Design and Performance of an Enhanced Bioremediation Pilot Test in a Tidal Wetland Seep, West Branch Canal Creek, Aberdeen Proving Ground, Maryland
By Emily H. Majcher2, Michelle M. Lorah1, Daniel J. Phelan1, and Angela L. McGinty
1 U.S. Geological Survey, Baltimore, Maryland.
2 Geosyntec Consultants, Inc., formerly of the U.S. Geological Survey.
3 EA Engineering, Science, and Technology, Inc., formerly of the U.S. Geological Survey.
Because of a lack of available in situ remediation methods for sensitive wetland environments where contaminated groundwater discharges, the U.S. Geological Survey, in cooperation with the U.S. Army Garrison, Aberdeen Proving Ground, Maryland, conceived, designed, and pilot tested a permeable reactive mat that can be placed horizontally at the groundwater/surface-water interface. Development of the reactive mat was part of an enhanced bioremediation study in a tidal wetland area along West Branch Canal Creek at Aberdeen Proving Ground, where localized areas of preferential discharge (seeps) transport groundwater contaminated with carbon tetrachloride, chloroform, tetrachloroethene, trichloroethene, and 1,1,2,2-tetrachloroethane from the Canal Creek aquifer to land surface. The reactive mat consisted of a mixture of commercially available organic- and nutrient-rich peat and compost that was bioaugmented with a dechlorinating microbial consortium, WBC-2, developed for this study. Due to elevated chlorinated methane concentrations in the pilot test site, a layer of zero-valent iron mixed with the peat and compost was added at the base of the reactive mat to promote simultaneous abiotic and biotic degradation.
The reactive mat for the pilot test area was designed to optimize chlorinated volatile organic compound degradation efficiency without altering the geotechnical and hydraulic characteristics, or creating undesirable water quality in the surrounding wetland area, which is referred to in this report as achieving geotechnical, hydraulic, and water-quality compatibility. Optimization of degradation efficiency was achieved through the selection of a sustainable organic reactive matrix, electron donor, and bioaugmentation method. Consideration of geotechnical compatibility through design calculations of bearing capacity, settlement, and geotextile selection showed that a 2- to 3-feet tolerable thickness of the mat was possible, with 0.17 feet settlement predicted for unconsolidated sediments between 1.5 and 6 years following installation of the reactive mat. To ensure hydraulic compatibility in the mat design, mat materials that had a hydraulic conductivity greater than the surrounding wetland sediments were selected, and the mixture was optimized to consist of 1.5 parts compost, 1.5 parts peat and 1 part sand as a safeguard against fluidization. Sediment and matrix properties also indicated that a nonwoven geotextile with a cross-plane flow greater than that of the native sediments was suitable as the base of the reactive mat. Another nonwoven geotextile was selected for installation between the iron mix and organic zones of the mat to create more laminar flow conditions within the mat. Total metals and sequential extraction procedure analyses of mat materials, which were conducted to evaluate water-quality compatibility of the mat materials, showed that concentrations of metals in the compost ranged from one-half to one order of magnitude below consensus-based probable effect concentrations in sediment.
A 22-inch-thick reactive mat, containing 0.5 percent WBC-2 by volume, was constructed at seep area 3-4W and monitored from October 2004 through October 2005 for the pilot test. No local, immediate failure of the mat or of wetland sediments was observed during mat installation, indicating that design estimates of bearing capacity and geotextile textile selection ensured the integrity of the mat and wetland sediments during and following installation. Measurements of surface elevation of the mat showed an average settlement of the mat surface of approximately 0.25 feet after 10 months, which was near the predicted settlement for unconsolidated sediment.
Monitoring showed rapid establishment and sustainment throughout the year of methanogenic conditions conducive to anaerobic biodegradation and efficient dechlorination activity by WBC-2. The median mass removal of chloromethanes and total chloroethenes and ethane during the performance monitoring period was 98 and 94 percent, respectively, within the 1.5-feet-thick zone between the base of the mat and the middle of the organic zone, whereas no mass removal was observed within the underlying 12-15-feet-thick wetland sediment in the seep area. Following mat installation, transient appearance of daughter volatile organic compounds, including trichloroethene, cis-1,2-dichloroethene, vinyl chloride, and methylene chloride, was observed in groundwater in the mat in association with decreasing concentrations of parent volatile organic compounds. In some areas of the mat, the non-toxic end products of ethene and ethane were consistently detected throughout the monitoring period. An apparent decrease in volatile organic compound degradation and methane production occurred in late winter to early spring, consistent with a decline in microbial activity during the colder months. Water-quality monitoring for pH, specific conductance, nutrients, major ions, and metals indicated that the mat had no adverse effect on the groundwater quality in the wetland sediments surrounding the reactive mat, nor on nearby surface-water quality.
Hydraulic head gradients (0.20 to 0.26 feet/feet) in and around the reactive mat remained dominantly vertically upward following mat installation, consistent with pre-installation calculations. Horizontal hydraulic gradients remained at least one to two orders of magnitude lower than the vertical hydraulic head gradients. Despite the dominantly vertically upward head gradients in the mat, an aerial thermal infrared flight during winter 2005 showed that the reactive mat was not warmer than surrounding wetland sediments, as observed on flights over the seep area prior to mat installation. Observed seep areas in the vicinity of seep 3-4W were consistent with previous flight imagery, however, and porewater sampling indicated that transport of volatile organic compounds had not shifted to these nearby seep areas as a result of the mat installation. This indicates that the design goal of normalizing flow across the mat area without affecting flow in the surrounding wetland was achieved.
Overall, the pilot test showed that the design goal of at least 90 percent mass removal of total chlorinated volatile organic compounds was achieved and maintained for 1 year in the reactive mat without any undesired geotechnical, hydraulic, or water-quality effects on the wetland and tidal creek. Additional monitoring, however, would be necessary to evaluate long-term sustainability of enhanced biodegradation in the mat. Future applications would benefit from either deeper placement within the native sediments or a thinner mat to minimize final elevation of the mat above land surface.