Constellation, le dépôt institutionnel de l'Université du Québec à Chicoutimi

Résumé

The screening of the data obtained from a primary Aluminium production cell equipped with CF4, anodic current, line intensity, and voltage sensors during a six-week measurement campaign shows that the Low-Voltage Anode Effects (LVAE) at the scale of a single anode enshrine typical but various footprints into the anodic current signals. These footprints can exhibit shapes in ‘V’, of a backward facing step profile, or a combination of both. ‘V’ shapes tend to be relatively short term while ‘Step’ shapes are usually longer-lasting. The drop magnitude of these signatures furthermore covers a relatively wide range. ‘V’ shapes tend to be deep and ‘Step’ shapes shallow. Anodic currents are efficient to identify cell zones prone to LVAE. This is certainly helpful to assess and optimise feeding schemes. On the other hand, the variety of the patterns imprinted within the anodic current time series by LVAE events exacerbates the challenges to reliably detect those using automated systems. The understanding of the shapes these patterns can take provides a guidance for the future design of a detection algorithm.