AWWA JAW37684

Arsenic "profiles" were obtained from full-scale conventional treatment (coagulation, Fe-Mn oxidation, or softening) plants, facilitating testing of theories regarding arsenic removal. Soluble As(V) removal efficiency was controlled primarily by pH during coagulation, by Fe+2 oxidation and Fe(OH)3 precipitation during Fe-Mn oxidation, and by Mg(OH)2 formation during softening. Insignificant soluble As(V) removal occurred during calcite precipitation at softening plants or during Mn+2 oxidation-precipitation at Fe-Mn oxidation plants. The extent of soluble As(V) removal during coagulation and softening treatments was lower than expected. Somewhat surprisingly, during coagulation As(V) removal efficiencies were limited by particulate aluminum formation and removal, because much of the added coagulant was not removed by 0.45-um-pore-size filters. At one utility, reducing the coagulation pH from 7.4 to 6.8 (at constant alum dose) improved removal of particulate aluminum, thereby enhancing soluble As(V) removal during treatment. Includes 9 references, tables, figures.