The least‐cost method of removing arsenic may not be the best method all around. The best arsenic treatment technique for a given utility will depend on arsenic concentration and species in source water, other constituents in the water, existing treatment processes, treatment costs, and handling of residuals. To evaluate these issues, a national survey investigated arsenic occurrence and speciation in US drinking water sources. In general, total arsenic concentration was higher in groundwater than in surface water supplies. Particulate arsenic was more abundant than previously suspected, and more arsenate than arsenite was present. The cost of arsenic treatment increased in the following order: modified conventional treatment << activated alumina or anion exchange < reverse osmosis. Nevertheless, the most cost‐effective treatment still might not be best, because secondary treatment benefits and residuals handling should also be taken into account.

Analysis of

A modified field technique can quantify particulate As, soluble As(III), and soluble As(V) in drinking water. Several problems with commonly used techniques that analyze for arsenic are reported, and solutions are proposed. Analytical techniques can accurately detect <0.5 μg/L total arsenic if certain conditions are met: iron, nitrate, and other interferences are overcome in hydride generation techniques, poor recoveries are overcome in graphite furnace atomic adsorption techniques, and chloride interferences are accounted for in inductively coupled plasma mass spectrometry techniques. Because no techniques adequately preserve arsenic species during storage, a speciation protocol that is fast (about 5 min/sample), relatively inexpensive (about $10/resin column), and user‐friendly and that can be applied in the field was developed. Particulate arsenic represents a significant fraction of total arsenic in drinking water.

Data from US arsenic occurrence surveys are used to estimate compliance violations given a range of possible arsenic standards. Three surveys of arsenic occurrence in US drinking water supplies were synthesized to estimate how possible arsenic standards would affect compliance. Detectable levels of arsenic (0.5 μg/L) were found in 73 percent of surface water sources and 58 percent of groundwater sources. For finished water, detectable arsenic occurred in 45 percent of surface water systems and 53 percent of groundwater systems. The authors' compliance projections for arsenic standards ranging from 2 to 20 μg/L agreed fairly well with initial estimates by the US Environmental Protection Agency. It was estimated that approximately 25 percent of all community water suppliers (11,550–11,890 systems) would violate an arsenic standard of 2 μg/L. Between 6 and 17 percent of systems (2,775–7,870) were projected to violate an arsenic standard of 5 μg/L, and 1–3 percent of systems (510–1,360) were estimated to violate an arsenic standard of 20 μg/L.

No abstract