|About this Abstract
||2016 TMS Annual Meeting & Exhibition
||Aluminum Reduction Technology
||Using Artificial Neural Network to Predict Low Voltage Anode Effect PFCs at the Duct End of an Electrolysis Cell
||Lukas Dion, Charles-Luc Lagacé, László Istvan Kiss, Sándor Poncsák
|On-Site Speaker (Planned)
Primary aluminum production is generating an important amount of greenhouse gases. CO2 is the dominant compound but during anode effects, important concentrations of perfluorocarbons(PFCs) are released as well. Continuous emissions of PFC have also been reported in small concentrations but the root causes of these emissions are hardly understood.
Measurements were taken at “Aluminerie Alouette” plant using Fourier-transformed Infrared Spectroscopy on individual cells to analyze the composition of the duct end gas. By correlating the variations of the concentration for the emitted gas with cell variables (voltage, intensity, and pseudo-resistivity) and individual anode currents, it was possible to develop a predictive model to quantify the tetrafluoromethane (CF4) emissions between 10 and 1000 ppb for individual cell emissions.
By analyzing the time history of the resulting data and by applying a post treatment process accordingly, it is possible to reduce the number of false predictions and increase precision of the final results.
||Planned: Light Metals Volume