Predicting As removal during metal hydroxide precipitation

Abstract

Data from two utility surveys led to the development of a simple model for predicting arsenate removal at full‐scale alum coagulation, ferric coagulation, and iron–manganese removal plants. A simplified isotherm is described that can predict the extent of arsenate removal at drinking water utilities practicing coagulation or iron–manganese (Fe–Mn) removal. If all possible sources of particulate iron and aluminum hydroxide present in the system are accounted for, the model predicts arsenic (As) removal to within ±13 percent (90 percent confidence) for Fe coagulation at pH 6.5–8 and alum coagulation at pH < 7.6. Analysis of full‐scale treatment data suggests that colloidal aluminum (Al) flocs with sorbed arsenate [As(V)] may pass through filters, thereby decreasing overall As removal efficiency. Thus, Al solubility and particle stability must be minimized to improve As removal. If stability and solubility of aluminum hydroxide flocs are not a problem, alum and Fe coagulants have nearly equal capacity for sorbing As(V). Survey results also demonstrate the importance of particulate As.