MODELING ARSENIC(V) REMOVAL BY COAGULATION WITH FERRIC SALTS: EFFECTS OF pH AND DOSAGE

In 2001 the U.S. Environmental Protection Agency (EPA) lowered the Maximum Contaminant Level for arsenic from 50 μg/L to 10 μg/L. Public water suppliers are required to comply with the lower standard by January of 2006 and several thousand water systems nationally are looking for efficient and inexpensive ways to remove arsenic. Ferric coagulation/filtration and ferric coagulation/microfiltration are two promising techniques for attaining the new arsenic water standard. Both processes remove arsenic by sorption to ferric hydroxides (and possibly co-precipitation as well). Models of arsenic removal by sorption to ferric hydroxide have been developed, but conflicting information has been given to practitioners about the relative importance of pH and iron dosage. Edwards (1994) modeled arsenic removal using an equilibrium chemistry approach and reported that pH effects were more important than coagulant dosage. However, a subsequent empirical approach (McNeill and Edwards 1997) ignored pH and initial arsenic concentration entirely and modeled removal percentage as depending on dose alone. While the more recent empirical approach has the advantage of being validated by observations at actual water treatment plants, understanding the equilibrium chemistry may help to define boundaries for the acceptable use of the empirical approach. In this study, the equilibrium chemistry approach of Edwards (1994) is followed and a model of arsenic(V) removal by ferric coagulation developed. This model runs in Microsoft Excel and is available free of charge upon request. Charts of arsenic removal as a function of pH, dosage, and ionic strength are developed to assist with the selection of appropriate initial conditions for pilot studies. Increasing ionic strength is found to modestly decrease arsenic removal. Above pH 8, arsenic removal is found to decrease rapidly with increasing pH. However, removals are relatively insensitive to changes in pH below pH 8. Similarly, at low coagulant doses, arsenic removal increases rapidly with increasing iron concentrations. However, further increases in dose produce diminishing additional removal. In the absence of competing anions, there is little benefit to increases in dose beyond 5 mg/L as iron. Initial arsenic concentration has little effect on removal percentage until initial concentrations exceed 200 μg/l, at which point the percentage removal decreases as initial arsenic concentration increases. This research indicates that equilibrium chemistry modeling generally supports the omission of pH and initial arsenic concentration from the empirical model of McNeill and Edwards (1997), but only if the pH is below 8 and the initial arsenic concentration is below 200 μg/l.