Abstract

This report details a research project on steam-enhanced remediation (SER) for the recovery of volatile organic contaminants (VOCs) from fractured limestone. The project was conducted at an abandoned quarry at the former Loring Air Force Base in Limestone, Maine. The project was carried out by:

• EPA’s National Risk Management Research Laboratory
• EPA Region 1
• The Maine Department of Environmental Protection (MEDEP)
• SteamTech Environmental Services, Inc.
• The United States Air Force
• Academic experts on characterization of fractured rock and steam injection remediation

EPA's Superfund Innovative Technology Evaluation (SITE) program participated in this research project to evaluate the SER technology in the fractured rock setting.

Loring Air Force Base was added to the Superfund National Priorities List in 1990, and the quarry there was one of more than 50 sites on base that were addressed. Historically, the quarry was used for the disposal of wastes from construction, industrial, and maintenance activities at the base. During remedial activities in the 1990s, approximately 450 drums were removed. Subsequent investigations showed that both chlorinated organics and fuel-related compounds were present in the ground water beneath the quarry.

Tetrachloroethylene (PCE) was detected at concentrations indicative of the presence of dense non-aqueous phase liquids (DNAPL). The Record of Decision, signed in 1999, recognized that it was currently impractical to restore ground water in fractured rock to drinking water standards. However, an agreement was made among the U.S. Air Force, MEDEP, and EPA Region I to use the quarry for research to further the development of remediation technologies in fractured rock, with the hope of recovering enough contaminant mass to reduce the timeframe for natural attenuation of the remaining contaminants. In addition, the regulatory agencies hoped to develop guidance on characterization techniques for fractured rock. A Request for Proposals for technologies to be tested at the site was issued in 2001, and SER was chosen from the proposals received.

Technology-specific objectives for the research project were developed. They included determining whether SER could:

• Heat the target area for remediation
• Enhance contaminant recovery
• Reduce contaminant concentrations in the rock and ground water

Secondary objectives included determining whether contaminants were mobilized outside of the treatment area, documenting the ability of SteamTech's effluent treatment systems to meet discharge requirements, determining operating parameters for fractured rock, and documenting costs.

Characterization activities began in 2001 with the installation of process boreholes based on the designated treatment area and the preliminary design of the treatment system. These borings were cored and logged, and rock chip samples were collected from fracture surfaces for determination of contaminant concentrations. Additional characterization activities included discrete interval transmissivity testing and ground water sampling, conventional borehole geophysical and acoustic televiewer logging, and interconnectivity testing. Based on the results of all the characterization activities and an updated conceptual site model, the steam injection and extraction system was revised to include steam injection at the eastern side of the target area, with extraction along the center line and the western side of the target area.

Construction of the system began in August 2002 and the extraction system starting operation on August 30. Steam injection was initiated on September 1 and continued until November 19, when funding for the project ran out. Extraction was terminated on November 26. Throughout operations, the SITE program collected effluent vapor and water samples to document the contaminant recovery rate and amount of contaminant recovered. SteamTech collected temperature data, using 22 thermocouple strings, and documented changes in subsurface resistivity caused by temperature increases or by steam replacing water in the fractures, using electrical resistance tomography (ERT).

Early on, it became apparent that steam injection rates were much lower than anticipated due to low transmissivities in the injection intervals and sparsely spaced fractures. In an attempt to inject more steam and increase the rate of heating, three extraction wells were converted to injection wells during the operation. Although this significantly increased the amount of steam being injected, the amount of energy that could be injected during the time-limited project was still low, and the entire target zone for treatment could not be heated. The highest recorded temperature away from the injection wells was approximately 50° C, which was recorded approximately 4.5 meters (15 feet) from the nearest injection well.

ERT was found to be capable of monitoring the heatup of the subsurface during SER; however, the magnitude of the resistivity changes determined was not consistent with the expected change based on prior laboratory measurements of the resistivity of limestone as a function of temperature. Based on the limited duration of steam injection during this project, it cannot be determined conclusively that steam injection would be capable of heating the entire treatment area to the target temperature. However, because the rock chip sampling showed that most of the contaminants were located at the fracture surfaces or within 0.3 meter (1 foot) of the fractured surface, the heat that was injected was concentrated where the contaminants were found. It is possible that adequate treatment might have been achieved even without achieving target temperatures throughout the target zone.

Despite the limited heating that occurred, effluent vapor and water samples showed that after approximately three weeks of operations, the extraction rates started to increase and they continued to increase for the duration of the project. The highest extraction rates were achieved at the end of the project, after steam injection had ceased and air injection was increased. This is believed to be due to air stripping of VOCs at the higher subsurface temperatures, which carried the vaporized contaminants to extraction wells. Effluent samples showed that more than 7.4 kilograms (kg) (16.2 pounds [lbs]) of contaminants were recovered during the project. The contamination consisted of 5.0 kg (11.12 lbs) of chlorinated VOCs; 0.55 kg (1.22 lbs) of gasoline range organics; and 1.77 kg (3.9 lbs) of diesel range organics (DRO). Based on the high concentrations of PCE and DRO in some wells during the last round of sampling, it is believed that DNAPL was about to be extracted.

Sampling of the effluent vapor and water streams just prior to discharge showed that the vapor and water treatment systems employed by SteamTech effectively treated these streams to meet discharge limitations. Ground water samples from two angled wells that extended below the treatment area showed that contaminants did not appear to have been moved downward by SER. Ground water samples from two wells just to the north and east of the treatment area showed that contaminants were not moved horizontally into those areas. Evaluation of operations data shows that higher steam injection pressure can be used in competent bedrock than is typically possible in unconsolidated media. The data also show the importance of the co-injection of air and pressure cycling to enhance the transport of mobilized contaminants to extraction wells.

As additional characterization information became available, and after the completion of the steam injection, the evolution of the conceptual site model allowed an evaluation of the importance of different characterization activities. The evaluation was done to understand ground water and contaminant transport in fractured rock, and to design and implement the SER system. It was determined that a variety of characterization activities are required to understand the flow system and contaminant distribution sufficiently for remediation system design and operation.

For simple or moderately complex large fractured rock sites, SER may be an efficient and cost-effective remediation technology for VOCs. However, for highly complex, low permeability fractured sites with low interconnectivity, such as the Loring Quarry, steam injection may not be the best technology for remediation.

In order for SER to be successful in such an environment, extensive characterization is needed and extremely long injection times are likely necessary. Even with long injection times, heat losses may limit the ability to heat the entire target zone. For sites such as this, thermal conductive heating (TCH) or electrical resistance heating (ERH) may be more capable of uniformly heating the target zone and may be effectively implemented with less characterization, which would result in an overall reduction in remediation costs. Further research is warranted on steam injection remediation in fractured rock in less complex sites and on the application of TCH and ERH to contaminated fractured rock sites.

Contact

Eva Davis

See Also

Superfund Innovative Technology Evaluation

You will need Adobe Reader to view some of the files on this page.
See EPA's PDF page to learn more.