Remediating DNAPLs

• Jul 01, 1999

Related Web sites

When dense non-aqueous phase liquids (DNAPLs) percolate downward through the subsurface, they become trapped below the groundwater table on low permeability stringers and aquitards. As they migrate below the groundwater table, DNAPLs commonly form networks of pools and interlacing strings. Once in an aquifer, DNAPLs represent distributed source terms that are extremely difficult to access and remediate.

To address the problem of DNAPL remediation at federal sites, the Department of Defense (DOD), Department of Energy (DOE), Environmental Protection Agency (EPA) and the National Aeronautics and Space Administration (NASA) have formed the Interagency DNAPL Consortium (IDC). The IDC was created specifically to identify and evaluate innovative remediation technologies with the potential to remediate DNAPL contamination.

After a rigorous evaluation process, three technologies have been selected by the IDC for demonstration and comparison at the Cape Canaveral Air Station in Florida:

1. Electrical resistive heating (Six-Phase Heating^TM (SPH))
The preferential application of SPH is in those areas where DNAPLs and groundwater are present with a heterogeneous lithology.

2. In situ chemical oxidation (using potassium permanganate (KmnO₄)
Oxidant injection is the injection of oxidizing compounds into the sub-surface to break down contaminants. It is typically used for the treatment of dissolved phase contaminants.

3. Steam injection
This is the injection of steam into the groundwater for heating and stripping groundwater contaminants. It is typically used for the treatment of heavier hydrocarbons like creosote and pentachlorophenol (PCP).

Technology demonstrations will be conducted in fiscal years 1999 and 2000; detailed cost and performance data will be published after the demonstrations.

The DNAPL problem
By definition, DNAPLs are liquids that are heavier than water and have a low aqueous solubility. Compounds commonly associated with DNAPL releases to the environment include the chlorinated solvents perchloroethylene (PCE), trichloroethylene (TCE) and dichloroethylene (DCE). When DNAPLs are released into the subsurface they tend to migrate rapidly downward toward the bottom of the aquifer. If not remediated, DNAPLs will slowly dissolve into the groundwater over a period of years. In addition to directly contaminating large quantities of groundwater, DNAPLs also degrade to form other hazardous substances, such as vinyl chloride (VC), that can threaten human health and inhibit land reutilization.

Chlorinated solvents were first manufactured in the United States in the early 1900s; their subsequent use mirrored the economic growth of the country. Production quantities ranged from hundreds of millions, to billions of kilograms (kg) per year. Historical use and disposal practices led to widespread releases at thousands of locations.

According to a recent National Research Council study, three of the top 10 most frequently detected contaminants at hazardous waste sites are DNAPL compounds. The Office of Management and Budget estimates that the federal government will spend between $234 billion and $289 billion on environmental remediation over the next 75 years at DOD, DOE, Department of Interior and Agriculture, and National Aeronautics and Space Administration (NASA) sites. Although no specific data is available, experts expect that a significant portion of these budgets will be directed toward DNAPL remediation.

DNAPL remediation
Although much progress has recently been made in developing and improving cleanup technologies, there is still no proven technology for the restoration of DNAPL source zones. The difficulty of this challenge is unprecedented in the field of groundwater engineering. Despite recent notable progress, "The technologies available for the removal of DNAPL from the groundwater zone at appreciable rates are still experimental, and no DNAPL source zone of significant size has been fully restored using any of them" (James F. Pankow and John A. Cherry, 1996, Dischlorinated Solvents and Other DNAPLs in Groundwater: History, Behavior and Remediation, pg. 503). It is also a well-recognized fact that lack of credible performance and cost data limits the deployment of new technologies.

To address these challenges, the DOD, DOE, EPA and NASA have combined complementary programs to form the IDC. The IDC seeks to organize and maximize the resources of the parent organizations, leverage budgets and expedite technology delivery to contaminated sites. Once established, the IDC initiated the Interagency DNAPL Source Remediation Project.

The Interagency DNAPL Source Remediation Project's initial efforts were to identify existing technologies that showed true promise for restoring DNAPL source zones. After evaluating the identified technologies, the most promising were rigorously screened, and three were selected for side-by-side field demonstrations at Cape Canaveral, Fla.

The technology demonstrations at Cape Canaveral will generate performance and cost data for the technologies under controlled field conditions. The ultimate goal of the project is to demonstrate that commercially mature technologies are available to address the removal of DNAPLs from the saturated zone. If successful, the project will also create a mechanism by which site managers can quickly evaluate site-specific cost and performance expectations for applying these technologies at their various facilities.

Six-phase heating
SPH is a polyphase electrical technology that uses in situ resistive heating and steam stripping to accomplish subsurface remediation. A voltage control transformer converts conventional three-phase electricity into six electrical phases. These electrical phases are then delivered to the subsurface by vertical, angled or horizontal electrodes installed using standard drilling techniques.

Because the SPH electrodes are electrically out of phase with each other, electrical current flows from each electrode to all the other out-of-phase electrodes adjacent to it. It is the resistance of the subsurface to this current movement that causes heating. The result is a uniquely uniform subsurface heating pattern that can be generated in both the saturated zones or vadose (unsaturated) zones.

Electricity takes pathways of least resistance when moving between electrodes, and these pathways are heated preferentially. Examples of low resistance pathways include silt or clay lenses - horizontal deposits of sediment that are thin or discontinuous - and areas of high free ion content. As chlorinated compounds sink through the lithology, they become trapped on these same silt and clay lenses. Over time, trapped solvents undergo biological dehalogenation, producing daughter compounds and free chloride ions. Thus, at DNAPL sites, the most impacted portions of the subsurface are also the low resistance electrical pathways that are preferentially treated by SPH.

By increasing subsurface temperatures to the boiling point of water, SPH speeds the removal of contaminants by both increased volatilization and in situ steam stripping. As subsurface temperatures climb, contaminant vapor pressure, and the corresponding rate of contaminant extraction, increases. The Henry's Law constant of typical DNAPL compounds, which represents the ratio of vapor phase concentrations to dissolved phase concentrations, increases by 15 to 20 times as temperatures rise from 10 to 100 degrees Celsius (¡C). Direct in situ volatilization is an extremely important SPH remediation mechanism for contaminants with boiling points below that of water, such as TCE, the isomers of 1,2-dichloroethene (1,2-DCE) and VC.

The ability to produce steam in situ represents the second significant mechanism for contaminant removal using SPH. Through preferential heating, SPH creates steam from within silt and clay stringers and lenses. The physical action of steam escaping these tight soil lenses drives contaminants out of these otherwise diffusion-limited portions of the soil matrix, which tends to lock in contamination via low permeability or capillary forces.

The released steam then acts as a carrier gas that, as it moves toward the surface, strips contaminants from both groundwater and the more permeable portions of the soil matrix. The presence of the steam also causes the boiling point of the DNAPL to become depressed, due to the partial pressure effects described by Dalton's law of partial pressure. Thus, the normal boiling point of PCE is depressed from 121ºC to 89ºC in the presence of steam in the groundwater, causing free phase PCE to be very rapidly volatilized from the subsurface by the SPH process.

Once in the vadose zone, rising steam and contaminant vapors are collected by conventional soil vapor extraction wells. A condenser then separates the mixture into condensate and contaminant-laden vapor. If these waste streams require pre-treatment before discharge, standard air abatement and water treatment technologies are utilized.

Applications of the SPH process
The SPH technology has been successfully demonstrated at several sites and has proven capable of remediating both fuel and chlorinated hydrocarbons from the vadose zone, as well as LNAPL and DNAPL from the saturated zone.

At the DOE's Savannah River, Aiken, S.C., site, a 25-day SPH demonstration reduced PCE concentrations in a 10-foot thick clay lens by over 99 percent. Because the low permeability clay lens was sandwiched between two high permeability sand lenses, the demonstration proved that the SPH process preferentially heats clay layers.

In August 1995, Armstrong Laboratory's Environics Directorate chose SPH as a promising technology for treating DNAPL in the saturated zone. In February 1997, a field test was performed at the Dover Air Force Base Groundwater Remediation Field Laboratory (GRFL) in Delaware. The test's objective was to determine if SPH could heat an aquifer sufficiently to remove DNAPL. The test was conducted in an uncontaminated aquifer using the tracers perfluoromethylcyclohexane (PMCH) and perfluorotrimethylcyclohexane (PTMCH), to mimic TCE and PCE.

In 17 days of heating, temperatures within the entire treatment volume had reached the boiling point of water. Aquifer boiling was continued for another 13 days while tracer recovery sampling was performed. A combination of vapor stream, soil gas and groundwater sampling verified full recovery of the tracers.

At the Dover GRFL demonstration, SPH successfully boiled a flowing aquifer and removed the selected DNAPL tracers. There was full recovery of the tracer PMCH and 35 percent recovery of the tracer PTMCH. Less than full recovery of the PTMCH was due to a failure of the analytical system. Electrical use during the demonstration equaled $16 per cubic meter of subsurface heated.

In each of three separate technology demonstration projects at Fort Richardson, Anchorage, Alaska, SPH successfully removed over 90 percent of the TCE and PCE mass from a complex heterogeneous lithology. Heating times were limited during these demonstrations to six weeks per treatment area. Two weeks were allowed for heating the aquifer to boiling, and an additional four weeks were allowed for remediation of the treatment area. In all three demonstrations, SPH was conducted to a depth of 45 feet below grade (bg), and the heated volume contained three stacked water tables and a soil matrix consisting of low permeability silty-clay lenses intermixed with sandy gravels. Based on cleanup effectiveness and projected economics from these demonstrations, the SPH technology is planned for full-scale deployment at the Fort Richardson site later this year.

During the Fort Wainwright site demonstration project in Fairbanks, Alaska, SPH was used successfully to augment the aerobic bioremediation of gasoline and diesel at a cold weather site. In this application of SPH, subsurface temperatures were increased to 30ºC and held there for 60 days. Once the biological testing period was complete, SPH was then used to boil the aquifer and steam strip gasoline directly from the soil matrix. A final report on costs and effectiveness has not yet been released.

Most recently, the first full-scale deployment of SPH has successfully remediated pools of TCE and TCA from the saturated zone at a former manufacturing facility in Illinois. Efforts to remove chlorinated solvents by combining steam heating, bioremediation, groundwater extraction and soil vapor extraction had been ongoing at the site since 1991. After seven years, these technologies had significantly reduced overall site impact, but had left behind three large DNAPL hot spot areas.

Site lithology consists of heterogeneous sandy silts to 18 feet bg and a dense silty clay till from 18 to 25 feet bg. A shallow groundwater table is encountered at 7 feet bg, and hydraulic conductivity through the remediation zone ranges from 10^-4 to 10^-8 cm/sec. At the start of SPH, most of the remaining solvent mass was pooled on top of the clay till at 18 to 20 feet bg.

A network of 107 SPH electrodes was installed covering two-thirds of an acre. To treat directly beneath a warehouse, 85 of those electrodes were constructed through the floor of the building. Electrically conductive from 11 to 21 feet bg, the SPH electrodes actively heat the depth interval from 5 to 24 feet bg. Once subsurface temperatures reach boiling, steam laden with chlorinated solvents is collected by 37 soil vapor extraction wells screened to 5 feet bg. Full-scale operations of the SPH system began on June 4, 1998.

Within 60 days, temperatures throughout the entire 24,000 cubic yard treatment volume had reached the boiling point of water. With another 70 days of heating, separate phase DNAPL in the area had been removed, and groundwater concentrations of both TCE and TCA reduced to below the targeted Tier III risk based cleanup levels.

Groundwater concentrations of chlorinated hydrocarbons across the SPH remediation area have now been reduced to below the Tier III cleanup levels established for this site, and most of the DNAPL pool area concentrations have been reduced to below the Tier I cleanup levels. Cleanup results are summarized in Figure 1.

This full-scale SPH remediation was performed for $32 per cubic yard of treatment area. Costs were calculated for turn key remediation services, including the installation and operations of the SPH, vapor extraction and condensate treatment systems. Cost calculations also include project permitting and the preparation of work plans, electrical use, waste disposal, interim sampling and progress reporting. Final demobilization, confirmatory sampling and reporting were not included in the cost calculations, as these activities are not complete. Costs for electrical energy were $6.41 per cubic yard of treatment volume. Electrical energy costs were significantly lower than the $16 per cubic meter costs at the Dover demonstration due to technology refinements and a less than full-scale deployment of the SPH technology as was necessary at Fort Wainwright.

This full-scale remediation project demonstrated both the efficacy and cost-effectiveness of the SPH technology for full-scale DNAPL treatment.

Current Environmental Solutions
www.cesiweb.com

Battelle
www.battelle.org

Pacific Northwest National Laboratory
www.pnl.gov

Interagency DNAPL Consortium
www.getf.org/dnaplguest

U.S. Department of Energy
http://home.doe.gov

An introduction to SPH
http://Marketing.cesiweb.com/papers/sph/sph.htm

Savannah River project
http://Marketing.cesiweb.com/white_sheets/savannah/savannah.html

Ft. Richardson project
http://Marketing.cesiweb.com/papers/ft_richardson/unitb.htm
http://Marketing.cesiweb.com/white_sheets/ftrich/ftrich.html

Dover AFB project
http://Marketing.cesiweb.com/papers/dover/dover.htm
http://Marketing.cesiweb.com/white_sheets/dover/dover.html

SPH case study
http://Marketing.cesiweb.com/papers/awma/awma.htm

Skokie site project
http://Marketing.cesiweb.com/projects/skokie/index.htm
http://Marketing.cesiweb.com/white_sheets/skokie/skokie.html

Six-phase heating
www.cesiweb.com

Dynamic underground stripping
www.gnet.org/filecomponent/4581.html

In situ chemical oxidation using KmnO₄
www.envnet.org/scfa/tech/dnapl/factsheets/Kmno4.htm

This article originally appeared in the 07/01/1999 issue of Environmental Protection.