Cache Creek drains part of northern California’s Coast Ranges and is an important source of mercury (Hg) to the Sacramento–San Joaquin Delta. Cache Creek is contaminated with Hg from several sources, including historical Hg and gold mines, native Hg in the soils, and active mineral springs. In laboratory experiments in a study conducted by the U.S. Geological Survey, in cooperation with the U.S. Environmental Protection Agency, the use of coagulants and sorbents to immobilize Hg in water samples from high- concentration sources in the Cache Creek watershed was investigated. Three sites were selected for the collection of surface-water samples containing high and low concentrations of particulate-associated Hg. The high-particulate Hg samples were collected from Cache Creek Settling Basin duringstormflow conditions. The low-particulate Hg samples werecollected from two geochemically contrasting sites duringbase-flow conditions (downstream from a geothermal springand at the emergence point of a connate-water spring).
Three coagulants were chosen for laboratory testing with the high-particulate sample— (1) ChitoVanTM HV 1.5 percent (shell based), (2) FerralyteTM 8131 (ferric sulfate based), and (3) UltrionTM 8186 (aluminum based). Each coagulant was tested at various dose amounts to determine the optimum dose rate for the high-particulate sample. The low-particulate source samples were passed through three sorbents—
(1) chitosan flakes, (2) coconut shell-based activated carbon,and (3) coal-based activated carbon. In-line columns were packed with each material, and the untreated sample waspassed through each column at three different flow rates (0.1,0.5, and 1.0 liter per minute, L/min).
For dose rates used in this study, ChitoVanTM reduced turbidity of the particulate sample by 85–91 percent, FerralyteTM reduced turbidity by 54–93 percent, and UltrionTMreduced turbidity by greater than 90 percent. At the lowest dose rate, ChitoVanTM achieved a 59- to 61-percent reduction in whole-water methylmercury (MeHg) concentrations and a 71- to 75-percent decrease in whole-water total mercury (THg) concentrations. FerralyteTM achieved a 37- to 48-percent decrease in whole-water MeHg concentrations and a 37- to 48-percent reduction in whole-water THg concentrations. UltrionTM achieved a greater than 90-percent decrease in whole-water MeHg and THg concentrations.
Mercury removal from the low-particulate samples was less efficient for the sorbent materials compared to the coagulants; less than 30 percent of THg was removed from any 500-milliliter aliquot using sorbent materials. The coal-based sorbent was the most versatile of the sorbents, removing THg to a similar extent from both low-particulatesource waters. The chitosan sorbent was the most effective atremoving THg from the low-particulate stream sample, but less effective for the low-particulate connate-spring sample. The Hg removal efficiency of the coconut sorbent decreasedquickly compared to the other two sorbents, indicating that sorption may be limited by the short contact times evaluated in this study.