This work presents an application of clustering in the phosphate industry. It focuses on the identification of fluorine losses in a phosphoric acid concentration unit. A clustering method using different algorithms was applied to group the data according to the importance of fluorine losses. As a result, the Gaussian mixture model was selected as the best method for this study.
Consequently, two important clusters of data were identified. The first one is comprised of data showing high losses and the second one is comprised of data showing low losses. In addition, a ratio R was calculated to quantify the losses. This Ratio enabled the detection and measurement of the level of fluorine losses in a phosphoric acid concentration unit.