Occurrence of natural and anthropogenic hexavalent chromium (Cr VI) in groundwater near a mapped plume, Hinkley, CA

Task 8

Fate of chromium during and after in-situ reduction

PG&E is injecting a carbon source (ethanol) into the alluvial aquifer downgradient from the compressor stations to create reducing conditions within the plume that reduce soluble Cr VI to insoluble Cr III; and thereby, remove Cr VI from solution. As previously discussed, there is concern from the CAC over dissolution of manganese and arsenic from aquifer solids within the reduced zone (known as the In-situ Reactive Zone, or IRZ) produced by this process. There also is concern that chromium removed from solution within the IRZ may be remobilized over decadal time-scales through reoxygenation of treated aquifer material as oxic groundwater recharged from the Mojave River flows through the aquifer.

To evaluate the plausibility of chromium reoxygenation and remobilization, consider a 1,000 cm3 volume of aquifer material having a porosity of 0.3 cm3/cm3. Assuming a mineral grain density of 2.65 g/cm3 (density of quartz), this 1,000 cm3 of aquifer material would contain about 1.8 kg of alluvium and 300 cm3 (0.3 liters) of groundwater. Given a Cr VI concentration in groundwater within the plume of 1,000 µg/L, 300 µg of chromium would be added to a 1,000 cm3 volume of aquifer material as a result of in-situ reduction of Cr VI to Cr III. In this example, the chromium concentration of the aquifer material would increase by 170 µg/kg as a result of in-situ reduction.

Chromium concentrations on aquifer material extracted using a citrate bicarbonate dithionite (CBD) extraction (Mehra, and Jackson, 1960) to completely digest iron, manganese, and aluminum oxide surface coatings from alluvium within the IRZ in Hinkley Valley ranged from 0.5 to 6 mg/kg (ARCADIS, Inc., 2011). This value is similar to the range of chromium extracted from oxide surface coatings on alluvium eroded from granitic rock in the elsewhere in Mojave Desert ranged from 0.3 to 3 mg/kg (Izbicki and others, in review) using the strong acid digestion proposed for this study (Chao and Sanzalone, 1989). Both data sets indicate an order of magnitude variation in total extractable chromium on aquifer material, and show the chromium added to aquifer material through in-situ reduction is small compared to the mass of naturally occurring chromium obtained through a complete digestion of crystalline iron, manganese, and aluminum oxides commonly occurring on the surfaces of mineral grains.

Consistent with these data, ARCADIS, Inc. (2011) did not find a significant difference in chromium concentrations in CBD extracts from iron, manganese, and aluminum oxide surface coatings on alluvium collected within the in Hinkley IRZ before and after in-situ reduction (ARCADIS, Inc., 2011). On the basis of X-ray spectrographic data, chromium in post-IRZ alluvial material was associated with iron oxide minerals on the surfaces of mineral grains (ARCADIS, Inc., 2011). As a consequence, reduced Cr III produced as a result of in-situ reduction was claimed to be relatively immobile. However, the experimental design at Hinkley, and similar work at the PG&E compressor station at Topock, CA (ARCADIS, Inc. 2009), did not distinguish between comparatively abundant naturally-occurring chromium associated with the crystalline oxide mineral surface coatings measured using the CDB extraction, and the smaller mass of chromium added by activities within the IRZ to the more reactive "non-specifically sorbed" and "specifically sorbed" operational fractions (measured as part of Task 2) (Wentzel and others, 2001).

In contrast to previous interpretations (ARCADIS, Inc., 2009, 2011), it is possible that reduced chromium added as a result of in-situ reduction may initially be sorbed to aquifer material through cation exchange, rather than directly incorporated into crystalline oxide mineral structures on the surfaces of mineral grains. Chromium extracted from "non-specifically sorbed" and "specifically-sorbed" operational fractions (Wenzel and others, 2001) within alluvium eroded from granitic rock in Antelope Valley was less than 6 and 4.5 µg/kg. These data suggest the mass of chromium added to aquifer material by in-situ reduction (170 µg/L in the example above) may be several orders of magnitude greater than chromium sorbed in the more reactive fractions under natural conditions, and chromium may be potentially more reactive than indicated in previous interpretations (ARCADIS, Inc., 2009, 2011).

In contrast to field sample collection and analyses of materials collected pre- and post- in-situ reduction, this study proposes controlled laboratory experiments to measure the fate of chromium within the IRZ. As part of these experiments, 6 sets of paired microcosms will be prepared from each of six homogenized samples of aquifer material selected to be representative of aquifer material within the IRZ (72 microcosms total). Selected materials will be described physically and characterized chemically using sequential extraction procedures (discussed in Task 2) to determine the range of naturally occurring chromium within the operational fractions defined by Wentzel and others (1991) and Chao and Sanzalone (1989). Extraction procedures will be modified to include an additional step to quantify Cr III sorbed through cation exchange.

Cr VI added to the experimental microcosms will be isotopically labeled using chromium-50 (50Cr) and chromium-54 (54Cr). 50Cr and 54Cr are naturally occurring stable isotopes of chromium having low crustal abundances of 4.3 and 2.3 percent, respectively. The labeled 50Cr/54Cr stable isotope ratios will be different from the natural ratios and can be used to trace the distribution of chromium on surface sorption sites within the microcosms for periods longer than 6 months.

Once prepared, labeled Cr VI within the microcosms will be reduced to Cr III to simulate removal of chromium from solution within the IRZ. After reduction, the samples will be placed on a shaker table and incubated at room temperature. One microcosm from each set will be fully capped to exclude atmospheric oxygen and maintain reduced conditions. The other microcosm will be loosely capped to permit entry of atmospheric oxygen and maintain oxic conditions. Both microcosms from each set will be harvested at selected intervals during a two-year incubation (times = 0, 3, 6, 12, and 24 months; the sixth set of microcosms will be held in reserved and analyzed if a longer incubation time is needed). The isotopically labeled chromium in various operational fractions will be sequentially extracted using procedures described previously in Task 2 (Wentzel and others, 1991; Chao and Sanzalone, 1989) and analyzed in duplicate using thermal-ionization mass spectroscopy.

Data will be evaluated to determine if chromium reduced within the microcosms is: 1) initially sorbed as Cr III, or directly incorporated into crystalline mineral oxides, and 2) to determine if the operational fraction of chromium on aquifer solids changes with time and becomes increasingly mineralized and potentially less mobile with time. Similar experiments on arsenic sorption to alluvium eroded from granitic rock in the Antelope Valley (Izbicki and others, 2012) showed that some arsenic, initially sorbed to exchange sites on the surfaces on mineral grains, was incorporated into less reactive amorphous and crystalline oxide minerals within one year. Precision of replicate analyses and recovery of labeled arsenic for that experiment ranged from 91 to 96 percent with a median recovery of labeled arsenic of 94 percent (Izbicki and others, 2012). Similar precision and recovery of labeled chromium is expected for this experiment.

Included for analyses within each time step will be five additional sets of two microcosms containing iron and manganese oxides precipitated on to an artificial substrate (either glass beads or glass slides). The oxide precipitates will consist of: 1) iron oxide, 2) three iron and manganese oxide mixtures bracketing the range of known proportions of these elements in natural oxide coatings, and 3) manganese oxide. The prepared artificial substrates will be incubated at room temperature on the shaker table with the other microcosms (one under reduced condition, one under oxic conditions) to serve as a control for aquifer materials. The artificial substrates will enable examination of isotopically labeled chromium reduced from Cr VI to Cr III and sorbed on to the oxide precipitate without the presence of naturally occurring chromium; thereby, permitting examination of potential reoxidation of Cr III to Cr VI through time.

The presence of Cr VI from reoxidation within the microcosms and on the artificial substrates will be measured using XANES spectroscopy. Detection of Cr VI using XANES would indicate reoxidation of chromium within the microcosms. Reoxidation can only occur in the oxic microcosm, and therefore the reduced microcosm will serve as a control for the experiment. Previous work failed to detect Cr VI in reduced alluvial material 9 months after in-situ reduction using XANES (ARCADIS, Inc., 2011). The use of material from controlled laboratory microcosms, including: 1) initial examination of materials, 2) use of artificial substrate that does not contain natural chromium, and 3) examination of materials at various time steps after oxidizing conditions are introduced (including time steps longer than 9 months), is intended to improve XANES resolution and detection of Cr VI. If necessary the sixth microcosm will be held for a longer time period to ensure sufficient time to measure reoxidation of Cr III to Cr VI.

Microcosms will be prepared, incubated, and analyzed in a U.S. Geological Survey laboratory in Menlo Park, CA under the direction of Larry Miller USGS, NRP. 50Cr and 54Cr analyses will be made by Thomas Bullen, USGS, NRP, Menlo Park, CA using thermal -ionization mass-spectrometry (TIMS, Finnigan MAT 261) with sample processing and purification procedures discussed in Task 3. XANES analyses will be done by Andrea Foster USGS, Menlo Park, CA.

Potentially reactive sorbed chromium and reoxidation of Cr III to Cr VI, if present within the microcosm experiments does not necessarily mean that in-situ reduction of Cr VI to Cr III within the aquifer is not a viable option or appropriate remediation technique for the site; however, reoxidation would suggest that in-situ reduction is not a permanent solution; and PG&E, regulators responsible for the site, and the community should be prepared to re-treat the site in the future.

Procedures for laboratory evaluation, quality assurance, and data management. Data for this study will be analyzed by a number of different highly specialized U.S. Geological Survey and contract laboratories In accordance with U.S. Geological Survey Policy (2006), performance data from laboratories used to analyze data as part of this study will be evaluated according to their ability to meet specific project needs as defined in this proposal. The laboratory evaluation is subject to approval by the Water Science Center Director and regular review to show the laboratory meets performance criteria throughout the study.

Approximately 10 percent of the analytical budget from tasks within this proposal has been reserved for quality control samples. Quality assurance will consist of a combination of replicate/duplicate analyses, blank samples of various types, and spike samples as appropriate. Higher frequency of data will be collected for analytes of higher importance to the project goals, and for analytes known to have high analytical variability on the basis of laboratory evaluation data. Lower frequency of quality assurance data will be collected for analytes of lesser importance to project goals, and for analytes know to have less analytical variability on the basis of laboratory evaluation data. Laboratory Evaluation Plans (LEPs) will be developed for each of the laboratories used within this study to ensure data from the laboratories will meet the quality objectives required for this study.

All data collected as part of this study will be publically available after internal project review and approval. All water chemistry and isotopic data collected as part of this study will be stored and archived in the U.S. Geological Survey National Water Information System (NWIS) data base, and will be available to the public through the USGS on-line data base NWIS-Web. Geophysical data including gravity data and geophysical log data will be archived according to U.S. Geological Survey policies and publically available. Analyze-Hole models will be archived in the USGS model archive. After review and approval data also will be transferred to the TWG to update project data maintained by the TWG in support of this study. Details for transfer to TWG and will be developed as the project moves forward: however, NWIS-Web and other U.S. Geological Survey databases are to remain the primary mechanism for public release of data.


A fact-sheet style Open-File report will be prepared early in Federal Fiscal Year 2014 to describe the study purpose and approach. A report will be prepared in FFY 2016 to describe selected preliminary results and how those results contribute to understanding of the movement of water and the occurrence of chromium in Hinkley Valley near the contamination plume. A final report will be prepared in FFY-2017 for review, and publication In FFY-2018. Due to the complexity of the study multiple reports describing important results may be required. The final report(s) will be accompanied by a fact-sheet style Open-File report describing the results of the study. Report findings will undergo USGS technical peer review prior to distribution to external stakeholders. Updates on study design, progress, and data interpretation will be provided periodically to the TWG, interested stakeholders within the community, and the LRWQCB as appropriate.


The study will require the near fulltime services of two Research Hydrologists (GS-14 and GS-12) familiar with trace element geochemistry in desert aquifers. In addition the study will require part-time services of 1) a Research Hydrologist and a Hydrologist familiar with groundwater flow modeling in desert aquifers (GS-13, and GS-12, respectively), and 2) part-time services of several Hydrologic Technicians familiar with groundwater-quality data collection, geophysical log collection, gravity data collection, description of geologic material, and data management (various GS levels ranging from GS-11 to GS-7). The project also will require part-time services of geospatial, reports, and fiscal support staff. These personnel are on staff within the San Diego Office of the California Water Science Center. The project also requires specialized services of U.S. Geological Survey research personnel within the Water and Geologic Disciplines. These personnel have been identified within the proposal and they have agreed to provide the services described.

Work Plan

The project will be done during Federal Fiscal Years 2014 through 2018. Data collection and interpretation will occur during Federal Fiscal Years 2014 through 2017. Interim reports will be prepared for release in 2014, and 2016. Preparation of a final report will begin in Federal Fiscal Year 2016 for release in 2018.

Cooperating Agency: Lahontan Regional Water Quality Control Board
Project Chief: John A. Izbicki
Phone: 619-225-6131
Email: jaizbick@usgs.gov

California Water Conditions

Real-Time California Streamflow Conditions