Permeable Reactive Barriers Action Team
Permeable Reactive Barrier Installation Profiles

TriangleChlorinated Solvents

Metals and Inorganics

Fuel Hydrocarbons



Other Organic Contaminants

Profiles Home View Profiles

Tetrachloroethene, Trichloroethene, 1,1-Dichloroethene

Reactive Media:

Funnel and Gate

Point of Contact:
Arun Gavaskar
Tel: 614-424-3403
Fax: 614-424-3667
Email: Gavaskar@
Environmental Restoration Department
505 King Avenue
Columbus , OH 43201-2693

Area 5, Dover Air Force Base, Delaware, DE

A pilot-scale field demonstration of a permeable reactive barrier (PRB) is being conducted at the Area 5 site at Dover AFB, DE. The demonstration is funded by the Strategic Environmental Research and Development Program (SERDP). The Dover site is contaminated with perchloroethylene (PCE), trichloroethylene (TCE), and dichloroethylene (DCE). The maximum concentrations encountered during site characterization were 5,617 µg/L of PCE, 549 µg/L of TCE, and 529 µg/L of DCE.

Area 5 lies within the Atlantic Coastal Plain Physiographic Province, consisting of Cretaceous to Recent sedimentary deposits of gravel, sand, silt, clay, limestone, marl, and chalk dipping to the southeast. Ground water is located 5-15 ft below ground surface. The clay aquitard is located 40-45 ft bgs. The hydraulic conductivity values used for design of the permeable barrier were based on an aquifer conductivity range of 10-50 ft/day.

Major objectives of the demonstration include comparing two reactive media schemes and examining innovative emplacement techniques designed to reduce the cost of construction for PRB systems. The funnel-and-gate system, installed in January 1998, consists of two gates that are 8 ft wide and 45 ft deep. One gate is filled with pure, zero-valent iron filings with a 10% iron/sand pretreatment zone to stabilize flow and remove dissolved oxygen. The second gate also is filled with iron, but it is preceded by a 10% pyrite/sand mixture. The mixture serves to moderate the pH of the reactive bed, thereby decreasing the precipitates formed.

The gates were constructed with 8-ft-diameter caissons that were removed after reactive media emplacement. The funnel sections were built using Waterloo interlocking sheet piling driven to the 45-ft depth and keyed into the underlying clay aquitard.

The total cost for the system was $800,000. This includes the cost of design, construction, materials, and the reactive material.

Following installation, the reactive and hydraulic performance of the PRB were evaluated primarily through two comprehensive monitoring events in July 1998 and June 1999. Limited monitoring events were conducted periodically throughout the demonstration to monitor specific operating parameters.

At the end of 18 months of operation, core samples of the gate and surrounding aquifer media were collected and analyzed for precipitate formation. Monitoring results showed that the PRB was functioning as designed in terms of capturing the plume and reducing the contaminants down to below target limits. The pretreatment zones (PTZ) in both gates succeeded in removing dissolved oxygen from the ground water before it entered the reactive cell. In addition, the use of pyrite did result in some degree of pH control while the ground water was in the PTZ of Gate 2. Magnesium, nitrate, and silica were the main inorganic species precipitating out of the low-alkalinity ground water as it flowed through the gates.

The reactivity of the PRB was assessed in terms of the ability of the two reactive cells to degrade the influent CVOCs (PCE, TCE, and cis-1,2,-DCE) and their byproducts (cis-1,2,-DCE and VC) down to respective MCLs. CVOCs persisted longer in Gate 2 than in Gate 1. This is because the pyrite in the Gate 2 PTZ is not as strongly reducing as the iron, and does not promote reductive dechlorination to the same extent as iron. No CVOCs were detected in the exit zones in both gates. The maximum concentration discovered in the expected capture zone of the PRB is 3,900 µg/L of PCE. Evaluation of the residence time in the reactive cells and the capture zone of the PRB were the main objectives of the hydraulic performance evaluation. Results show the estimated range of residence times in each reactive cells is 1 to 9 days. This estimate is based on a representative range of gate velocities of 0.46 to 4.1 ft/day measure during the demonstration. Aquifer heterogeneities, differential packing of the media in the two gates, and fluctuations in ground-water flow volume and direction contribute to this relatively wide range of estimates. A relatively strong hydraulic gradient along the flowpath through both gates indicates that ground water is being captured by the PRB. However, due to low hydraulic gradient in the upgradient aquifer, the size and orientation of the capture zone cannot be determined with certainty. In all probability, the capture zone is about 40 to 50 ft wide, an estimate based on the estimated range of flow velocities in the gates, the measured hydraulic gradients through the gates, and the measure water levels in the upgradient aquifer. The geochemical performance indicated the DO and ORP levels in the ground-water decline along the flowpaths, indicating that a strong reducing environment is created in both gates. The PTZ containing iron (Gate 1) was more efficient at removing DO than the PTZ containing pyrite (Gate 2). Very little precipitate buildup was observed in cores from both gates after 18 months of PRB operation. Available information is insufficient to make any conclusions about the long-term performance (i.e., longevity) of the PRB.

The demonstration is being used to validate the document "Design Guidance for Application of Permeable Barriers to Remediate Dissolved Chlorinated Solvents," developed with input from state and federal regulators and published in February 1997. The guidance document will be updated to reflect lessons learned at the completion of the project.

Lessons Learned

The innovative caisson-based installation technique was used successfully to install the reactive media in a relatively deep aquifer. Caisson-based installation was required at Dover AFB because conventional excavation techniques that use a standard backhoe or clamshell were not feasible at Dover due to the greater depth of the aquifer, presence of underground utilities, and higher cost.

Both reactive media in the PTZ of the gates were successful in their primary function of scrubbing out dissolved oxygen. However, once the water left the PTZ and entered the reactive cell, the tendency of the iron to raise the pH overwhelmed any pH control achieved by the pyrite. Achieving pH control was desirable objective because low pH has the potential to reduce precipitation in the reactive cell and thus increase the longevity of the barrier.


Remediation Technologies Development Forum
Sponsored by the Technology Innovation Program

Date Last Modified: January 14, 2000