Removal of arsenic from water by ferric chloride

Abstract

Widespread arsenic contamination of groundwater has become a major public health concern in Bangladesh, where water supply is heavily dependent on groundwater. The present study focused on removal of arsenic from groundwater by coagulation with ferric chloride. Removal of both As(V) and As(III) was found to increase with increasing coagulant (ferric chloride) dose. For higher coagulant (ferric chloride) doses, removal efficiencies appear to be independent of initial arsenic concentration. Removal efficiency of As(III) by ferric chloride was found to be significantly lower than that of As(V). For example, 25 mgIL of ferric chloride could bring down As(V) concentration to less than 10 f.lg/L from an initial concentration of 0.25 mg/L; whereas in case of As(III) a ferric chlonde dose as high as 125 mgIL resulted in a residual arsenic concentration of about 55 f.lg/L, higher than the drinking water standard (50 f.lg/L) set by the Bangladesh EQS. It was found that removal efficiency of As(III) pre-oxidized with bleaching powder was almost identical to that of As(V). For a sorbate/sorbent ratio (expressed as f.lg As/mg Fe) of about 50 or less, removal efficiencies exceeding 95 percent could be achieved, irrespective of the initial arsenic concentration. This can serve as a guideline value or thumb rule in determining required dose of ferric chloride for arsenic removal. A maximum adsorption density of about 353 f.lg As/mg Fe was achieved with a As(III) (pre-oxidized) concentration of 2.0 mg/L and ferric chloride concentration of 25 mg/L. However, at optimum sorbate/sorbent ratios (i.e., 50 or less) maximum adsorption density achieved was much less, about 50 f.lgAs/L permg/L Fe. . Since significant fraction of arsenic in the groundwater of Bangladesh can exist as As(III), it appears that arsenic removal technology based on ferric chloride coagulation must involve a pre-oxidation step for converting As(III) to As(V). In this study, good removal of As(III) was achieved in jar tests using a chlorine (bleaching powder) dose 3 to 4 times that calculated from stoichiometric consideration. However, the unstable nature of bleaching powder is a cause of concern in' its use as an C'v;nant. Results obtained from this study enabled determination of ferric chloride dose required for achieving a particular removal goal for arsenic present at different initial concentrations. It was found that ferric chloride requirement for satisfying the WHO standard (10 flg/L) was significantly higher than that required for satisfying the Bangladesh standard of 50 flg/L. Thus, the cost of arsenic removal by ferric chloride would depend on the water quality standard it is designed for. This is important in view of the possible revision of the drinking water standard for arsenic in many countries. In Bangladesh, arsenic removal by coagulated floes of ferric hydroxide could be a very effective technique, particularly in view of the presence of elevated levels of iron in many regions of the country. Removal of As(V) appears to be maximum in the neutral pH range and slightly depressed at higher pH values, exceeding 8. Since pH of groundwater usually falls between 6 and 8 in Bangladesh, pH is not likely to play a major role in determining the removal of arsenic from groundwater by ferric chloride. Although pH has a more pronounced effect on As(lII) removal, this may not be an important consideration for arsenic removal by ferric chloride involving pre-oxidation. Phosphate appears to have a dramatic effect on arsenic removal from groundwater. Removal of arsenic present at an initial concentration of 0.10 mg/L (and treated with 10 mg/L of ferric chloride) came down frs", 01. percent (in the absence of phosphate) to about 46 percent in the presence of 2 mg/L of phosphate. Relatively high concentrations of phosphate, reaching as high as 5.8 mg/L, have been found in some places of Bangladesh and it appears that ferric chloride doses determined from laboratory batch studies (without the presence of phosphate) can significantly underestimate the actual dose requirement. Effect of phosphate on arsenic removal by ferric chloride also suggest that phosphate ions can play an important role in the release of arsenic in the groundwater environment in Bangladesh.