Groundwater contamination due to explosives such as TNT, RDX, and HMX is a problem at many Army ammunition plants. This ESTCP project evaluated the ability of selected aquatic macrophytes (aquatic plants) in constructed wetlands to clean up explosives-contaminated groundwater through a phytoremediation process. A field demonstration of two types of artificial wetlands was conducted at Milan Army Ammunition Plant (MAAP) in Tennessee. The first type was a two-celled lagoon-based wetland containing submergent plants to test the concept of explosive degradation via nitroreductase enzyme produced by the plants. The second type was a two-celled gravel-based wetland planted with emergent wetland plants to test the concept of explosive degradation via both microbial and plant enzyme processes. A variety of submergent and emergent aquatic macrophytes were screened for their ability to remediate the contaminated water.

Each wetland received 5 gpm contaminated groundwater containing 3,250 ppb of nitrobodies (including TNT, RDX, HMX, TNB, 2AA-DNT, 4A-DNT) from June to November 1996, and 9,200 ppb from November 1996 to August 1997. While both systems could remove explosives, the gravel-based system was clearly superior. The gravel-based wetland remediated groundwater contaminated with a variety of explosives, including both TNT and RDX, and was able to reduce TNT concentrations to less than 2 ppb and total nitrobody concentrations to less than 50 ppb during all but the coldest months. During winter operations, the gravel-based system had difficulty meeting the total nitrobody reduction goals due to reduced microbial activity and a decrease in treatment efficiencies at low water temperatures. In contrast, the lagoon-based wetland was unable to satisfactorily remove RDX and HMX, or meet the total nitrobody removal goals, and was only able to meet the TNT reduction goal of 2 ppb during the initial stages of the demonstration.

A gravel-based system can be economically resized to overcome the winter performance issues. Assuming a 95% system availability and 30-year life, the total cost for treating 200 gpm of MAAP groundwater with a gravel-based system was estimated at $1.78 (surface water discharge) or $2.06 (groundwater reinjection discharge) per thousand gallons of groundwater. This includes capital costs estimated at $3.465 million for surface water discharge and $4.125 million for groundwater reinjection discharge.

Wetland phytoremediation is an environmentally friendly technology that is widely accepted, relatively self-sustaining, and cost-effective to maintain. In addition, unlike granular activated carbon (GAC), this technology does not produce secondary waste streams. The plants appear to be metabolizing the explosives over time, thus keeping the level of explosives or metabolites to a low level in plant tissues. As a result, the explosive concentrations in the plants should not pose any adverse environmental effects. (Project Completed - 1999)