In the 1991 proposed role for radionuclides in drinking water, EPA stated that GAC was not a best available technology for radon removal, although it has been shown to remove radon from drinking water (Kinner and others 1989; Lowry and Lowry 1987; Lowry and Brandow 1985). The agency cited problems with radiation buildup, waste disposal, and contact time. Since then, the SAB (July 1993, EPA-SAB-RAC-93-014 and July 1993, EPA-SAB-DWC-93-015) (EPA-SAB 1993b) has suggested that GAC might be an option for small systems with modest raw-water radon concentrations and that there could be problems with the thoroughness of EPA's analysis of the risk and disposal issues related to the use of GAC. In addition, new data that have become available since 1991 suggest that GAC might require shorter empty-bed contact times than originally thought (Cornwell and others 1999).
GAC was first shown to be an effective technique for removing radon from drinking water in the early 1980s (Lowry and Brandow 1981). As a result of its simplicity, it was installed in many private homes in New England where radon levels were high (1,111,000 to 11,111,000 Bq m-3) (Lowry and others 1991). By the late 1980s, studies of GAC units were producing data that suggested that gamma emissions from 214Bi and 214Pb, the short-lived progeny of radon, were substantial (2 × 10-6-5 × 10-4 Gy h-1) (Kinner and others 1989; Lowry and others 1988; Kinner and others 1987). In addition, accumulation of long-lived species (such as uranium, radium, and especially 210Pb) on the GAC was creating disposal problems (Kinner and others 1990; Kinner and others 1989; Lowry and others 1988). The radon-removal efficiency of some GAC units also decreased with time (Kinner and others 1993; Lowry and others 1991; Kinner and others 1990; 1989). It was the data from the studies in the late 1980s that led EPA to question the use of GAC as a best available technology in its 1991 proposed role. Furthermore, economic evaluations suggested that the cost of GAC treatment is high and in most situations not competitive with aeration, because of the large amount of carbon needed, especially for large radon loadings (high flow or high influent radon concentration) (EPA 1987b). A description of the GAC process is found in appendix C.
By their very nature, GAC systems are designed to sorb and retain contaminants. Thus, while a GAC unit is operating, the bed is accumulating radon. In addition, because of radon's short half-life relative to the GAC unit's run-time (months to years), radon comes to secular equilibrium with its progeny. The solid progeny remain sorbed to the GAC (Cornwell and others 1999; Kinner and others 1990; Kinner and others 1989; Lowry and Lowry 1987) which is not surprising inasmuch as GAC is known to have a high affinity for metals (such as lead)