As of 2005, the Department of Defense (DoD) has identified nearly 6,000 sites at its facilities that require groundwater remediation and has invested $20 billion for site cleanup over a 10-year period. At many of these sites, unsaturated chlorinated volatile organic compound (CVOC) source zones located above the water table are producing and sustaining groundwater plumes. Many of these unsaturated sources are currently being treated with soil vapor extraction (SVE) technologies. Long-term SVE projects can be very costly, as the treatment process for the recovered vapors is expensive.

The overall objective of this project was to demonstrate hydrogen-based treatment (H₂T) as a remediation technology for the unsaturated zone, either as the initial remediation technology applied at a site or as a polishing technology that will allow DoD site managers to shut down an existing expensive, low performance SVE system, but where monitored natural attenuation may not be sufficient to control the groundwater plume that is sourced by the residual contaminants in the unsaturated zone. With such a technology, the cost for remediating these groundwater plumes can be greatly reduced, and a much more sustainable remedy can be implemented. This demonstration answers key questions about the performance, implementability, and cost of the technology.

In the H₂T system, a mixture of nitrogen, hydrogen, propane, and carbon dioxide gases are injected into an unsaturated treatment zone through a series of widely spaced injection points to degrade chlorinated organic compounds. Nitrogen serves as a non-explosive carrier gas to flush oxygen from the soil gas, enhancing conditions for the anaerobic degradation of chlorinated solvents. Propane is used as an inexpensive electron donor for scavenging oxygen (i.e., naturally occurring aerobic bacteria will use the propane to remove oxygen). Hydrogen is used as the electron donor for dechlorinating bacteria to stimulate biodegradation of chlorinated organic compounds, forming innocuous daughter products such as ethane or ethene. Nitrogen and hydrogen can be purchased and delivered to the site (which are refilled or changed out regularly by the gas provider as part of the gas delivery contract) or generated on site depending on the size of the H2T application (i.e., total flowrate and treatment time). The stoichiometry of the dechlorination reaction indicates that for every 1 mg of hydrogen utilized by dechlorinating bacteria, 21 mg of tetrachloroethene (PCE) can be completely converted to ethene.

In the unsaturated zone, the H2T process relies on a gas injection skid consisting of piping, gages, safety equipment, a process control system, and gas supply vessels that could connect to a piping manifold and injection wells at the site. At some sites, one advantageous configuration could be the conversion of a low-performance SVE system to H₂T, where the existing SVE blower and treatment system is decommissioned and replaced by the H₂T injection skid connected to the existing manifold and injection wells.

Over the 6-month test, a total of 830,000 standard cubic feet of gas was injected into a fine-grained vadose zone at a former missile silo site in Nebraska with the following average composition: 10% hydrogen, 79% nitrogen, 10% propane, and 1% carbon dioxide. The hydrogen gas was designed to stimulate biodegradation of the chlorinated solvent contaminants that persisted in this zone even after 3 years of SVE. Because of inconclusive sampling results during the test, the total gas flow rate and hydrogen composition were doubled for the last month of the injection phase (2.5 scfm to 5.0 scfm and 10% to 20%, respectively). A subsequent increase in hydrogen and propane concentrations and decrease in oxygen concentrations were observed.

The molar mass of chlorinated compounds was unchanged (7.1 moles before vs. 7.1 moles after). Therefore, while the system was successful at converting TCE, a cis-DCE stall condition appeared to be present at the site. Key conclusions from the test:

• The H₂T process removed half the TCE from the test zone that was remaining after this zone had been treated with SVE for 3 years. This indicates the process may be effective for treating finer-grained units that are difficult to treat with SVE.

• In-test measurements of redox-related parameters (oxygen, methane) indicated that deeply anaerobic conditions were not achieved uniformly through the test zone, a likely contributing factor for the observed cis-DCE stall. For example, the average oxygen content in the treatment zone soils ranged from 0.1% to 11%, indicating partial anaerobic conditions for most of the treatment zone.

• Laboratory microcosm work where the gas mixture was added to soil samples from the site indicated that samples that had been bioaugmented with dechlorinating bacteria performed much better than unamended soils. Additional microcosm results indicated that low moisture may have been a contributing factor to this bacterial limitation. The research team concluded that the system’s inability to create deeply anaerobic conditions was likely a major factor in the cis-DCE stall.

• It is possible to safely inject the hydrogen, nitrogen, propane, carbon dioxide gas mixture in the test zone. The radius of influence (ROI) from the injection point was approximately 15 feet.
• In-test vapor VOC monitoring data were not helpful in evaluating the progress of remediation.

The demonstrated H₂T system was more successful than the existing SVE system at removing TCE from the fine-grained soils at this test site, but it was not successful at removing a significant fraction of the cis-DCE. To help drive a full-scale H2T treatment zone to deeply anaerobic conditions, some type of barriers over the top and around the sides of the treatment zone (even something as simple as adding water to reduce the gas permeability of the soils) might help break out of a cis-DCE stall condition.

Key H₂T implementation issues are summarized below.

• Specific permits for H₂T may be required by local codes and will include drilling, well installation permits, and hazardous materials storage permits. Other permits may be necessary and will be dependent on local codes.

• One of the main safety concerns associated with H₂T application is the flammability of hydrogen and LPG and the potential production of methane gas. It was shown in this demonstration that the safety concerns could be addressed easily by following the safety codes (e.g., NFPA50A, NFPA55, etc.).
• Soil permeability, heterogeneity, and moisture can greatly affect the performance of an H₂T system.
• A suitable population of dechlorinating organisms (Dehalococcoides) is needed to ensure complete conversion of PCE or TCE to non-toxic products (e.g., ethane).
• Practitioners considering this alternative should first conduct laboratory testing with and without bioaugmentation to assess whether there is a microbial limitation and then measure the electron acceptor influxes to estimate initial and sustained hydrogen demand, before designing and implementing a field pilot test.

The unit cost for a full-scale H2T system (assumed to be about 50,000 cubic yards) is projected to be $49 per cubic yard. This would compare to the following costs per cubic yards: $37 for a new-build soil vapor extraction system; $20 to keep an existing SVE system in operation for another two years; and $97 for excavation. Sensitivity analyses were performed to evaluate the effect of gas flowrate and ROI on the unit cost of H2T implementation. It was concluded that while the cost of H2T was greater than SVE system operation, the decision to switch to H2T operation over an SVE system should be made based on the overall performance and not only on the cost assessments.