Residues of 2,4,6-Trinitrotoluene (TNT), Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX), and Octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) can accumulate on military ranges and open burn/open detonation (OB/OD) areas, and can pose a risk to groundwater because these high explosive (HE) compounds are moderately soluble in water and bind weakly to soil particles.  Under certain conditions, HEs also can be degraded in soils, biologically or abiotically, limiting any transport to groundwater.

This project evaluated the transport and attenuation of TNT, RDX, and HMX in variably saturated soils at an active hand grenade range (RG40) located at Fort Bragg, NC. Two different management approaches were evaluated in adjoining grenade throwing bays: (a) monitored natural attenuation (MNA); and (b) enhanced attenuation (EA). MNA relies on the use of natural attenuation processes to control off-range migration of HEs to environmentally acceptable levels. MNA is most effective in fine textured soils that occasionally become saturated during high rainfall periods, reducing oxygen transfer through gas filled soil pores. EA involves addition of soluble, biodegradable organic substrates that increase oxygen consumption, generating reducing conditions to stimulate anaerobic biodegradation of HEs. 

The soils in the two throwing bays differed in their capacities for natural attenuation. Much of the original surface soil had been removed to form containment berms, leaving subsoils with somewhat different properties in the two bays. The soils in the two bays had similar textures and pH values, but the organic carbon (OC) and clay contents were significantly higher in Bay C than in Bay T, and as a result, the potential for natural attenuation was greater in Bay C. The field testing therefore evaluated MNA in the soils from Bay C and EA in Bay T (enhanced through additions of glycerin and lignosulfonate). 

Laboratory studies included screening of potential amendments, followed by microcosm and column tests of natural and enhanced degradation. The initial laboratory studies were conducted to identify inexpensive organic materials that could be spray applied on ranges and could be transported 0.5-1.0 m into the soil profile. Glycerin (GL) and lignosulfonate (LS) were selected for further testing. The two were combined because GL was degraded quickly, causing a rapid onset of anoxic conditions, while the LS sorbed to soil and was biodegraded more slowly, providing a longer-term enhancement of TNT, RDX, and HMX biodegradation.

TNT was extensively degraded in both batch microcosm and column experiments, consistent with prior reports. In microcosm experiments with field sand and soil from both bays, TNT rapidly biodegraded under aerobic conditions with transient increases in the concentrations of the amino-dinitrotoluene (ADNT) degradation products, 2-ADNT and 4-ADNT.  In the column testing, less than 0.1 % of the added TNT was discharged in the effluent from either the control or the GL+LS amended columns, and there was no accumulation of any TNT degradation products in either column. However, the fraction of TNT sorbed to soil was much lower in the amended columns, indicating GL+LS addition enhanced TNT degradation. 

There was no evidence of RDX or HMX biodegradation in aerobic microcosms, but both were rapidly degraded under anaerobic conditions, with somewhat slower degradation in the Bay C soil. Addition of GL and/or LS significantly increased RDX degradation rates in Bay C soil, but did not further increase degradation rates in the Bay T soils. During anaerobic RDX degradation, the aqueous concentrations of dissolved nitroso derivatives (Hexahydro-1-nitroso-3,5-dinitro-1,3,5-triazine [MNX], Hexahydro-1,3-dinitroso-5-nitro-1,3,5-triazine [DNX] and TNX) increased initially and then declined. The variable saturation column experiments demonstrated that natural attenuation of RDX was possible under ambient conditions. However, the results from the different bays showed that relatively small differences in soil properties can result in large differences in RDX removal.  Effluent RDX concentrations were much higher in the un-amended Bay T columns, indicating substantial RDX leaching, despite evidence that RDX was degraded in the columns, even when oxygen was present in the soil gas. On the other hand, there was extensive degradation of RDX in the control Bay C columns, probably because the higher clay and organic carbon contents resulted in much more of the column distance being anoxic. 

Addition of the GL+LS amendment stimulated RDX degradation in columns from both the Bay T and Bay C columns. In the replicate Bay T columns, RDX was degraded faster in both columns (86 and 88% removal) than in the untreated controls (64 and 77% removal), although one of the replicates was more aerobic than the other, and therefore had higher effluent RDX concentrations. In the Bay C columns, GL+LS addition also reduced the amount of RDX and RDX degradation products remaining in the soil at the end experiments. However, treatment with GL+LS did not substantially reduce RDX leaching from the Bay C columns, because leaching was already very low in the untreated Bay C controls. 

Amendment addition did not have a significant impact on HMX leaching or on the final HMX concentrations in the soil, for either the Bay C or Bay T soils. The limited HMX removal is probably due to high RDX concentrations which are reported to inhibit biodegradation of HMX. 

Field testing included evaluation of MNA in Bay C and EA in Bay T. The results differ for the three explosive compounds. For TNT, natural attenuation was protective in both soils, with no TNT accumulation in surface soils or soil pore water in either Grenade Bay T or C, even though both bays have been used regularly for training for over 20 years. These results are not surprising, as natural attenuation of TNT in soils appears to be common.

There was also strong evidence for natural attenuation of RDX in Bay C. At the start of the project, concentrations of RDX were relatively high in both pore water (1 to 454 µg/L) and soil (3 to 100 µg/Kg), likely due to a combination of intensive training and lower than average rainfall. During the low rainfall period, the low soil moisture probably resulted in more oxidizing conditions, with accumulations of nitrate and RDX. In June 2013, a period of high rainfall combined with warm summer temperatures resulted in reducing conditions (low Eh and an increase in dissolved Mn), and a rapid decline in both nitrate and RDX concentrations. With the return of drier conditions and lower temperatures, soil Eh increased quickly, but nitrate and RDX levels recovered much more slowly. 

Effective control of RDX in Bay T soil required EA. As in Bay C, the RDX concentrations in Bay T were elevated initially, but the higher permeability of the Bay T soils allowed more rapid drainage even during wet periods, maintaining oxidizing conditions and limiting any RDX degradation. Adding the GL+LS amendment caused Total Organic Carbon (TOC) concentrations in the soil water to increase after several months, causing declines in Eh and NO₃ and an increase in the dissolved Mn, followed by a rapid decrease in the RDX concentration. At 500 days after amendment application, TOC declined, causing Eh and NO₃ to increase again and dissolved Mn levels to decrease. At two years after amendment application, RDX concentrations also showed some evidence of rebound. 

Conceptual Model of Explosives Attenuation

This project has improved the conceptual model of the major factors that control leaching of HEs through the vadose zone. This conceptual model includes: (a) the amount of HE deposited by low order detonations; (b) the chemical properties of the individual HE compounds, which control their aqueous solubility, sorption to soil and biodegradability; and (c) the rate of oxygen transport and consumption in the soil. Oxygen transport in soil is controlled by the soil physical properties (permeability and water retention characteristics), precipitation and evapotranspiration. Oxygen consumption is controlled by the amount and type of organic carbon and soil temperature.

This conceptual model can help managers determine when MNA might be appropriate, and when enhancements may be needed. In most cases, RDX is the critical contaminant of concern for groundwater protection because TNT will naturally attenuate in the vadose zone at most sites, due to the ease of biodegradation under both aerobic and anaerobic conditions. Natural attenuation of RDX is more variable, however, due to its resistance to biodegradation under aerobic conditions.

At some sites (e.g., Fort Bragg Grenade Bay C), the soil can undergo extended anoxic periods, leading to extensive RDX removal and limited leaching to groundwater. These anoxic conditions may not need to be continuous or occur throughout the soil profile. At many ranges, it may take several years for RDX to migrate through the upper two meters of soil due to slow infiltration and sorption to soil, so if a significant volume of the soil becomes anaerobic for even a few weeks during this period, much of the RDX can be degraded, greatly reducing the risks of groundwater contamination. 

However, in soils with lower fines and/or organic content (e.g., Fort Bragg Grenade Bay T), the soils can remain generally aerobic, resulting in less degradation and greater RDX leaching to groundwater, and EA may be needed. Organic amendment additions can be effective in reducing RDX leaching by inducing anaerobic conditions at least during wetter periods.  Given the relatively long travel time for munitions through soils, it may not be necessary to maintain continuously anaerobic conditions to protect groundwater, and periodic amendment applications could provide effective long-term control.

The lessons learned from this project are important for managers of many sites where explosive compounds have been used. Key lessons include:

1. Natural attenuation processes may be more important in reducing RDX leaching than previously assumed. 
2. Spray application of the organic amendments (GL+LS) on the soil surface is an effective and cost-effective method to generate reducing conditions and thereby reduce RDX leaching.
3. There can be large spatial variations in soil drainage rates between different bays, and these differences can have a dramatic impact on the amount of RDX transported to groundwater.
4. The geochemical conditions controlling RDX attenuation can vary tremendously over both space and time. These large variations will make it much more difficult to predict when MNA processes are sufficient to prevent RDX leaching and when EA will be required.
5. There are important technical challenges to using organic amendments to enhance RDX degradation. High amendment loading rates will be needed in permeable soils with rapid oxygen transfer. Also, amendments can be washed off the soil surface by heavy rains or migrate very slowly into low-permeability soils. 
6. Tilling the amendments into the soil during routine maintenance for UXO clearance and/or crater smoothing would likely be more effective than the spray application approach employed in this project. Another effective strategy can be to inject the amendment a few inches below the soil surface using injection nozzles like that used for land application of biosolids.
7. Periodic reinjections will probably be needed if relying on EA. RDX degradation can be delayed for several months after applying amendments, until after the TOC increases and Eh decreases. Once TOC declines and Eh rebounds, dissolved RDX concentrations can remain low for several months.