- Program Areas
- Installation Energy and Water
- Environmental Restoration
- Munitions Response
- Resource Conservation and Resiliency
- Weapons Systems and Platforms
Remediation of TNT and RDX in Groundwater Using Zero-Valent Iron Permeable Reactive Barriers and Zero-Valent Iron In Situ Treatment Wells
Dr. Rick Johnson | Oregon Health and Science University
ER-200223 - In Situ Treatment Wells
The effectiveness of zero-valent iron permeable reactive barriers (ZVI PRB) for removal of explosives from groundwater has been recently demonstrated. As an added benefit, ZVI can treat a variety of contaminants that may co-occur in groundwater at RDX- and TNT-impacted sites (e.g., chlorinated solvents, chromate). ZVI PRBs, however, are generally applicable in unconsolidated media and at depths of less than 70 feet below ground surface (bgs). To expand the applicability of ZVI to a broader range of settings, the injection of fine-grained ZVI into the subsurface has been proposed. An alternate approach, evaluated at the Cornhusker Army Ammunition Plant (CAAP) site near Grand Island, Nebraska, involved installation of a pair of dual-screened wells in which the conventional filter pack around the well screens has been replaced with coarse granular iron. Water moving into or out of the well screens passes through the ZVI and explosives are removed. The rate of explosives reaction with ZVI is sufficiently fast that good capture of groundwater by the wells can be achieved.
Objectives of the Demonstration
The objectives of this technology demonstration were:
- Demonstrate that TNT and RDX can be degraded in situ to acceptable levels (i.e., the method detection limit) using a zero-valent iron in situ treatment well (ZVI ISTW)
- Evaluate ISTW well pair hydraulics
- Identify design and operational factors that influence successful implementation and continued operation of the ZVI ISTW approach.
In this demonstration, granular iron was placed outside of the well screens in a pair of dual-screened wells. The wells were installed inside a large-diameter temporary casing such that the iron adjacent to the upper and lower screens could be isolated from one another in a manner that would promote groundwater circulation between the pair of dual-screened wells.
The groundwater hydrology of the well pair was evaluated and met design expectations. A primary concern in the methodology was that water entering the treatment zone would contain materials (e.g., sulfate) that would plug the treatment zone over time. Based on calculations for the site, this was not expected to be a problem, which turned out to be the case. Unfortunately, an unanticipated problem arose in which water moving out of the upper screen (which was located near the water table) became oxygenated and plugged the treatment zone. An additional problem with oxygen arose because of regionally decreasing water table elevations in the area of the demonstration. This exposed a portion of the iron adjacent to the upper screen and, together with drawdown during pumping, led to a loss of permeability in the well in which water was extracted from the upper screen. These problems could have been avoided if the screened interval was below the water table and if a packer had been placed in the well casing above the screen.
Tracer test data indicate that water recirculated between the two ISTW wells relatively quickly. Measurement of explosives concentrations in groundwater also showed that the performance of the ISTWs met design expectations. A year after installation of the ISTWs, reactivity of the iron was still sufficiently high to reduce explosives concentrations to below detection limits.
An inherent disadvantage of the design used at CAAP was that the iron could not be readily replaced. In retrospect, if the reactive material would have been emplaced as a removable “cartridge” within a large dual-screen well, it would have provided an opportunity to remedy the plugging issue. However, plugging of injection screens is an inherent problem with circulation wells, and it is difficult to say with confidence if the well design improvements discussed would represent a long-term solution.
The cost of each of the two ISTW wells was approximately $40,000. It is likely that modifications to the design would increase the cost per well. However, if the technology was implemented at a full scale in a similar setting, it is believed that the cost per well would be similar to well costs for this demonstration.
ER-200223 - Permeable Reactive Barriers
As the result of past practices, groundwater at many Department of Defense (DoD) facilities is contaminated with explosives such as TNT and RDX. Because of their chemical characteristics, these contaminants can persist for long periods in groundwater. Conventional treatment (i.e., pump and treat) often lasts for decades and still cannot completely remove these contaminants. In-situ destruction of the explosives using passive permeable reactive barriers (PRB) represents a cost-effective alternative to long-term pump and treat. In this project, a PRB filled with metallic (zero-valent [ZVI]) iron was used to intercept a TNT/RDX groundwater plume. A multidisciplinary team examined performance of the PRB in terms of contaminant destruction and its impact on the flow of groundwater.
The primary advantages of ZVI PRBs for groundwater remediation include: no aboveground remediation equipment required, rapid conversion of groundwater to reducing conditions in which explosives are degraded, low operation and maintenance costs, and long-lasting (more than 20 years) in-situ treatment. Their cost-effective use may be limited by the depth to groundwater and the ability to install the PRB in some geologic media. At sites without these physical constraints, the approach can be highly effective.
Objectives of the Demonstration
The objectives of this technology demonstration were:
- Demonstrate that TNT and RDX can be degraded in situ to acceptable levels (i.e., the method detection limit) using a zero-valent iron permeable reactive barrier (ZVI PRB)
- Evaluate barrier hydraulics
- Identify design and operational factors that influence successful implementation and continued operation of the ZVI PRB approach.
The demonstration was conducted at the Cornhusker Army Ammunition Plant near Grand Island, Nebraska. Groundwater at the site is at 15-20 feet below ground surface (bgs). The shallow aquifer consists of medium sands with some silty material. Groundwater velocity is approximately 1-2 feet/day.
The demonstration activities included a field study that involved installation of a mixed iron/sand permeable reactive barrier (30% by weight iron). The PRB was approximately 50 feet long by 15 feet deep by 3 feet thick. Monitoring activities were conducted over a 20-month period to evaluate performance of the PRB. The PRB was located within a large groundwater plume from a diffuse source resulting from production of munitions.
The ZVI PRB reduced concentrations of TNT and RDX to below detection limits throughout the duration of the project. In addition to removal of the explosives, significant changes in groundwater chemistry occurred due to the PRB. Dissolved sulfate concentrations decreased substantially as groundwater flowed through the PRB. Detailed groundwater concentration data and measured hydraulic conductivity data suggest that a portion of the water up-gradient of the PRB was diverted beneath the PRB. The reason for this is not entirely known; however, the project team believes it is not due to the nature of the contaminants being remediated (i.e., explosives) and it is almost certainly related to the use of guar during installation of the PRB. Guar could have entered the formation up-gradient of the PRB and not been fully removed at the completion of the installation. This may be the primary reason for flow reduction; however, it is also possible that the guar led to strongly reducing conditions just up-gradient of the PRB and the removal of sulfate as sulfide precipitates in the natural aquifer materials just up-gradient of the PRB, as observed in core samples.
All of the primary performance criteria for this project were met. TNT and RDX values were consistently reduced to below detection limits in the aquifer downgradient of the PRB. Barrier hydraulics were successfully characterized, and the project team was able to identify design and operational factors that influence successful implementation and continued operation of the ZVI PRB. The installation costs for the pilot-scale barrier were $138,000 with a cost per square foot of approximately $180/ft2. This is consistent with other demonstration-scale ZVI PRB installations. The greatest uncertainty in cost relates to the longevity of the PRB. If the PRB had been installed without guar, a 20-year lifetime could have been expected based on the observed hydraulic conductivity and reactivity of the PRB.
Points of Contact
Dr. Rick Johnson
Oregon Health and Science University
SERDP and ESTCP