‐c Based Methods

The average crystallite size (L-sub-c or LC) is an important property of carbon materials for aluminium electrolysis; LC is used for characterizing the petroleum coke calcination level and sometimes also to estimate the baking level of anodes. This paper discusses problems when comparing LC results from different laboratories using precision statements from ASTM and ISO standards. The main cause is peak broadening errors introduced by the XRD instrument and sample preparation. The LC standards ASTM D5187 and ISO 20203 neglect these errors. Two ways are demonstrated to minimize the peak broadening effect to improve the standards, 1) by using thin sample thickness and 2) by embedding the coke in a high absorptive medium. Using LC to determine the anode baking level is discussed and three practices are discussed; measurement on the anode directly or two methods for using a reference coke that is baked with the anode. It is shown that precision is better for the latter methods. Especially for underbaked anodes a baking level estimated from measurement of the anode LC can be misleading.

Understanding the Anode Porosity as a Means for Improved Aluminium Smelting

The anode pore structure affects gas diffusion as well as air and carboxy reactivity burn-off during the aluminium electrolysis. The coarse porosity was investigated using X-ray computed tomography, the mid-range porosity using optical microscopy, and the finest porosity using mercury intrusion porosimetry. These methods were combined to gain a deeper understanding of porosity and how this affects anode quality and the potential for energy savings. Test materials were received from Hydro Aluminium with different coke sources, particle sizes, binder levels, and mixing temperatures. The paper will present measurement results and link these to production and process factors. The findings indicate that direct comparison for each technique is challenging because of the physical basis of the measurement and the different measurable pore size ranges. However, the combination of different techniques gives valuable insight to understanding anode porosity. Coke type, granulometry, pitch content and baking process affecting anode porosity, properties and reactivity.

Effect of Coke Properties on the Bubble Formation at the Anodes During Aluminium Electrolysis in Laboratory Scale

The anodic reaction of aluminium electrolysis cells leads to the formation of CO2 bubbles, which partly screen the anode surface and leads to an increase in the cell voltage. An advantage of these bubbles is that the formation and release contribute to the stirring of the electrolyte, however, the screening of the surface increases the irreversible energy losses. The voltage and current oscillation due to the bubble evolution during electrolysis for different anode materials have been determined in a laboratory cell. The effect of coke sulphur content and grain sizes were investigated. Anodes with finer coke fraction showed lower oscillations than coarser fraction equivalents. Additionally, the influence of current density on the amplitude of the anode potentials was measured. A 64% increase of current density caused an increase of anode potential oscillations from 79 to 170%.

Formation of Carbon Build-Up on the Flue Wall of Anode Baking Furnace

A hard carbon build-up layer often forms on the flue wall surface in anode baking furnaces. The layer accumulates over thermal circles and needs to be mechanically removed regularly to ensure sufficient space for the anodes between flue walls. The underlying mechanisms are still unknown and the extent of the carbon build-up varies from plant to plant. The build-up on the flue wall, taken from an autopsy of an open top furnace, has been examined. Microstructure and phase compositions of the carbon build-up, especially towards the refractory interface, were studied by optical microscopy, X-ray computed tomography (CT), SEM/EDS, and XRD. Pyrolytic carbon was found to be the main part of the carbon build-up layer in addition to packing coke particles. The transport of silicon from the refractory material, condensating on the flue wall surface, is found as nucleation sites for the formation of carbon build-up. Formation mechanisms of the carbon build-up are proposed with reaction schemes supported by thermodynamic calculations.

Cathode Wear Based on Autopsy of a Shutdown Aluminium Electrolysis Cell

To investigate cathode wear, an autopsy of a shutdown aluminium electrolysis cell was conducted. The original lining consisted of a fully impregnated and graphitized carbon block and the cell was shut down after 2461 days operation. The cell was cleaned down to the surface of the carbon cathode, revealing the profile of the cathode wear. Generally, the cathode wear was uneven across the cell with typical potholes. At a finer length scale, the wear was characterized by small “pitholes” resembling wide shallow pitting corrosion. Samples of the cell lining were obtained by drilling cylindrical samples at different locations in the cell. These samples were analysed with respect to phase composition and microstructure by a combination of X-ray computed tomography, optical and electron microscopy. The findings are discussed in relation to the current understanding of the underlying mechanism(s) for cathode wear.