Three municipal wastewater treatment plants (WWTPs) in southeastern Pennsylvania were sampled to determine the presence and concentrations of 12 natural and synthetic estrogen hormones in the wastewater influent and effluent. The target estrogens were 17alpha-estradiol, estrone, estriol, equilin, 17alpha-dihydroequilin, 17beta-estradiol, 17alpha-ethinyl estradiol, gestodene, norgestrel, levonorgestrel, medrogestone, and trimegestone. One WWTP uses a biofilm reactor (packed-bed trickling filter),and the other two use suspended-growth media (continuously stirred activated sludge reactor and sequential batch reactor). Estrone was detected in all the three plants; estriol and estradiol were detected at two WWTPs; and 17 alpha-dihydroequilin and 17 alpha-ethinyl estradiol were detected at one WWTP. The concentration of estrogens in the influent and effluent of the three treatment plants ranged from 1.2 to 259 ng/L and 0.5 to 49 ng/L, respectively. The percentage removal of estrogens from the aqueous phase ranged from 41 to 99%, except in the case of 17alpha-dihydroequilin; the removal of 17alpha-dihydroequilin was negligible. The suspended-growth media systems showed higher removal efficiencies for estrogens than the biofilm system. The analytical method uses a Varian C-18 solid-phase extraction (Varian Inc., Palo Alto, California), followed by a derivatization with bis(trimethylsilyl)trifluoroacetamide. The detection limits for the estrogen compounds ranged from 0.1 to 10 ng/L using a sample size of 1 L. The method recoveries ranged from 71 to 120%, and the relative standard deviation ranged from 6 to 14% for all the hormones.
