All boreholes must be properly sealed to prevent water from infiltrating. This is usually done with a bentonite/cement mixture but bentonite pellets like those shown here can be easier to work with when allowed by the regulators.

There are a number of site-specific factors – such as geochemical and lithological conditions– that can negatively affect the performance of in situ groundwater remediation efforts and sometimes yield unexpected, secondary effects. While working under a limited budget, it may be hard to plan for every aspect of an injection job. But proper planning and developing a thorough understanding of a site prior to implementing an in situ remedy can significantly increase the chance of success.

Toward this end, this summary presents some of the lessons learned from hundreds of field sites using EHC, (a patented combination of carbon plus zero-valent iron; ZVI) used for in situ chemical reduction of chlorinated solvents (CVOCs) and immobilization of heavy metals. These lessons are also applicable on a more general level to the use of other amendments. An overview of some relevant factors in each of several potentially troubling in-situ injection scenarios will be discussed; including transient observation of compounds related to the injection (e.g., acetone and MEK), unexpected daughter products of CVOC degradation, mobilization of metals sorbed to or present in the soil matrix, concentration spikes after an injection, and daylighting/short-circuiting of amendments during injection. We also present some thoughts on how to manage these issues.

Unexpected transient catabolites from CVOC degradation

Conventional catabolites generated from sequential reductive dehalogenation reactions are usually anticipated and monitored. However, we have occasionally come across other metabolites generated from less common pathways. For example:

• Carbon disulfide (CS2) may be formed during reductive dechlorination of carbon tetrachloride (CCl4), especially under sulfate reducing conditions.^[1] We observed CS2 at a peak concentration of 79 µg/L when treating a site in Kansas with CCl4 concentrations in excess of 2 mg/L. The CS2 concentrations quickly tapered off within a few months after EHC injections, and they have remained below criteria (generally not detected) for over three years.
• The formation of lower concentrations of chlorinated ethanes (DCA and CA) has been observed at two sites where EHC was injected to treat chlorinated ethenes (PCE/TCE). While the chlorinated ethanes formed made up no more than approximately 1 percent of the starting PCE/TCE concentrations in both of these instances, the concentrations of DCA and CA exceeded MCLs temporarily, before being further degraded to ethane. The pathway is hypothesized to be hydrogenation of DCE and VC to form DCA and CA respectively (addition of two hydrogen atoms and converting the carbon double bond to a single bond).

The viscosity of the slurry must be adjusted to fit the site conditions. This site required a slurry that was about 30 wt percent.

Addition of a carbon substrate to an aquifer can result in the production of MEK (i.e., 2-butanone) and acetone by indigenous microorganisms (a presentation from U.S. Naval archives, available at www.navylabs.navy.mil/Archive/emdq-2005/2-3-1.pdf(pdf), provides good evidence and references for this effect). Field data showing the occasional, transient presence of acetone and/or MEK seem to occur when alkanes are present, along with high-organic carbon levels, in sub-oxic, especially methanogenic environments. It therefore is assumed that MEK/acetone production may result before the system goes fully anoxic, or soon after injection of a carbon source. The production of MEK and acetone can be quite substantial, but it is a short-lived phenomenon. Acetone also has been observed to be produced in samples stored using standard preservation techniques.

Metals mobilization

With any in situ treatment approach that alters a site's geochemistry, there always is the risk of mobilizing metals sorbed to, or present in, the soil matrix. For example, conventional reductive anaerobic dechlorination reactions may result in a release of metals bound within iron oxides, manganese oxides or other clay minerals in the aquifer material. Examples of such metals include As, Cr, Se, Fe, and Mn.^[2] This process may occur via reductive dissolution or desorption from those mineral phases, as a result of changes to pH and redox potentials.

EHC maintains a near-neutral pH, as the acidity of carbon substrate degradation is offset by the alkalinity resulting from ZVI corrosion. This minimizes some of the metal liberation effects relative to alternative carbon-only substrates (such as oils or simple hydrogen releasing compounds). Hence, one mechanism for heavy metals to be sequestered through the use of EHC is the complexity and effects of iron phases created in the treated matrix. The presence of ZVI in the agent enables metal adsorption and co-precipitation through a continuous generation of dissolved iron and subsequent formation of iron oxides, oxyhydroxides and sulfides. The long-term stability of metals, including As, in the presence of ZVI has been demonstrated in previous extensive research on permeable reactive barrier (PRB) applications ^[3,4].

For sites where mobilization of metals is of specific concern, a formulation of EHC is used that includes a source of sulfur to promote precipitation as metals sulfides. It is therefore recommended to consider the potential for metals mobilization beforehand, and measure natural sulfate levels to evaluate the need for an additional source of sulfur to immobilize metal cations under reducing conditions.

Concentration spikes after injection

Concentration spikes following injection can result from various processes. For example, Figure 1 shows data from a site where seasonal groundwater table fluctuations resulted in the flushing of compounds from the smear zone. Other factors that can result in concentration spikes include:

• Dissolution of residual NAPL (if present) or flushing of untreated source zones (e.g. inaccessible or unmonitored source areas) can cause concentration spikes.
• Pore volume (PV) displacement for some amendments – Depending on injectate volume and aquifer porosity, water level rises may occur during injection, thereby flushing compounds from unsaturated soils. This may also cause artificial concentration drops (i.e., perceived "treatment" via dissipation / dilution) if plume displacement is occurring [Author note: we have not seen water table increases with EHC, as we typically only displace 2 to 10 percent of the PV].
• Compound degradation chemistry – Increases in daughter products of many organic compounds can be observed immediately following an injection before low redox conditions are established. However, EHC established in situ chemical reduction (ISCR) conditions and the accumulation of typical intermediates is avoided under these conditions^[5], as shown in Figure 2.

Figure 1: Total CVOC concentrations and depth to groundwater measured at four wells within an EHC treatment area. A general increase in the CVOC concentrations were observed during the first two post-injection monitoring events, conducted in September 2006 and January 2007. However, historical and background data suggests that this increase was related to introduction of CVOCs from the smear zone in relation to an increase in the water table. Subsequent sampling events showed smaller rebounds during the rainy season (high groundwater table).

Daylighting and short-circuiting

Daylighting is the surfacing of injectate, which can occur when injecting any type of amendment under pressure and must be anticipated during an injection job. The following has been found to be helpful in minimizing potential daylighting:

• Minimize injected volume – typical source area injection volumes of 2 to 10 percent of PV are used. In some PRB (non-source area) installations, a maximum of 20 percent of PV is displaced. Many other injectates require significantly more PV displacement.
• Reduce pressure and flow-rate – these injection parameters are site-specific, but the lowest pressure for a sustainable low-flow rate (<10 gpm) seems generally best.
• Ideal borehole abandonment – slurry grouts are unacceptable, unless time is allowed for setup. Ideally, some form of bentonite pellets, such as AquaBlok's HoleBlok+ product could be used. If historical boreholes were not properly abandoned, they will likely become exit points during the injection process.
• Ideal viscosity – field experience shows that a slurry containing ~30 wt percent EHC is generally optimum for injecting via direct push technologies. Again, site-specific results may vary and thicker slurries have been applied for shallow applications (from 2 feet bgs) to limit surfacing of the slurry around the rods.
• Top-down vs. bottom-up injection protocol – use an injection tip that allows for flexibility in the field, such as hydro-punch pressure-activated injection tooling. This will reduce the potential for pooling the amendment in the bottom of the injection hole.
• Allow pressure dissipation – Use of multiple rods and moving around the injection grid help to enable localized subsurface pressures to dissipate.
• Finally, it is recommended to use an injection contractor experienced with and prepared to work over a range of pressures, flow-rates, and installation approaches. It is best to plan for a flexible implementation. PE

Figure 2: Effect of EHC and ZVI on 1,2-DCA and CA in groundwater measured under flow-through conditions. Treatment performance improve with time as reducing conditions are being established.

References:

1. Devlin J.F. and Muller, D., "Field and Laboratory Studies of Carbon Tetrachloride Transformation in a Sandy Aquifer under Sulfate Reducing Conditions," in Environmental Science and Technology, 1999, v. 33, p. 1021-1027.

2. AFCEE, 2004. Principles and Practices of Enhanced Anaerobic Bioremediation of Chlorinated Solvents. Dzombak, D.A. and Morel, F.M.M., 1990. Surface Complexation Modeling: Hydrous Ferric Oxide. New York: Wiley Science

3. Manning, B. A., M. L. Hunt, C. Amrhein, and J. A. Yarmoff. "Arsenic(III) and Aresenic(V) reactions with zerovalent iron corrosion products" in Environmental Science and Technology, 2002, v. 36, p. 5455-5461.

4. Wilkin, R.T. and McNeil, M.S. "Laboratory evaluation of zero-valent iron to treat water impacted by acid mine drainage" in Chemosphere, v. 53, p. 715-725.

5. Brown, R.A., J.G. Mueller, A.G. Seech, J.K. Henderson, J.T. Wilson. 2009. "Interactions Between Biological and Abiotic Pathways in the Reduction of Chlorinated Solvents." REMEDIATION Winter 2009, pages 9-20.