Arsenic Contamination in Groundwater

Bridget Scanlon, principal investigator

Lowering the Federal standard for arsenic in drinking water from 50 ug/L to 10 ug/L, results in much more widespread arsenic contamination in groundwater in Texas. Potential anthropogenic sources of arsenic, such as arsenical pesticides in the southern High Plains and the southwestern Gulf Coast, were examined using GIS overlay analyses and soil sampling. Potential geologic sources of elevated arsenic concentrations in groundwater were evaluated in the southern High Plains and southwestern Gulf Coast using relationships between arsenic concentrations and different geologic units. Relationships between arsenic concentrations and other ions, particularly oxyanions, were evaluated using existing databases (TWDB, NURE, and TCEQ) to assess sources of arsenic.

Groundwater arsenic contamination is widespread in Texas. Approximately 6% of wells exceed the MCL of 10 ug/L. Contamination is focused in the southern High Plains (32% of wells exceed than the MCL) and the southwestern Gulf Coast (29% of wells exceed than the MCL).

Southern High Plains

The southern High Plains (SHP) was subdivided into two areas: a northern area (SHP-N) characterized by low total dissolved solids (TDS < 500 mg/L) and a southern area (SHP-S) characterized by high TDS (> 500 mg/L). Groundwater arsenic contamination is much greater in the SHP-S region (51% of wells > 10 ug/L) than in the SHP-N region (7% of wells > 10 ug/L). Regional analyses of groundwater arsenic concentrations do not support a surfical source of arsenic contamination. Arsenic concentrations are not correlated with land use, cotton production (arsenic acid used for defoliating cotton), distance from cotton gins, nitrate concentrations, water table depth, or aquifer saturated thickness. These results suggest the source of arsenic is not from the land surface.

Unsaturated zone studies were conducted to assess the potential for arsenical pesticides to provide a source of arsenic to groundwater. Results of drilling and sampling 18 boreholes in the southern High Plains indicate that the distribution of arsenic is not related to the distribution of cotton production and application of arsenical pesticides. High arsenic concentrations in a rangeland profile (peak 77 ug/kg) indicate that background levels of water soluble arsenic are high in soils. Arsenic levels in cultivated areas are variable. Some profiles have highest arsenic levels near the surface which are correlated with nitrate and phosphate that may suggest a fertilizer or arsenical pesticide source. These data indicate that arsenic related to arsenical pesticides is probably restricted to the near surface zone. Other profiles have peak concentrations in the middle of the profile or at depth. The unsaturated zone data indicate a widespread source of water soluble arsenic in soils in the southern High Plains that may contribute to groundwater arsenic contamination.

Groundwater arsenic contamination occurs in generally oxidizing conditions in the High Plains and limited data indicate that arsenic is in the form of arsenate. Groundwater arsenic concentrations were compared with concentrations of other ions to evaluate potential arsenic sources. Correlations between arsenic and other constituents (vanadium, r 2 0.65; fluoride r 2 0.30; molybdenum r 2 0.18; boron r 2 0.17; selenium r 2 0.14) suggest a geologic rather than an anthropogenic source. Arsenic concentrations are related to geologic units and are highest in the Ogallala aquifer and much lower in the Dockum aquifer. Potential sources of arsenic include volcanic ash beds in the Ogallala, black shales in the Cretaceous (Kiamichi Shale), and saline lakes. Additional studies will be required to assess geologic sources, including geophysical logging and stratified water sampling.

Southwestern Gulf Coast

Groundwater arsenic concentrations are much higher in the southwestern area of the Gulf Coast (29 percent of wells exceed the MCL) than elsewhere in the Gulf Coast (3.5 percent of wells exceed the MCL).

GIS analysis indicates that groundwater arsenic concentrations are not related to cotton production. Some counties with high levels of arsenic contamination do not have any cotton production (Live Oak and Duval Counties). Results of drilling and sampling 10 boreholes in the unsaturated zone indicate that arsenic concentrations are highest in a rangeland site where gin waste was ploughed into the field (1854 ug/kg at 1.3 m depth). Restriction of elevated arsenic related to gin waste to the upper ~ 2 m soil zone suggests that this is an unlikely source of groundwater arsenic. High arsenic concentrations in the shallow subsurface and correlation with nitrate suggests fertilizer or arsenical pesticide sources for another profile. High arsenic concentrations were found throughout an irrigated profile. The remaining profiles had low arsenic levels (< 10 ug/kg) that showed no systematic variation with land use or with depth.

Redox conditions range from mildly oxidizing to reducing in the Gulf Coast. Although arsenic concentrations are not related to dissolved oxygen, high arsenic concentrations do not occur at low redox potentials ~ -100 mV). However, conditions are not reducing enough to immobilize arsenic in sulfides. High arsenic concentrations occur along the Rio Grande valley, in the few counties west and southwest of Corpus Christi, and along the Catahoula Formation outcrop extending into the north eastern Gulf Coast . Correlations between arsenic and other constituents (vanadium, r 2 0.43; molybdenum r 2 0.36; boron r 2 0.12) suggest a geologic rather than an anthropogenic source. Arsenic concentrations are highest in the Jasper aquifer (48 percent > 10 ug/L) which immediately overlies the Catahoula Formation and are much less in younger stratigraphic aquifers (Evangeline aquifer; 21 percent > 10 ug/L and Chicot aquifer; 27 percent > 10 ug/L). Therefore, volcanic ashes associated with or reworked from the Catahoula Fm. are the most likely source of high arsenic concentrations in the southwestern Gulf Coast aquifer. Correlations between arsenic and other oxyanions typically associated with volcanism (molybdenum, vanadium) as well as the general decrease in arsenic contamination away from this formation strongly support this hypothesis.

This study represents an initial assessment of arsenic contamination in the southern High Plains and southwestern Gulf Coast and has resulted in a number of questions that may be addressed in future studies. The widespread distribution of water soluble arsenic in soils in both regions should be evaluated to determine if arsenic in rangeland and in deeper portions of cultivated profiles is related to volcanic ashes. Gamma logs should be conducted in boreholes to determine if there are high gamma levels that would indicate ashes. Water from leaching the soils should be analyzed for oxyanions and fluoride to assess relationships between soluble arsenic and these ions. Arsenic speciation should be conducted on selected samples to determine whether there are any organic arsenicals in the water. Playa water and saline lake water should be sampled to determine arsenic levels in these potential sources. Groundwater sampling should be conducted in different geologic units to assess potential correlations with arsenic contamination. Geophysical logging and multilevel sampling of groundwater should be conducted to determine if arsenic concentrations are stratified and if arsenic contamination can be linked to specific geologic units.

March 2006