Thor Anders Aarhaug

Particulate Emissions from Electrolysis Cells

In the dry cleaning of the exhaust gas from the aluminium cells impurities are accumulated in the finer fractions of secondary alumina from the dry scrubbers. The present work describes new methods for the determination of dust composition, aiming at increasing the understanding of the effect of cell operation on the amount and the composition of dust in the fume. New and advanced analysis methods are used to characterize a broad specter of emissions. An Electrical Low Pressure Impactor is used to sample and analyze the dust from the cells. The equipment enables real-time particle size distribution analysis of 12 particle classes in the range 30 nm — 10 μm. The size classified samples are analyzed by means of SEM/EDS and XRD to determine the characteristic chemical composition of the different fractions. Understanding the evolution, evaporation, and condensation of particulates in the cell emissions under different operational conditions may facilitate new standards for environmental friendly and energy efficient high amperage electrolysis cells.

Correlation between Moisture and HF Formation in the Aluminium Process

Hydrogen fluoride (HF) emission to the working atmosphere is still a problem in the aluminium industry. Moisture in secondary alumina fed to the cell and humidity in the ambient air reacts with fluorides in the bath and fluoride vapours to form hydrogen fluoride. The relation between the various sources of water and the resulting HF emission is still not well understood. In this work, industrial measurements have been done to determine where HF escapes from the bath. The quantities of HF and moisture at the specific sites have also been determined. Measurements were done in the duct during normal operation as well as during anode change, above the feeder hole and above an open hole in the crust. A strong correlation between feed cycle and HF levels was measured. Increased HF emissions were also recorded during anode change.

HF Measurements Inside an Aluminium Electrolysis Cell

HF emissions to the working atmosphere may still be a problem for the aluminium industry. The objective in the present work was to study how the HF evolution is distributed between feeder holes, other openings in the crust, gases diffusing through the crust, fumes from the secondary alumina residing on top of the crust etc. A movable “gas sniffer” connected to a Tunable Diode Laser was used to measure the HF concentrations at the above mentioned locations. The stationary HF level in an open flaming feeder hole was approximately 9000 ppm, when measured a few cm above the bath surface. In comparison, when the probe was positioned 5–10 cm above a crust area with good integrity, the HF concentration was in the range 5–10 ppm. The results support the notion that most of the HF evolves from open feeder holes and the tapping hole.

PFC evolution characteristics during aluminium and rare earth electrolysis

In addition to aluminium electrolysis, the electrolysis of rare earth (RE) metals from fluoride melts is a significant source of perfluorocarbon (PFC) emissions to the atmosphere. These processes have many similarities, they are both based on molten fluoride salt electrolysis at temperatures around 1000 °C, and are utilizing carbon materials as the anode. Although PFC emissions from aluminium industry and rare earth electrolysis have similar overall reactions, they are often reported to have different characteristics. In order to get a better understanding of these differences and similarities, different laboratory experiments focusing on anode reactions and gas compositions in Al2O3 saturated cryolite and REF3-LiF melts during aluminium and rare earth metal electrolysis were studied. The results obtained, combined with thermodynamic data analysis allowed to better understand onset, evolution and termination behaviour of PFC evolution in molten fluoride systems of different chemistries.

Anode Effect Initiation during Aluminium Electrolysis in a Two-Compartment Laboratory Cell

Most laboratory cells used in the investigation of the alumina reduction process use a single anode. When investigating the initiation of the anode effect an approach with more than one anode might give better results, as the probability of obtaining partial anode effect is higher. Additionally, the design is closer to the industrial, where several anodes are connected in parallel. The system constructed consisted of two anodes in separate electrolyte compartments connected in parallel with a single combined cathode. The results indicate that an anode can go in and out of partial anode effect with little influence on the current, although, kept untreated a full anode effect is likely imminent. The results also show that under certain current and alumina conditions, with only two anodes in parallel, an anode can handle approximately the whole load of a fully passivated anode for a certain time.

Raw Gas Particles and Depositions in Fume Treatment Facilities in Aluminium Smelting

The trace element distributions in depositions found in the gas treatment centres (GTCs) exhibit striking similarities with the coarse pot gas particulates captured in a standard cyclone. To gain a better understanding of possible scale formation mechanisms and components involved, several analysis techniques were applied and compared. Pot exhaust coarse and fine particles, as well as scale samples collected form the gas treatment centre of the same Al-smelter were evaluated. LECO and Sintasyzer results, XRD, IR spectra and XPS pattern of grey scale samples and raw gas particle fractions are presented. Recrystallization of sodium fluoroaluminates due to HF adsorption reactions in combination with moisture is suggested as one formation mechanism.

Validation of Online Monitoring of PFC by QCL with FTIR Spectroscopy

Monitoring of perfluorocarbon (PFC) evolution from aluminium smelting is gaining attention, not only because of their high greenhouse gas potentials but also due to process optimization purposes. Conventionally, PFC monitoring has been conducted by extractive sampling and subsequent analysis by fourier transform infrared (FTIR) spectroscopy. With FTIR, the quantification can be performed by IR spectral features specific for PFC. The downside is a requirement of gas scrubbing to remove HF detrimental to the instrument as well as relatively poor gas dynamics due to the large internal gas volume of the instrument. With emerging quantum cascade laser (QCL) technology, online monitoring can now be conducted with duct mounted lasers with calcium fluoride optical windows. However, due to a strong spectral overlap of CF4 and other gas constituents present in the process (e.g. methane), the QCL instruments currently suffer from some cross-interference. In this work, QCL single cell PFC monitoring has been validated by simultaneous monitoring with FTIR.

A study of anode baking gas composition

A method has been developed to measure gas composition inside the pit in open anode baking furnaces. The gas composition can be used to understand attack on and degradation of the refractory lining, baking behavior and combustion energy contribution through the baking cycle. A probe was installed in the packing coke near the bottom of the pit while extracting gas over several days with continuous analysis with an FTIR spectrometer. The results show a clear temperature dependence of CO and CO2 composition. Methane was found to be the dominating gas species at the beginning of the measuring cycle. Fluoride gases was also present, indicated by reactions with the glass wool filter to yield SiF4 that was detected in small amounts. PAH condensates were observed but not systematically determined in the present campaign. Earlier results from PAH measurements before the scrubbing, showing large fluctuations, will be discussed in relation to present findings.

Perfluorocarbon Formation During Rare Earth Electrolysis

A challenge during rare earth (RE) electrolysis is to avoid emissions of perfluorocarbon (PFC) green-house gases. The objective of this work was to study how to operate the RE electrolysis process with neither PFC formation nor anode effect. Linear Sweep Voltammetry was carried out at 1050 and 1100 °C, and electrolysis was performed in REF3-LiF melts at ca. 1050 °C during on-line off-gas analysis. To avoid anode effect, the current density values must be strictly less than 0.43 and 0.68 A cm−2 at working temperatures of 1050 and 1100 °C, respectively. The optimal REO batch feed rate for avoiding PFC formation could be established by correlating the onset of PFC with the values and the changes that occurred in the anode potential.

Partial Anode Effect in a Two-Compartment Laboratory Alumina Reduction Cell

Most laboratory systems investigating the aluminium production process utilize a single anode set-up. When approaching alumina depletion under constant current conditions in such a system, the potential will increase to high levels (>10 V) and initiate an anode effect and perfluorocarbon generation. However, it has been discovered by industrial measurements that perfluorocarbon generation may also occur at normal cell voltages. With the use of a two-anode setup in parallel with an electronic load this phenomena was investigated in the laboratory. The results indicate that as long as the rest of the cell can acquire the extra load, partial passivation of one or more anodes is possible and can be accompanied by small amounts of PFC evolution (0–3 ppm mol CF4). Individual anode potentials can be highly elevated, albeit the changes get buried in the total cell voltage. Only when the total load becomes too large the voltage rises abruptly and substantial amounts of PFC can be produced (≫1000 ppm mol CF4).