The objective of this project was to quantify relationships between changes in the redox reactivity of abiotic reduced (or indirect biotically reduced) aquifer sediments and complex conductivity for four in situ reduction technologies in order to predict the longevity of permeable reactive barriers at field scale. These relationships are quantified at different scales starting with fundamental geochemical-geophysical relationships for single mineral phases at a millimeter scale, to one-dimensional (1D) columns (30 to 100 cm), to a two-dimensional (2D) laboratory system containing spatial heterogeneities (iron oxide inclusions), and finally to field scale in order to evaluate effects of spatial averaging. The influence of iron oxide minerals and mass in natural sediments on the resulting reduced zone performance will also be evaluated.

Specific research questions included the following:

• Can geochemical changes that occur (i.e., iron sulfide/oxide precipitation, adsorbed Fe²⁺) as a result of abiotic or bioreduction of minerals be quantifiable by changes in complex electrical conductivity?
• During reductant injection into aquifer sediment, how sensitive is electrical resistivity tomography, induced polarization (ERT/IP) for evaluating geochemical phase changes for four reduction technologies that produce different iron phase mass, mineralogy, and spatial distribution?
• During reduced sediment oxidation, how sensitive is ERT/IP for evaluating the longevity of the reduced zone reactivity?
• How significant is spatial averaging of geochemical reactions and ERT/IP at a particle scale and discontinuous iron oxide zone scale (10s of cm)?

Five tasks were used to evaluate the relationship between redox reactivity changes and complex conductivity of four in situ reduction technologies:

1. Single mineral phases (homogeneous) at a small (millimeter) scale (all 3 years)
2. Reactant injection in sediments with particle scale geochemical heterogeneities at a 1D column scale of 30 to 100 cm (year 1 and part of year 2)
3. Long-term barrier oxidation at a 1D column scale of 30 to 100 cm (year 2 and part of year 3)
4. Injection and oxidation in a 2D flow system with inclusions/layers (year 3)
5. Long-term performance at field scale (year 3)

In Task 2, because most technologies (except nanoscale zero-valent iron [nZVI]) rely on some aspect of the sediment mineralogy, the project team also correlated sediment properties (iron oxide mineralogy and concentration) with resulting ferrous iron precipitates. The significance of spatial averaging of contaminant reduction and also ERT/IP in systems with a) sub-particle scale mineral phase distributions (in Task 3.2) and b) larger scale (10s of cm) chemical heterogeneities (discontinuous iron oxide lenses, Task 4) was evaluated in modeling studies for geochemical spatial averaging and through the use of different electrode spacing for ERT/IP.

General SIP Imaginary Conductivity Response for: a) Ferrous Iron Phases or Microbes, and b) Reduction Technologies, Which Have One or More Ferrous Iron Phases

Geochemical-complex conductivity (i.e., spectral induced polarization [SIP]) characterization studies were conducted on the following:

• model sediments or sediment components (i.e., 2:1 clay, microbes, ferrihydrite-coated sand, goethite coated sand, sand),
• a Department of Defense (DoD) Umatilla aquifer sediment with reductive additions (i.e., adsorbed ferrous iron, mackaniwite FeS, pyrite FeS₂),
• Umatilla aquifer sediment with four different reduction technologies (Na-dithionite, Ca-polysulfide, bioreduction, nZVI), d) a DoD Ft. Lewis aquifer sediment with five different reduction fractions using Na-dithionite, and
• reduction with Na-dithionite of a sediments varying in iron and clay content (i.e., Cloudland, GA, Ocala, FL, Kenoma, KS, Norborne, MD, Hanford, WA, and Mt Simon, IL).

These studies were conducted to quantify correlations between increasing ferrous iron in different phases and SIP imaginary conductivity. Geochemical characterization of the reducing environment was characterized by: a) iron extractions (directly identifying surface phases that act as electron donors), and b) probe compound reduction rates (i.e., chromate, iodate, nitrate, and RDX), which provide a measure of the redox capacity and potential). The SIP characterization included measurement of complex conductivity (i.e., conductivity and phase shift over a frequency range from 0.001 to 10,000 Hz) for untreated, partially reduced, fully reduced, and at differing stages as reduced sediment was oxidized.

SIP Imaginary Conductivity Changes for Reduction of Sediments or Phases

The frequency-specific SIP imaginary conductivity changes for all of the sediment components, model sediments, additions, and reduction technology modifications to the Umatilla aquifer sediment, and other sediments are shown in the figure below. Diagrams in this figure are generalized response based on SIP data. The SIP imaginary conductivity changes identified in the Umatilla sediment (high iron oxide content and low clay content) that correlated well with specific ferrous iron phases are frequency specific, as the addition of iron sulfides to sediment generally decreased the imaginary conductivity in the 0.001 to 0.01 Hz range, and the addition of adsorbed ferrous iron generally increased the imaginary conductivity in the 0.01 to 10 Hz range (Figure 60a). In some cases, the addition of adsorbed ferrous iron increased the imaginary conductivity in the 0.001 to 0.01 Hz range, counteracting the decrease in imaginary conductivity due to iron sulfide addition. The increase in imaginary conductivity for the adsorbed ferrous iron was expected (i.e., adsorbed ferrous iron acts to increase the capacitance of the sediment or temporary storage of electrons), but the decrease in imaginary conductivity for the iron sulfides was not expected. However, multiple data sets confirmed these results: a) five different iron sulfide additions to Umatilla sediment, b) iron sulfide additions to silica sand, c) iron sulfide additions to Hanford sediment, and d) Ca-polysulfide treatment of Umatilla sediment (with iron extractions showing a large FeS increase). Microbial biomass resulted in an increase in imaginary conductivity in the 1 to 100 Hz range (13, 25), which overlaps with the increase as a result of adsorbed ferrous iron. A 2:1 smectite clay (montmorillonite, SWy-1) with adsorbed ferrous iron removed that was chemically reduced with Na-dithionite resulted in reduction of structural iron in the clay ‎(22). The SIP response to reduction of this clay was an increase in imaginary conductivity in the 0.001 to 0.1 Hz range (see table below).

The SIP imaginary conductivity response identified for separate components or these components added to a sediment generally can be used to predict the SIP imaginary conductivity behavior for differing technology reductions of the Umatilla sediment and to some extent, other sediments.

In order to use the results of this study at field scale, the effects of spatial heterogeneities needs to be evaluated. Spatial heterogeneities of mineral types and size commonly occur in natural systems at different scales. The influence of one type of spatial heterogeneity (discrete iron oxide zones) was evaluated to quantify the effects of geochemical and SIP spatial averaging. Results from a 2D experimental system which contained reduced iron oxide inclusions showed that a homogeneous equivalent model (which incorporated only the mass of the Fe oxide zones, not location or shape) could predict nitrate and iodate reduction, but not chromate reduction behavior. SIP data collected at different times during the 2D experiment would be used to predict the spatially averaged iron oxide content. In addition, inclusion prediction was found to be dependent on: a) having low pore fluid conductivity, b) having a higher contrast between the inclusion and matrix porous media, c) SIP electrode spacing, and d) SIP electrode survey type.

Overall, to predict reduction reactivity with complex conductivity (ERT/IP) in the field, these correlations between imaginary conductivity and geochemical properties (contaminant reduction and iron surface phases) are useful, as frequencies that can be used in the field are 0.001 to 100 Hz. Prediction of ferrous iron mass changes (or contaminant reduction rates) based on the imaginary conductivity changes at a new site would require laboratory experiments to develop calibration curves between adsorbed Fe²⁺/iron sulfide additions and imaginary conductivity response. Results to date have also shown that spatial variability of ferrous iron phases has significant influence on the geochemical reactivity and could be predicted from spatially averaged ERT/IP data for specific cases. (Project Completion - 2022)