This demonstration installed a passive subsurface biobarrier to treat the nitramine explosives hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX), octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX), and comingled perchlorate in groundwater at an operational military range. The optimal areas for application of this technology include open burn/open detonation (OB/OD) sites, munitions test ranges, explosive ordnance disposal (EOD) training areas, target areas, munitions disposal sites, and other regions where high concentrations of munitions constituents are likely to occur. The Churchill Range in the Explosives Experimental Area (EEA) of the Naval Surface Warfare Center Dahlgren Division (NSWCDD) in Dahlgren, VA, was chosen as the demonstration site. The barrier, which was placed downgradient of a location where testing activities occur at NSWCDD, consisted of an emulsified oil substrate (EOS) and buffer applied to the subsurface. The buffer was used to increase the natural acidic pH of the groundwater and maintain that pH in the circumneutral range to promote rapid biodegradation kinetics of the target contaminants. Important to the mission of operational U.S. Department of Defense (DoD) ranges, the barrier had no surface structure and no significant impact on typical range activities. A key objective of this demonstration was to limit further contaminant migration in groundwater cost-effectively, with minimal impact to range activities.
When injected into a groundwater aquifer, EOS promotes the growth of indigenous bacteria capable of anaerobically biodegrading perchlorate, RDX, and HMX to low concentrations. The effectiveness of the barrier for reducing migration of perchlorate and explosives in groundwater at the EEA of NSWCDD was determined using a series of groundwater monitoring wells (MWs). The 100-ft-long biobarrier was installed by injecting emulsified oil and buffer through a series of 20 injection wells (IWs) placed cross-gradient to groundwater flow. Two injection events were conducted over time. Upgradient, in-barrier, and downgradient wells were monitored for perchlorate, RDX, HMX (and nitroso intermediates of RDX and HMX), field parameters, total organic carbon (TOC, as a measure of oil concentration), fatty acids, dissolved metals, anions, and field parameters for approximately 30 months after the initial emulsified oil injection. The remedial approach was designed to treat the contaminants in the ground with no surface structure and to minimize impacts to ongoing range activities.
This barrier promoted the rapid in situ biodegradation of perchlorate, RDX, and HMX (Figure E1). Upon emulsified oil injection, RDX concentrations decreased significantly downgradient of the biobarrier, with a degradation “front” slowly moving down the centerline of the plot. The RDX removal averaged 83 ± 17% for the in-barrier wells and 75 ± 21% for the centerline wells from the first emulsified oil injection to the end of the demonstration. However, these averages included periods of time when the TOC from the emulsified oil injection(s) was depleted leading to increased RDX in downgradient wells. When TOC from emulsified oil or its degradation products was adequate, and time was allowed for degradation to occur, RDX concentrations reached extremely low levels in the centerline wells. For example, approximately eight months after the initial oil injection, the RDX within the barrier to a distance of 30 feet (ft) downgradient ranged from <0.03 to 6 micrograms per liter (µg/L). RDX removal in these wells was >94%. Similarly, ten months after a second emulsified oil injection, RDX concentrations along the centerline wells ranged from <0.03 µg/L (5/7 wells) to 2 µg/L (2 wells) as far as 40 ft downgradient of the barrier, with removal percentages >98% over this large distance. Thus, this technology was highly effective for promoting RDX biodegradation when adequate TOC and appropriate biogeochemical conditions were achieved.
The RDX metabolites hexahydro-1-nitroso-3,5-dinitro-1,3,5-triazine (MNX), hexahydro-1,3-dinitroso-5-nitro-1,3,5-triazine (DNX), and hexahydro-1,3,5-trinitroso-1,3,5-triazine (TNX) increased as RDX degraded in response to the initial and secondary emulsified oil injections, and conversely decreased as RDX degradation slowed. The trends indicate that the nitroso metabolites were being produced in measurable, albeit not stoichiometric, concentrations and were being further transformed, degraded, or otherwise attenuated, and therefore were not expected to be present at any appreciable concentration further downgradient. To that end, during the final sampling event of the demonstration approximately 30 months after the initial oil injection, MNX, DNX and TNX were below detection (<0.08 µg/L) in 11 of the combined in-barrier and downgradient wells, and were present at a maximum of 1.1 µg/L in the remaining four wells that had detectable intermediates. The data suggest that the RDX ring structure was being broken during biodegradation, leading to non-toxic or otherwise labile products.
The concentration of HMX in the groundwater before the demonstration averaged 15 ± 5 µg/L, and concentrations in the upgradient well remained in the same range for the duration of the study (17 ± 3 µg/L). As with RDX, HMX concentrations decreased significantly downgradient of the biobarrier after emulsified oil injection, with a degradation “front” slowly moving down the centerline of the plot. HMX removal was slightly lower than RDX removal, averaging 77 µg/L ± 20% for the in-barrier wells and 61 µg/L ± 32% for the centerline wells from the first emulsified oil injection to the end of the demonstration. However, as noted for RDX, during periods with sufficient TOC, low HMX concentrations were achieved in the centerline wells. For example, approximately six months after the initial oil injection, the HMX within the barrier to 40 ft downgradient ranged from <0.03 to 4 µg/L with an average of 1.2 µg/L. Similarly, 10 months after the second emulsified oil injection, HMX concentrations along the centerline wells ranged from <0.03 µg/L (6/7 wells) to 2 µg/L (1 well) as far as 40 ft downgradient of the barrier. Like RDX, the data suggest that this technology was also highly effective for HMX removal when appropriate biogeochemical conditions were achieved.
The concentration of perchlorate in the groundwater before the demonstration averaged 36 ± 11 µg/L, and concentrations in the upgradient well remained in the same range for the duration of the study (34 ± 5 µg/L). Perchlorate removal was greater than both RDX and HMX, with 91 µg/L ± 9% removal in the barrier wells, and was comparable to total RDX removal along the centerline at 76 µg/L ± 21% from the first emulsified oil injection to the end of the demonstration. During periods with sufficient TOC, low perchlorate concentrations were achieved in the centerline wells. For example, six months after the initial oil injection, the perchlorate within the barrier to 40 ft downgradient ranged from <0.5 µg/L (4 wells) to 17.2 µg/L (1 well) with an average of 3.2 µg/L. Similarly, 10 months after the second emulsified oil injection, perchlorate concentrations along the centerline wells ranged from <0.5 µg/L (5/7 wells) to 2.2 µg/L (1 well) as far as 40 ft downgradient of the barrier, with an average concentration of 0.9 µg/L—an overall reduction of >97%.
This field trial at NSWCDD suggests that an emulsified oil biobarrier is a viable alternative to reduce the migration of co-mingled perchlorate and explosives in groundwater at this and similar range sites. The optimal areas for application of this technology include OB/OD sites, munitions test ranges, EOD training areas, target areas, munitions disposal sites, and other regions where high concentrations of munitions constituents are likely to occur. Despite heterogeneous subsurface lithology, low pH, and low hydraulic conductivity in the aquifer at NSWCDD, emulsified oil and buffer were well distributed to form a subsurface biobarrier. RDX, HMX, and perchlorate were reduced by ≥92% in the centerline of MWs extending 40 ft downgradient of the biobarrier after the second injection of emulsified oil, and accumulation of nitroso-degradation products from RDX was minimal. Moreover, the biobarrier required no operation and maintenance (O&M) other than injection and reinjection of oil substrate, and resulted in no impacts to ongoing range activities.
A cost analysis for full-scale application was completed for several different applicable treatment technologies, using a base case in which a shallow aquifer is contaminated with perchlorate and RDX from 10 to 40 ft below ground surface (bgs) with a plume width of 400 ft. The passive emulsified oil biobarrier had the lowest capital costs under this scenario, and had overall 30-year life cycle costs like several other in situ alternatives, including a zero-valent iron (ZVI) barrier, a mulch biobarrier, and a semi-passive emulsified oil barrier, each of which ranged from $2.4M to $2.6M. The actual costs of these technologies would depend on the longevity of each treatment under site geochemical conditions. All of the in situ technologies were appreciably less costly than an ex situ pump and treat (P&T) option, which was >$3.6M over a 30-year lifespan.
The general implementation concerns of these end users are likely to include the following: (1) technology scale-up technology and applicability under local site conditions, (2) secondary impacts to the local aquifer, and (3) technology cost versus other remedial options.
This technology is amenable for use at a variety of testing and training ranges. Consideration should be given to emplacing the barrier in an area that is not likely to be impacted either directly by detonations or by unexploded ordnance (UXO). While not feasible at all sites, emplacement of permanent, flush-mounted IWs should be preferred over using Geoprobe® injection methods, both in terms of ease of follow-on injections to maintain barrier effectiveness, and in terms of limiting UXO clearance activities to only those needed for IW installation. At more aggressive ranges, hardened IW vaults may be required to protect the infrastructure.
Emulsified oils have been widely used for other applications, such as treatment of chlorinated solvents, so scale-up for an application with explosives and perchlorate should not be problematic. In the case at NSWCDD, the biobarrier could have easily been scaled from 100 to 300 ft or so, which would have been a full-scale design for one of the two identified plumes. Due to the relatively low hydraulic conductivity of the groundwater aquifer at the NSWCDD site, another way to implement this approach full-scale would be to install a sand/gravel trench barrier cross-gradient to groundwater flow, with lines for the addition of emulsified oil. This trench system would replace the closely-spaced biobarrier IWs, and could be quickly rejuvenated with additional emulsified oil on an annual or semi-annual basis as necessary.
Some of the potential limitations of this approach include (1) cost or technological barriers at increased depth (beyond that easily obtained by a direct-push technology [DPT] rig), (2) difficulty injecting emulsified oils in low permeability formations, and (3) secondary groundwater impacts. Aquifer depth is one of the limiting factors for all fully passive designs, which become increasingly expensive due to close spacing of injection points or technically impractical (e.g., for passive trench barriers) as the depth to the water table increases. In addition, emulsified oils are most effectively injected in aquifers where the hydraulic conductivity >4 x 10-3 centimeters per second (cm/sec) (~10 feet per day [ft/day]), and become impractical <~1 x 10-4 cm/sec (~0.3 ft/day).
Secondary groundwater impacts typical of passive approaches include mobilization of metals (notably iron [Fe], manganese [Mn], and arsenic [As]) and production and accumulation of methane. In a typical application of emulsified oil, Fe and Mn will be mobilized within the treatment zone to milligrams per liter (mg/L) concentrations, but these metals will generally be oxidized and precipitated rapidly, so that dissolved concentrations return to background levels within several meters downgradient of the IWs. Similar results are expected for methane, which is usually oxidized in an aerobic aquifer via methane-oxidizing bacteria.
The emulsified oil biobarrier is generally a cost-effective option, particularly for an active range where other options such as P&T are technically impractical due to the surface structure and O&M required. Important cost factors for such biobarriers include (1) plume characteristics, particularly the plume width and depth, which will determine the costs of well or trench installation as well as the quantities of emulsified oil required; and (2) hydrogeology and aquifer characteristics, such as the rate of groundwater flow and general geochemistry (e.g., presence of alternate electron acceptors such as oxygen [O2], nitrate [NO3-], and sulfate [SO42-]), which will determine the rate of oil consumption and the necessity for other amendments (e.g., buffer or inorganic nutrients). (Project Completion - 2016)