OBJECTIVES

The purpose of this study was to determine the feasibility and sustainability of in-situ removal of arsenic from water infiltrated through unsaturated alluvium.

BACKGROUND

Arsenic is naturally present in aquifers throughout the southwestern United States and elsewhere. In January 2006, the U.S. Environmental Protection Agency (EPA) lowered the Maximum Contaminant Level (MCL) for arsenic from 50 to 10 micrograms per liter (g/L). This raised concerns about naturally-occurring arsenic in groundwater. Although commercially available systems using sorbent iron or aluminum oxide resins are available to treat high-arsenic water, these systems are expensive to build and operate, and may generate hazardous waste.

Iron and aluminum oxides occur naturally on the surfaces of mineral grains that compose alluvial aquifers. In areas where alluvial deposits are unsaturated, these oxides may sorb arsenic in the same manner as commercial resins, potentially providing an effective low-cost alternative to commercially engineered treatment systems.

APPROACH

The Antelope Valley within the Mojave Desert of southern California contains a shallow water-table aquifer with arsenic concentrations of 5 g/L, and a deeper aquifer with arsenic concentrations of 30 g/L. Water was pumped from the deep aquifer into a pond and infiltrated through an 80 m-thick unsaturated zone as part of field-scale and laboratory experiments to treat high-arsenic groundwater and recharge the shallow water table aquifer at the site.

The field-scale recharge experiment included the following steps: 1) construction of a recharge pond 2) test drilling for sample collection and instrument installation adjacent to the pond 3) monitoring downward migration of water infiltrated from the pond 4) monitoring changes in selected trace-element concentrations as water infiltrated through the unsaturated zone

Data from instruments within the borehole adjacent to the pond were supplemented with borehole and surface geophysical data to evaluate the lateral spreading of water as it moved downward through the unsaturated zone.

Three laboratory studies were undertaken. Sequential extraction was used to evaluate the abundance of iron, aluminum, and manganese oxides and selected trace elements on operationally defined sites on the surfaces of mineral grains collected before and after infiltration from the pond. Secondly, radio-labeled arsenic-73 microcosm experiments evaluated the potential for incorporation of arsenic sorbed to exchange sites on mineral grains into less reactive crystalline mineral structures with time. Finally, column studies evaluated arsenic sorption and the pH dependence of sorption for selected unsaturated zone materials.

RESULTS/CONCLUSIONS

Between December 2010 and July 2012, more than 120,000 cubic meters (m3 ) (about 97 acre-feet) of high-arsenic groundwater was pumped from the deep aquifer into a 0.11 hectare (about 0.27 acres) pond and infiltrated though an 80-meter (about 260 feet) thick unsaturated zone to recharge a water-table aquifer.

Arsenic concentrations were lowered from 30 to 2 g/L as water infiltrated though the unsaturated zone at the site. Some uranium, possibly associated with past agricultural land use at the site, was mobilized to concentrations as high as 66 g/L within the unsaturated zone during the experiment. Uranium was resorbed and the high uranium concentrations did not reach the water table at the site. Concentrations of other trace elements, including antimony, chromium, vanadium, and selenium were low throughout the study.

Infiltration rates from the pond were as high as 0.4 meters per day (1.1 feet per day, ft/d), and the wetting front moved downward about 25 centimeters per day (cm/d) (0.8 ft/d) to a depth of about 50 m (about 165 feet). Clay layers at that depth slowed the downward movement of the wetting front to about 5 cm/d (0.16 ft/d). Lateral movement of the wetting front was monitored using sequential direct-current (DC) surface and sequential electromagnetic (EM) and DC borehole resistivity. Most lateral movement occurred on a clay layer about 50 m (about 165 feet) below land surface. Infiltrated water reached the water table in January 2013. At the water table, the “wetted footprint” of water infiltrated from the pond, indicated by surface resistivity data, was about 13 hectares (about 32 acres). On the basis of data collected at the site, there is enough sorbent material to operate this pond and treat groundwater having an arsenic concentration of 30 g/L to 2 g/L for about 500 years. Toxicity Characteristic Leaching Procedure (TCLP) data showed arsenic concentrations to be below hazardous levels beneath the pond after the experiment. Pond maintenance may be required to keep infiltration rates high, and prevent accumulation of organic material on the pond bottom, although organic material on the pond bottom may increase removal of other trace elements in infiltrated water including chromium, selenium, and vanadium.

Laboratory results are consistent with the field data and show sorption of arsenic in 10 cm (0.3 feet) columns to about 2 g/L over a pH range of 6 to 8, and at influent arsenic concentrations as high as 300 g/L, without breakthrough in 50 pore volumes. Column results suggest that the insitu treatment may remove arsenic in a range of hydrogeologic settings, and would not necessarily be restricted to alkaline alluvial aquifers common throughout the southwestern United States. Radiolabeled arsenic-73 experiments show that although arsenic is initially weakly sorbed (and potentially mobile), with time arsenic is incorporated into amorphous materials. One year after sorption onto surface exchange sites, most sorbed arsenic is incorporated into crystalline oxide minerals on the surfaces of primary mineral grains and is less mobile.

Results of the study suggest that long-term land use restrictions on sites used for in-situ treatment of arsenic may not be needed to control water applied to surface materials. This minimizes some regulatory concerns about future land use at sites used for in-situ arsenic treatment. However, future land uses that may alter reduction-oxidation conditions in the subsurface should be avoided, such as infiltration of stormwater recharge or recharge with other water having high organic carbon concentrations (including unsewered residential land use, dairy or other confined animal operations).