Greenhouse gas (GHG) generation is inherent in the production of aluminium by a technology that uses carbon anodes. Most of those GHG are composed of CO 2 produced by redox reaction that occurs in the cell. However, a significant fraction of the annual GHG production is composed of perfluorocarbons (PFC) resulting from anode effects (AE). Multiple investigations have shown that tetrafluoromethane (CF 4 ) can be generated under low-voltage conditions in the electrolysis cells, without global anode effect. The aim of this paper is to find a quantitative relationship between monitored cell parameters and the emissions of CF 4 . To achieve this goal, a predictive algorithm has been developed using seven cell indicators. These indicators are based on the cell voltage, the noise level and other parameters calculated from individual anode current monitoring. The predictive algorithm is structured into three different steps. The first two steps give qualitative information while the third one quantitatively describes the expected CF 4 concentration at the duct end of the electrolysis cells. Validations after each step are presented and discussed. Finally, a sensitivity analysis was performed to understand the effect of each indicator on the onset of low-voltage PFC emissions. The standard deviation of individual anode currents was found to be the dominant variable. Cell voltage, noise level, and maximum individual anode current also showed a significant correlation with the presence of CF 4 in the output gas of an electrolysis cell.