SUMMARY OF THE REMEDIATION TECHNOLOGIES DEVELOPMENT FORUM
IN SITU FLUSHING ACTION TEAM MEETING

Ray Close Conference Room--Building 100
Hill Air Force Base, Utah
May 8-9, 1997

INTRODUCTION AND WELCOME
Dr. A. Lynn Wood, U.S. Environmental Protection Agency (EPA)
Mr. Stephen Shoemaker, E.I. Dupont DeNemours & Company, Inc.
Mr. Robert Elliott, Hill Air Force Base (AFB)

Dr. Wood and Mr. Shoemaker, the two co-chairs for this Remediation Technologies Development Forum (RTDF) In situ Flushing Action Team, opened the meeting by welcoming participants, thanking them for their attendance, and explaining that this meeting is a followup to one held with a precursor group. Both chairpeople were impressed to see how interest in in situ flushing technologies has grown, based on the number of participants present at the meeting (see Attachment A for a list of participants). The co-chairs identified four goals for the meeting:

• Introduce participants to RTDF
• Provide an overview of unresolved issues pertaining to in situ flushing technologies
• Provide an overview of recent technological advancements
• Define a mission and role for the RTDF action team

Mr. Elliott welcomed participants to Hill AFB and talked about Hill AFB's nonaqueous phase liquid (NAPL) ground-water contamination. NAPL removal has posed a challenge, but Hill AFB is optimistic that in situ flushing technologies could provide a solution. A series of in situ flushing demonstration projects have been performed at Hill AFB; these demonstrations have been made possible by the cooperative relationship that exists between Hill AFB and regulatory agencies, namely EPA and the Utah Department of Environmental Quality.

RTDF OVERVIEW
Mr. Richard Steimle, Technology Innovation Office (TIO), Office of Solid Waste & Emergency Response, EPA

Mr. Steimle presented an overview of the RTDF program. EPA established the RTDF in 1992 when industry spokespeople expressed interest in identifying ways for the government and industry to work together to solve complex hazardous waste remediation problems. Today, a number of RTDF teams have been formed, all of which focus on a particular hazardous waste issue or a remedial technology. Updates on the progress of each RTDF team will soon be available on the RTDF Internet site, maintained by EPA.

The RTDF allows government and industry to work together to develop and improve environmental technologies so that hazardous waste cleanup goals are met in a safe and cost-effective manner. The RTDF fosters public- and private-sector partnerships to undertake research, development, demonstration, and evaluation efforts needed to achieve common cleanup goals. The RTDFs mission is three-fold:

• Identify priority remediation technology development needs
• Establish and oversee action teams to plan and implement collaborative research projects to address remediation problems
• Address scientific, institutional, and regulatory barriers to innovative treatment technologies

RTDF teams:

• Share information about planned and ongoing research
• Define research needs, develop detailed research project plans, and implement projects that often entail field-scale demonstrations
• Ensure that all research is founded on sound scientific and engineering principles
• Enlist partners to support and participate in collaborative research efforts
• Produce and disseminate scientifically credible results to facilitate broad acceptance of technology

No strict guidelines outline how RTDF action teams should be managed. In general, two co-chairs, one from industry and one from EPA, lead each team. Co-chairs from all RTDF teams convene once a year to discuss their respective team's progress. Beyond the co-chairs, the organizational structure of each team is unique. According to a 1996 fact sheet, 47 percent of RTDF members were from industry, 32 percent from government, and 21 percent from academia.

Membership for RTDF teams is unrestricted and all interested organizations are welcome to participate. Members should be dedicated to carrying out the RTDF mission rather than joining just as a means to stay up-to-date on the latest advancements in a particular field. Members also should be willing to donate time or money to further the development of the technology. (The RTDF originally required a cash downpayment from all members, but this requirement has been suspended.)

All of the RTDF programs are currently supported on a budget of $20 million. Government sources (e.g., EPA, Department of Energy, Department of Defense) provide 48 percent of this funding, while corporations contribute 43 percent and industrial associations contribute 9 percent. No strict guidelines describe how RTDF teams should bankroll projects, although teams are encouraged to find ways within each team to fund their projects. For large projects, proposals can be written to government agencies.

OVERVIEW OF THE GROUND-WATER REMEDIATION TECHNOLOGIES ANALYSIS CENTER (GWRTAC)
Mr. Richard Steimle, TIO, Office of Solid Waste & Emergency Response, EPA

Mr. Steimle presented information about GWRTAC, a large information center established in 1995 and operated by the National Environmental Technology Applications Center (NETAC) in association with the University of Pittsburgh's Environmental Engineering Program. GWRTAC's activities are overseen by a guidance committee consisting of 25 people. Its mission and role are to:

• Serve as a focal point for information on innovative ground-water technologies
• Support customers' and stakeholder groups' need for the information
• Improve the understanding and commercial use of the technology

GWRTAC keeps up-to-date on the status of innovative ground-water remediation technologies. Several databases are accessible through GWRTAC, including the Technology Database, the Vendor Database, and others that link to various environmental remediation technology databases. Several types of technical reports are available through GWRTAC, including technology evaluation reports, information reports, technology overview reports, and status reports. GWRTAC advertises upcoming events (e.g., demonstrations, conferences, technical courses), and provides news briefs summarizing technological demonstrations, initiatives, and regulatory acceptance issues. Additionally, GWRTAC conducts a variety of customized information searches on a fee basis.

GWRTAC can be accessed at:

SUMMARY OF IN SITU FLUSHING ACTIVITIES
Mr. Richard Steimle, TIO, Office of Solid Waste & Emergency Response, EPA

Mr. Steimle explained that in situ flushing technologies reduce ground-water contaminant concentrations by injecting an aqueous solution into a zone of contaminated soil/ground water. Subsurface contaminants are solubilized or mobilized by the injected flushing agent and then brought to the surface via a downgradient extraction well. Flushing offers a variety of benefits over other ground-water remediation technologies:

• Does not require excavation, handling, or transport of large quantities of contaminated soils
• Is potentially applicable to a wide range of contaminants in both vadose and unsaturated zones
• Accelerates site remediation and achievement of cleanup goals
• Is useful in treatment train applications

A total of 40 in situ flushing demonstration projects have been completed or are currently underway in the United States. Attachment B lists these 40 sites as well as the contaminants located at each and the flushing agent utilized. Mr. Steimle presented a map that indicated how many projects had been conducted in each EPA region. Regions 5 and 8 lead the nation with ten and eight projects, respectively. No projects have been conducted in Region 1 or 9. Mr Steimle speculated that the lack of projects in Region 1 could reflect an unwillingness to use this technology in hard rock/fractured rock environments.

The success of in situ flushing is highly site-dependent and, as a result, requires a detailed site-specific predesign stage. Numerous geological factors can affect the outcome; for example, high clay content, high organic content, fractured rock, soils with large surface areas, and low hydraulic conductivity can hinder success. Soil pH and buffering capacity and contaminant properties may also affect the success of flushing systems. Properties that can affect outcome include degree of water solubility, soil sorption constants, octanol/water partitioning coefficients, vapor pressure, liquid viscosity, and liquid density.

A variety of agents can be utilized as flushing agents, with different degrees of success for different types of contaminants. Mr. Steimle offered a list of agents and the conditions where they are applicable:

Of the 40 demonstration projects listed in Attachment B, 24 involve surfactants, 11 utilize water, 5 utilize cosolvents, and 10 use other flushing agents. Researchers are currently investigating the potential of alternative flushing agents such as cyclodextrins, humic acid, and food additives.

Many issues need to be resolved before in situ flushing can be considered for widespread use, including:

• Risk of spreading contaminants horizontally or vertically into areas that were previously uncontaminated
• Risk of plugging soil pores with precipitates that form from reactions between flushing agents and contaminants
• Uncertainties associated with the performance and predictability of the system, including:

• Ability to achieve cleanup levels
• Ability to address multiple contaminants with varying properties
• Ability to perform in areas with geological features that are not conducive to success (e.g., low permeability, high clay or organic content, high degree of heterogeneity or secondary permeability)
• Ability to predict the length of time required to achieve cleanup goals (remediation times can be lengthy for dissolved phase contaminants due to slowness of the diffusion process)
• Ability to model in situ flushing operations

• Gaining approval from regulators, who are concerned about:

• The toxicity of residual flushing solutions
• Whether injected materials will spread to previously uncontaminated areas. (Regulators will need to be convinced that in situ flushing systems are completely contained either by physical or hydraulic controls.)
• Allowing this technology to be used in areas close in proximity to sensitive recharge areas or potable aquifers
• Permitting issues, such as determining when Class V underground injection permits or National Pollutant Discharge Elimination System (NPDES) permits are needed

Mr. Steimle noted that state agencies are generally receptive to in situ flushing technologies, especially when the flushing agents are nontoxic. EPA contacted every state and found Nebraska to be the only state with policies that prohibited use of surfactants and other injectables. (One participant in the audience noted that some Nebraska regulators have shown interest in use of in situ flushing.) Mr. Steimle's comments prompted discussion among participants. Some have not found state regulatory agencies to be as agreeable as Mr. Steimle has indicated, although they did admit that the situation is improving. Members from industry note that states are more apt to be receptive when they are included early in the decision-making process and treated as valued team members.

• Convincing decision-makers to utilize this technology by providing:

• Demonstrated field performance
• Realistic prediction of duration to achieve goals
• Cost-benefit analysis

TESTING AND DEVELOPMENT NEEDS OF INNOVATIVE REMEDIATION TECHNOLOGIES
Dr. P. Suresh Rao, University of Florida

Dr. Rao, a researcher from the University of Florida, discussed some issues pertaining to the health of the innovative technologies market. The National Research Council (NRC) is scheduled to release a report offering findings and recommendations related to this issue at the end of the summer.

Characteristics of the Remediation Market

Dr. Rao said the remediation market is amorphous and diverse. Problems associated with different sites are highly variable, with each site presenting different contamination issues and hydrogeologic characteristics. Because the problems are so diverse, solutions must be flexible; one technology will not have the same success at all sites. Clients served by remedial technologies are also diverse, with about two-thirds managed by private owners and one-third managed by the government.

Dr. Rao described the remediation market as fragmented and difficult to define or predict. Optimists claim that market potential has not yet been reached and the market is about to "take off." Pessimists, however, state that the market is on a steady decline.

Barriers to Utilization of Site Remediation Technologies

In the last 10 years, many companies in the remedial technology industry have had difficulty breaking into the market. As a result, many small companies have been forced to merge with larger ones. The current market is regarded as "mature," with only a few large companies providing the majority of services.

Many barriers restrict more widespread use of innovative remedial technologies. In the majority of cases, sites choose to delay remedial activities or to utilize conventional technologies (e.g., ex situ pump and treat remediation, soil vapor extraction [SVE]). Such choices are driven by economic, technical, and regulatory factors.

Technological Testing Issues

To convince site managers to use innovative technologies, vendors will need to prove that the technology will effectively solve site-specific problems in a cost-effective manner with minimal risk of failure. Claims of successful performance will need to be supported by convincing field testing and multiple lines of evidence. Controlled experiments will need to be performed to convince clients that the technologies are actually reducing contaminants rather than just displacing them. In an ideal situation, vendors visiting sites will be able to present performance data and automatically gain the confidence of clients at different sites. Dr. Rao asked the following questions:

• What is the minimum data set needed for vendors to approach different sites and gain the confidence of clients? How many successful demonstrations will need to be performed before site managers and the public have faith in a technology?
• Is it better to do large pilot demonstrations or duplicate small experiments?
• How big do pilot demonstrations need to be to represent the range of heterogeneities at the site?
• Can performance and cost be predicted at a large scale?
• Will we ever be able to develop protocols that will address concerns at all sites and eliminate the need for peer review?

OVERVIEW OF OPERATIONAL AND ENVIRONMENTAL ACTIVITIES AT HILL AFB
Dr. Jon Ginn, Hill AFB

Dr. Ginn provided an overview of the operational activities and environmental issues at Hill AFB. The War Department established Ogden Arsenal (now known as Hill AFB) in the 1920s for use as a storage center for unused ordnance. Since its inception, Hill AFB has been used for a variety of functions, including manufacturing of explosives, bombs, artillery shells, and small munitions; handling and storing munitions; and storing, testing, and repairing missiles. Hill AFB was placed on the National Priorities List in 1987 and nine operable units (OUs) have been identified. For the purpose of this meeting, the NAPL ground-water plumes associated with OU-1 and OU-2 are of interest because several in situ flushing demonstrations have been conducted.

Ground water at OU-1 has been contaminated by petroleum products from activities at fire training areas and by oils, fuels, and solvents from disposal practices at chemical disposal pits #1 and #2. The OU-1 plume is characterized by light nonaqueous phase liquids (LNAPL). A Record of Decision (ROD) to determine remedial action for OU-1 will be signed in 1998.

Ground water at OU-2 has been impacted by disposal activities at chemical pit #3. Estimates suggest that 100,000 to 1 million gallons of wastes (consisting primarily of trichloroethylene [TCE], tetrachloroethylene [PCE], and 1,1,1-trichloroethane [1,1,1-TCA]) were disposed in these pits. The OU-2 plume is characterized by dense nonaqueous phase liquids (DNAPL). In 1993, a solvent recovery system was installed and about 30,000 gallons of DNAPL were removed. An ROD was recently signed for this OU and the AFB is planning to install an interceptor trench to prevent off-base migration. The base is open to suggestions to upgrade remedial efforts and will be analyzing the success of different demonstrations that have been conducted at this plume, namely surfactant flushing, surfactant/foam flushing, steam technologies, and SVE.

REGULATORY PERSPECTIVES REGARDING IN SITU FLUSHING TECHNOLOGIES
Mr. Robert Stites, EPA, Region 8

Mr. Stites, the EPA Region 8 project manager for Hill AFB, offered the group insight into regulatory concerns regarding in situ flushing technologies, including:

• Defining remedial goals. One question regulators struggle with is: "How clean is clean?" Opinions about appropriate remedial goals vary between organizations, branches within an organization, and individual regulators. EPA generally believes that remedial efforts should restore media to their original beneficial use (e.g., potable ground water should be restored to drinking water standards). EPA does realize, however, that it is impossible to achieve this goal in some situations and thus often uses less conservative standards. In the case of NAPLs, which have proven to be difficult to remove by conventional technologies, EPA expects site managers to remove as much as "practicable," a term EPA defines as "between practical and possible." The view of what is practicable may change as new innovative technologies become available. EPA issues Technical Impracticability Waivers in rare instances when the Agency determines that remedial goals cannot be achieved at a particular site. These waivers could become even harder to obtain in the future if in situ flushing technologies raise regulatory expectations.

• Permitting issues. Regulators need to issue permits before new materials can be injected into the subsurface. Some states have nondegradation/antidegradation policies that do not allow additional contaminants to be introduced to the environment. States with such policies struggle with the definition of "contaminant" when deciding whether to allow flushing technologies. Many regulators think that introducing flushing agents is acceptable, particularly if agents are injected at concentrations below their maximum concentration limits (MCLs). Some agents used for flushing do not have MCLs or are relatively nontoxic (e.g., food additives). Although regulators are generally more accepting of nontoxic agents, they still worry about the issue of biodegradability. Mr. Stites suggested that, when trying to decide whether to allow use of flushing technologies, regulators consider the following question: Even if some of the flushing agent remains in the subsurface, how does it compare to what's down there now?

• Undesired mobilization. Regulators need to be convinced that in situ flushing technologies will not mobilize contaminants into previously uncontaminated areas.

• Waste management issues. Regulators need to be convinced that the effluent from the system will be disposed in an efficient and safe manner.

• Uncertain performance capacity. Information touting the success of in situ flushing technologies is relatively scant. Regulators who want to learn more about this technology have had to rely on "beehive-type" communication. Mr. Stites is hopeful that the new Web site being formulated will help this situation.

• Cost effectiveness. Regulators do look at cost-benefit ratios when they evaluate technologies. Some regulators are concerned by in situ flushing technology's high upfront costs. In Mr. Stites opinion, however, this technology may be less costly in the long-term because it will not require long-term operation and maintenance and will require fewer engineering controls.

Mr. Stites discussed ways to gain regulatory approval, underscoring the importance of involving regulators early-on in the decision-making process. Mr. Stites also suggested concentrating efforts on better dissemination of information. (Comprehensive Environmental Response, Compensation, and Liability Act [CERCLA] training programs offer training in innovative technologies, but in situ flushing technologies are not presently included.) A report summarizing what has worked would be very helpful to regulators.

STRATEGIC ENVIRONMENTAL RESEARCH AND DEVELOPMENT PROGRAM (SERDP) ENHANCED SOURCE REMOVAL

Project Description
Dr. Carl Enfield, EPA

Dr. Enfield, the technical point of contact for the SERDP project on enhanced source removal, defined SERDP's role and gave a brief overview of in situ flushing demonstrations conducted under SERDP. This SERDP project evaluates the effectiveness of emerging technologies by performing side-by-side field demonstrations and comparing them to each other and to conventional pump and treat technology. (Pump and treat is used as a comparison because this technology is currently being used at more than 90 percent of sites undergoing remedial activities.) Under the SERDP project, researchers characterize the sites, install and operate innovative technologies, evaluate the performance of the technology, and write manuals incorporating results from the different demonstration projects.

A number of in situ flushing demonstration projects have been performed under SERDP at Hill AFB. (The results of some of these projects will be summarized below.) The different demonstrations utilized different flushing agents but similar setups; for example, all the SERDP demonstrations at OU-1 were performed in cells isolated with physical barriers (sheet pilings). A preponderance of data will be required before people are convinced that in situ flushing technologies are effective. These SERDP demonstrations, therefore, evaluate performance in multiple ways, including:

• Comparing pre- and post-flushing soil core data for a selected number of contaminants of regulatory concern
• Calculating the mass of the analytes removed from effluent streams
• Comparing pre- and post-flushing partitioning tracers. (This comparison helps determine how much NAPL is present. The degree of the chromatographic separation between the peaks indicates how successful the flushing was. Less separation is indicative of success.)
• Comparing pre- and post-flushing conservative tracers. (This comparison helps determine how the hydrogeologic properties of the system change.)
• Comparing pre- and post-flushing concentrations of ground-water analytes

The field work for the SERDP demonstrations at Hill AFB have been completed but investigators are still interpreting the data. SERDP is planning additional studies at Dover AFB in Delaware.

Cosolvent Mobilization--Cell 3 at Hill AFB's OU-1
Dr. Ronald Falta, Clemson University

Dr. Falta opened his discussion by explaining the two mechanisms involved in cosolvent NAPL removal. Flushing agents that solubilize NAPLs increase the total aqueous solubility of the contaminants. Agents that mobilize NAPLs reduce the interfacial tension. The degree of solubilization versus mobilization that occurs during a cosolvent flood can be controlled by choosing cosolvents that place emphasis on these different mechanisms.

The Cell 3 team's goal was to mobilize the NAPL. Analysis of the LNAPL, however, indicated that mobilization of this particular NAPL would be a challenge. As a result, the flushing agent chosen was nearly pure alcohol, with a small amount of added water to prevent it from freezing. The alcohol flood consisted of three phases: first, 1,000 gallons of tert-butanol were injected, followed by 2,000 gallons of a tert-butanol/hexanol mixture, and finishing with 4,000 gallons of tert-butanol. After the alcohol flood was complete, water was injected to remove any remaining cosolvent. Dr. Falta noted a few challenges that arose along the way:

• Buoyancy. The Cell 3 team had to overcome a buoyancy issue to remove the light alcohols injected into the subsurface.
• Conductivity issues. Early in the process, the Cell 3 team noted dramatic conductivity differences within the test cell. (Conductivity between upper and lower portions of the cell differed by a factor of 20.) To overcome this problem and to direct the cosolvent to required areas, the Cell 3 team screened the injection and extraction wells selectively in order to better direct the flow of injectate.

A preliminary data analysis offers the following conclusions:

• It is technically feasible to mobilize and remove LNAPLs using a cosolvent.

• The cosolvent effectively mobilized the NAPL and reduced contaminant concentrations. (An estimated 25 to 30 gallons of NAPL were mobilized.)
• Tracer tests indicate that 75 percent of the NAPL was removed.
• Soil core data suggest contaminant removal percentages ranged from 75 to 95 percent, depending on the contaminant analyzed. (Removal percentages of decane and undecane were between 76 and 81 percent. Removal of toluene and 1,1,1-TCA was higher, exceeding 90 percent.)

• Economic feasibility will hinge upon efforts to recycle cosolvents.
• The mobilization of DNAPLs by cosolvents will probably require density modification.

Cyclodextrin Solubilization--Cell 4 at Hill AFB's OU-1
Dr. Mark Brusseau, University of Arizona

Dr. Brusseau opened his discussion with a description of cyclodextrin, a reagent that differs markedly from alcohols and surfactants. Cyclodextrins, glucose based molecules, are used in many pharmaceutical, food processing, and chemical processes. Dr. Brusseau used a beta-cyclodextrin, consisting of seven glucose molecules, each of which has three alcohol (-OH) functional groups attached. The -OH groups, arranged on the outside of the molecule, make the outer portions of the molecule highly soluble in water. The inner "core is an area of low polarity. Cyclodextrins enhance the solubility of hydrophobic NAPL constituents by forming inclusion complexes. The NAPL constituents can partition into the cyclodextrin core, where they are shielded from surrounding water.

Cyclodextrins have several properties that make them excellent candidates for use in flushing technologies because they:

• Are relatively nontoxic
• Have a relatively high solubilization power
• Have low reactivity with soils (i.e., in the laboratory cyclodextrins do not sorb or interact with soils--even soils with high organic and clay content)
• Are environmentally stable (i.e., their solubilization power is not strongly influenced by wide changes in pH or ionic strength)
• Are produced at commercial scale
• Exhibit conservative behavior (i.e., they are easy to remove from the subsurface)
• May simultaneously enhance the biodegradation of contaminants

For the demonstration at Hill AFB, the Cell 4 team injected about 10 pore volumes (PVs) of 10 percent cyclodextrin into the subsurface. The team used two different methods to calculate the percent removal for several contaminants. Results obtained using the different methods differed by less than a factor of two for the majority of contaminants and ranged from 39 to 83 percent. For four contaminants (1,1,1-TCA, o-xylene, decane, and undecane) the results differed by more than a factor of three. The Cell 4 team is currently investigating this discrepancy.

The Cell 4 team was able to predict field behavior fairly well with laboratory data. Before performing the field experiment, the team calculated the amount of LNAPL that they expected to be extracted. For many contaminants, the predicted numbers were very similar (within a factor of two) to those actually measured in the field. In other cases, the values measured in the field were less than predicted, perhaps as a result of competition, complexation by the cyclodextrin in a multicomponent solution, or enhanced biodegradation of labile contaminants. The Cell 4 team will be investigating this matter further.

The Cell 4 team compared the solubility enhancement potential of cyclodextrins with water so that they could draw a comparison with conventional pump and treat systems utilizing water flushing. The results indicate that cyclodextrins offer superior solubility enhancement.

Dr. Brusseau noted that further investigation needs to be performed to determine the feasibility of recycling cyclodextrins. Recycling is a particularly important issue with cyclodextrins because they are more expensive than some surfactants.

Surfactant Solubilization--Cell 6 at Hill AFB's OU-1
Dr. Robert Knox, University of Oklahoma

According to Dr. Knox, this test cell is referred to as Hill AFB's "problem cell." Although the team installed sheet pilings, they experienced problems with leakage. Water level monitoring indicated that leakage rates exceeded 0.3 percent PVs per day; as a result, the team was forced to alter its original plan and inject surfactant into only three injection wells rather than four (the fourth well was used to pump water). The Cell 6 team used Dowfax 8390 (solutions of 4.3 weight percent) as its flushing agent, injecting two PVs of water, followed by 10 PVs of Dowfax solution, and then followed by another two PVs of water.

The Cell 6 team formulated preliminary conclusions based on the following evaluation methodologies:

• Comparison of pre- and post-flushing soil core samples. Removal percentages varied markedly across different contaminants, ranging from negative values to 68 percent, with an average removal percentage of 37 percent. (Negative values resulted because contaminant leaked into the cell.) Dr. Knox notes several issues that still need to be addressed regarding their soil core data (e.g., failure to collect pre- and post-flushing samples from the same depth, averaging issues, choice of sampling location).
• Comparison of pre- and post-flushing partitioning tracer tests. The Cell 6 team experienced some difficulties operating at the right levels during the post-partitioning tracer test. As a result, some of the curves calculated are nonsensical. Curves plotted for extraction Well #2, regarded as the well exhibiting the best data, indicate that 50 percent of LNAPL was removed during the flush (i.e., LNAPL saturation decreased from 8.3 percent to 4.0 percent).
• Analysis of effluent samples from extraction wells. Total petroleum hydrocarbon (TPH) increased by two orders of magnitude over the course of the demonstration. Dr. Knox notes that the team needs to resolve issues pertaining to the analytical methods used to calculate concentrations. (He recently discovered that the team's method differs from Michigan Tech's method.)
• Comparison of pre- and post-flushing equilibrium ground-water samples. TPH decreased from concentrations of 25 milligrams per liter (mg/L) to 10 mg/L.

Surfactant Mobilization--Cell 5 at Hill AFB's OU-1
Dr. David Sabatini, University of Oklahoma

The Cell 5 team mobilized NAPL using a middle phase microemulsion (MPM) surfactant system. Dr. Sabatini explained that the MPM phase appears when the hydrophilic-lipophilic balance (HLB) of a surfactant system is adjusted so that it is right at the interface between being water and oil soluble. MPM phases are characterized by ultra-low interfacial tension and can cause bulk oil displacement. Dr. Sabatini notes three ways that a surfactant can be manipulated to achieve the MPM phase:

• Adding salinity to surfactants with high HLBs
• Adding a hydrotrope to surfactants with low HLBs. (This is normally the procedure that Dr. Sabatini prefers, but the complexity of the Hill AFB NAPL prevented its use.)
• Adding calcium chloride to adjust the HLB. (This approach was utilized at Hill AFB. The surfactant utilized was Aerosol-OT/TMAZ.)

Achieving MPM phases with a NAPL as complex as the one at Hill AFB was a challenge, but not impossible. Soil core data indicate that more than 90 percent of the analyzed contaminants were removed, except naphthalene (75 percent removed). Analysis of partitioning tracer tests indicate that 75 percent of the NAPL was removed, with NAPL saturation at 8.5 percent prior to flushing and 2.0 percent afterward.

Dr. Sabatini recommends further analysis of the following:

• The importance of sweep efficiency
• Short- and long-term risks associated with the surfactant
• The use of coupled processes to achieve regulatory compliance

AIR FORCE CENTER FOR ENVIRONMENTAL EXCELLENCE SURFACTANT DEMONSTRATION--HILL AFB'S OU-2
Dr. Gary Pope, University of Texas

Dr. Pope's team performed a solubilization flush on the DNAPL at Hill AFB's OU-2. The NAPL at OU-2 was less complex than the NAPL at OU-1 and consisted primarily of 1,1,1-TCA, TCE, and PCE. The field demonstration at OU-2 differed from those at OU-1 in that hydraulic controls were used rather than physical controls.

Dr. Pope's team used a flushing agent consisting of 8 percent Aerosol MA (also called sodium dihexyl sulfosuccinate) and 4 percent isopropyl cosolvent. The flushing agent was chosen based on successful performance in the laboratory. Aerosol MA is a particularly attractive choice for use at Hill AFB because it does not sorb to Hill AFB's soils. Budgetary restrictions prevented Dr. Pope from adding a polymer to the flushing agent, although he is certain that the addition of a polymer would have offered mobility control and would have reduced costs (by decreasing the amount of surfactant required), shortened remediation times, and mitigated heterogeneities.

Dr. Pope's demonstration was conducted in two phases. Phase I proved that partitioning tracer tests work well and provided insight for design of Phase II. Phase II involved injecting 2.5 PVs of the flushing solution into the subsurface. The flush resulted in a removal of 510 to 530 gallons of DNAPL. Estimates for the percent removal of contaminants ranged from 97 percent (calculated by comparing ground-water contaminant concentration in wells before and after flushing) to 99 percent (calculated by comparing DNAPL saturation provided by partitioning tracer tests).

ADVANCED APPLIED TECHNOLOGY DEMONSTRATION FACILITY (AATDF) DEMONSTRATIONS

Overview
Dr. Stephanie Fiorenza, Rice University

Dr. Fiorenza provided an overview of the AATDF program, which was formed in May 1993 and is scheduled for completion by October 1997. AATDF's funding ($19.3 million) supports projects investigating innovative remedial technologies.

Dr. Fiorenza described how AATDF determined which of the 170 proposals they received to fund. After narrowing the prospects to 38 projects, AATDF conducted "due diligence" to make its final selection. Over a course of 4 to 5 months, principal investigators and other related experts argued for certain projects. In the end, 12 projects were selected. Four of these projects will be discussed by presenters below.

AATDF identified two types of challenges that in situ flushing technologies face: technical challenges pertaining to subsurface heterogeneity and unintended migration, and economic challenges posed by the high cost of surfactants and cosolvents. AATDF hopes field-demonstration projects will provide insight on how to increase mobility control and improved sweep efficiency; achieve more efficient miscible displacement to reduce PVs; and recover and reuse surfactant.

Single-Phase Microemulsion (SPME)/Cosolvent Comparison--Hill AFB's OU-1
Dr. Michael Annable, University of Florida

Dr. Annable worked on two different projects at Hill AFB. Both projects focused on solubilization of NAPL rather than mobilization because the investigators did not want to address concerns regarding migration to previously uncontaminated areas. One of the demonstrations utilized SPME technology and the other used cosolvents. Both projects were conducted at Hill AFB's OU-1, but at different cells.

SPME Study

Dr. Annable opened his discussion by describing the basis of SPME technology. This technology utilizes a surfactant/cosurfactant combination that forms a water-continuous, low-viscosity, microemulsion when it comes into contact with NAPL. The microemulsion can be diluted in water and transported through porous media as a single-phase, low-viscosity fluid. Using SPME technology to solubilize agents is advantageous because it aims to remove all fractions of the oil phase rather than just specific components of the NAPL. Dr. Annable notes the difference between the SPME phase, which solubilizes NAPL, and Dr. Sabatini's MPM phase, which mobilizes NAPL via immiscible displacement. The flushing agent used in the demonstration consisted of 3 percent surfactant (Brij 97® [C₁₈O(C₂H₄O)₁₀H]) and 2.5 percent cosurfactant (n-pentanol). A total of nine PVs were used, but the majority of the contaminants were removed in the first one or two PVs.

Dr. Annable concludes that this technology effectively removes NAPL in a reasonable timeframe. Preliminary conclusions can be drawn based on the following evaluation methodologies:

• Comparison of pre- and post-flushing soil cores. The percentage of contaminant removed ranged from 87 percent (for naphthalene) to 96 percent (for 1,2-dichlorobenzene and undecane).
• Comparison of pre- and post-flushing partitioning tracer tests. The percentage of contaminant removal was approximately 75 percent.
• Mass balance estimates. Estimates for the percent of NAPL removed differed when different methodologies were employed to calculate mass balance. Using soil core data to calculate initial mass, removal percentages over 90 percent were calculated. Using information provided by partitioning tracer tests, percentages dropped to 59 to 76 percent.

Cosolvent Study

The flushing agent used in the cosolvent study consisted of 70 percent ethanol and 12 percent pentanol. When choosing the cosolvent, the team believed that ethanol was the only alcohol regulators deemed acceptable for injection into the ground. Many people are finding, however, that regulators are receptive to other cosolvents as well. Given this new development, Dr. Annable wanted to note that blending cosolvents may allow researchers to use solutions that contain only 20 to 30 percent of cosolvent.

The cosolvent study, which was completed about 2 years ago, is offering some interesting information on spatial issues, an area that requires further exploration. Different performance levels were observed between upper and lower layers of the cell. For example, the percent removal of 1,2-dichlorobenzene was 99 percent at 16.25 feet but only 49 percent at 20 feet. This difference in performance may be attributable to the fact that the low density flushing agents have difficulty reaching deeper layers. The team tried to mitigate this problem by using gradient injection. Dr. Annable stresses that the issues of density override (when trying to sink flushing agents) and underride (when trying to displace flushing agents) will have to be examined further.

Comparison of the SPME and Cosolvent Studies

Dr. Annable concluded that the efficacy between the two methods is comparable. In terms of efficiency:

• Both utilized similar PVs.
• The flushing agent used in the cosolvent study had a much higher percentage of chemical additives and was therefore more expensive.

Dr. Annable notes that the issue of recycling and reusing surfactants/cosolvents should be resolved if either of these technologies is going to be used in a full-scale field study.

Surfactant With Mobility Control--Hill AFB's OU-2
Dr. Clarence Miller, Rice University

Dr. Miller's team's demonstration was similar to that conducted by Dr.Pope's team, with one novel exception: it incorporated mobility control via foam. Laboratory experiments analyzing the ability of a surfactant (Aerosol MA) to extract a DNAPL (TCE) in a heterogeneous environment (high permeability upper layer and low permeability lower layer) indicate that the surfactant does not effectively extract TCE from low permeability areas. By adding air to the laboratory test cells, and thereby creating a foam, the TCE is extracted more effectively and at a more constant rate. Using such a technology is very beneficial for Hill AFB's OU-2 because characterization studies indicate that many of the contaminants are in areas of low permeability.

Based on promising laboratory results, Dr. Miller's team initiated a study in the field. Their flushing agent was composed of 4 percent Aerosol MA and 11,500 parts per million of sodium chloride (NaCl). The NaCl was added to promote the formation of a third phase (see Dr. Sabatini's discussion). Although Dr. Pope's team found that the efficacy of their surfactant was increased when combined with a cosolvent, alcohols could not be used for Dr. Miller's demonstration because alcohols degrade foam. The team simultaneously injected air and three PVs of surfactant into the subsurface. At one point, the rate of air injection was decreased because the researchers were unsure how substantial pressure increases would impact injection wells; later, this fear was deemed unfounded and air injection rates were increased again. A lot of the air escaped and never reached subsurface targets, foam was observed in two monitoring wells, however, and was present at a range of depths (even the bottom foot above the clay layer). Dr. Miller's team completed work at OU-2 in April 1997 and, like the other presenters at this conference, has not had a chance to fully evaluate its data. Preliminary analysis indicates that DNAPL recovery exceeded that which was initially present, reflecting that the cell was not completely isolated from incoming DNAPL sources.

Surfactant Recovery--HILL AFB's OU-2
Dr. Jeffrey Harwell, University of Oklahoma

As was noted by several participants, efforts to recover and reuse surfactants are vital if in situ flushing technologies are going to be cost effective. Dr. Harwell explained that this AATDF project attempted to determine optimal recovery schemes for surfactant recovery and to determine under what conditions the recovery systems can be economically employed. Recovery involves two processes: 1) decontamination (separating the surfactant from contaminants), and 2) reconcentration (removing excess water). Previous studies indicate that it is economically advantageous to decontaminate entire waste streams before reconcentrating them. For heavy surfactants (molecular weight > 1,000), monomers might be recoverable through ultrafiltration; lighter molecules generally require multistage foam recovery processes. Challenges facing researchers trying to recover surfactants from an in situ flush include:

• Separation processes are less effective in dilute solutions.
• Micelles reduce thermodynamic activity.
• Surfactants stabilize foams and emulsions.

Dr. Harwell's team analyzed four different recovery processes:

In the laboratory, each of these recovery systems was tested for its ability to recover nonionic and anionic surfactants from a stream that also contained isobutylbenzene (IBB), PCE, and oil (squalene). The vacuum stripper foamed under all conditions, even when antifoam agents were added; the packed column air stripper did not foam as long as the liquid loading or the air/water ratios were kept low; the hollow fiber air stripper did not foam as long as the air pressure exceeded water pressure by 1 to 2 pounds per square inch absolute (psia). The liquid/liquid extraction system did not produce an emulsion and was the only system that readily stripped the lowest IBB and PCE concentrations.

After the lab studies were completed, the team decided that design criteria for recovery systems should be based on 90 percent removal rather than 99 percent because achieving the latter doubles recovery costs and is unnecessary since the material is just going to be reinjected back into the most contaminated part of the aquifer. This decision allows air stripping to be considered as an option, even with low volatility contaminants. Use of air stripping is preferable if it can be designed so that foaming does not occur. Packed column air stripping is preferable over membrane stripping because the latter is 40 percent more expensive.

After completing all lab work, the team conducted field scale demonstrations for the packed column air stripper and the hollow fiber membrane module using surfactant wastestreams generated by Hill AFB in situ flushing teams. They concluded the following:

• Two technologies are available that can handle volatile organic compounds.
• Both the packed column and the hollow fiber unit could be operated reliably without foaming.
• When air temperatures dropped to -2.0 degrees Celsius, both units were able to operate properly but required heating.
• Hollow fiber units are easy to operate, easy to scale-up, and their design can be based on thermodynamic data without pilot scale studies.
• No design algorithm exists for scaling-up the packed column without foaming.
• Air stripping range can be extended by steam stripping or preheating the feed.
• The larger columns needed for full-scale air stripping will somewhat offset the higher cost of membranes.
• Non-volatiles will require solvent stripping and solvents will need high temperatures and/or vacuum regeneration.

Technology Practices Manual for Surfactants and Cosolvents
Dr. Thomas Simpkin, CH2M Hill

Dr. Simpkin presented the "Technology Practices Manual for Surfactants and Cosolvents," a document that he coauthored. He explained that this manual is not a guidance manual, research summary, or textbook. It is to be used as a tool for informed decision-making. Two editions are planned for this manual; the first has been completed and is available at www.ch2m.com/ch2mhill/services/environ/ RICEMANUAL/TOC.htm. Dr. Simpkin welcomes any comments on this draft and encourages new data to be submitted by July so that they can be incorporated into the second edition (scheduled for release in fall 1997). The manual provides a technical description and update to the current status of in situ flushing technology, information on hydrogeology and the basic mechanisms involved with flushing technologies, a stepwise approach to implementation (not to be confused with a protocol), and information on cost considerations. The section on cost estimates analyzes several different site scenarios.

ECONOMICS OF SURFACTANT-ENHANCED EXTRACTION
Mr. Stephen Shoemaker, E.I. Dupont DeNemours & Company, Inc.

Mr. Shoemaker presented the results of a report, presented in October 1996, entitled "Economic Study of Surfactant-Enhanced Pump and Treat Remediation." The goals of this study were to:

• Examine the economics of surfactant-enhanced pump and treat remediation of PCE from spills of different sizes
• Provide guidelines for the use of surfactant in pump and treat remediation
• Provide a template for evaluating specific remediation sites

The study evaluated a variety of variables such as volumes of PCE contamination, affected aquifer volumes, type of surfactant, surfactant concentration, pumping rate, remediation time, and solubilization versus mobilization. The following conclusions were drawn:

• Present cost is less than $5 million for surfactant solubilization of PCE from a hypothetical half-acre, 20-ft thick DNAPL impacted source zone. Over a 5-year period, the cost distribution is:

• 38 percent for the surfactant (this estimate is optimistic in that a portion of the surfactant can be recovered)
• 27 percent for sheet piling
• 30 percent for process costs (about half of the process costs represent costs for the recovery system)
• 5 percent for wells

• Given the high costs of surfactants, optimizing the use of the agent and recovering the agent is important.
• Remediation time requirements do not significantly affect costs.
• Optimal flow-rates per acre can be established.
• Mobilization appears more costeffective than solubilization.

INTEGRATED DEMONSTRATION OF SURFACTANT ENHANCED AQUIFER REMEDIATION WITH SURFACTANT REGENERATION/REUSE
Ms. Laura Yeh, Naval Facilities Engineering Service Center (NFESC)

Ms. Yeh opened her discussion by explaining that the Navy is very interested in technologies that utilize surfactants because approximately 100 Navy sites are contaminated with DNAPL (e.g., TCE), and no available technologies have been proven to effectively remove DNAPL.

Ms. Yeh is leading an effort to test surfactant recovery processes. Her team's objective is to prove that surfactants can be recovered and reused, and that use of such a recycling system can significantly reduce the overall cost of using surfactants for subsurface remediation. The recovery system involves a combination of pervaporation and micellar-enhanced ultrafiltration (MEUF).

Like other researchers interested in promoting in situ flushing technologies, Ms. Yeh's team has faced several barriers, the largest of which has been finding a site where they can perform a demonstration. She considers the following characteristics to represent conditions of an "ideal" testing site:

• Substantial amounts of residual DNAPL concentrated in a small area/volume
• Thick, continuous, confining layers present in the contaminated aquifer, or the aquifer has no specific supply use
• Permeable soils (hydraulic conductivity should be greater than 0.0001 centimeters per second)
• Homogeneous subsurface

Recently, the team has altered its strategy for finding a site and hopes that it will soon be successful. Once it has completed its demonstration it will conduct an economic analysis.

REMEDIATION OF CHLORINATED SOLVENTS AT THE BACHMAN ROAD SITE USING INNOVATIVE TECHNOLOGIES
Mr. Tohren Kibbey, University of Michigan

Mr. Kibbey discussed a project being conducted by the University of Michigan, Michigan State, and Georgia Tech to investigate surfactant-enhanced aquifer remediation at the Bachman Road site in Michigan. The team has just initiated work and plans to characterize the site within the next year, perform a pilot study in 1998, and complete a full-scale demonstration by 2001. The team has not decided which surfactant will be used, but is planning to initiate surfactant screening tests after the site characterization is complete.

SUMMARY OF OPEN-FLOOR DISCUSSIONS

Mission Statement

Meeting members formulated a mission statement for this RTDF group:

To facilitate the development and implementation of in situ flushing technology for site remediation.

Identification of Action Items and Future Activities

A list of action items that need to be addressed by this RTDF group was generated. It was generally agreed that the different action items should be tied together under one "umbrella" project and that a document focusing on designing a conceptual, generic, large in situ flushing study should be created. The group expressed the opinion that creating a manual will force the group to address the majority of issues/needs that were discussed during the 2-day meeting, and it will foster the conduct of consistent demonstration projects. Five action areas will need to be further pursued to compile this document:

• Design a full-scale demonstration.
• Further develop technical practices and produce a general protocol. This action area will address:

• Technical issues (e.g., ways to deal with heterogeneities, ways to maximize sweep efficiency via mobility control, ways to tailor a surfactant to a particular contaminant, ways to treat extracted wastewater streams).
• Standardization of field activities, documentation practices, and the peer review process. Participants noted that the criteria for valid experiments differs between academics and regulators and clients. Researchers investing efforts into this field need a clear outline of what is expected in areas such as quality assurance/quality control (QA/QC). Standardizing the way that investigators collect and present data will enable peer reviewers to draw better comparisons between different technologies.

• Perform economic evaluations and compare the costs of in situ technologies against conventional treatments.
• Further develop systems that allow for surfactant recovery and reuse.
• Analyze end-point issues and define success criteria by answering "How clean is clean?"

It was agreed that it would be most effective to split the participants into subgroups to work on these different areas. The two co-chairs plan to find out who is interested in each action item and to identify subteam leaders. Dr. Wood suggested setting deadlines to achieve certain goals. Dr. Pope suggested setting up an e-mail group so that RTDF members can update each other on their activities without having to physically meet.

The co-chairs tentatively set a date for the next meeting in September/October 1997. The goals for the next meeting are to formalize the mission and action plans. Dr. Pope recommended that the co-chairs outline a list of specific elements to be addressed under each subteam, e-mail it to everyone, allow people to comment on the list, and then incorporate the comments before the next meeting.

Other Issues

Other action items identified by participants include:

• Disseminating information to regulators and site managers to increase their confidence in using this technology and to eliminate misconceptions
• Focusing on identifying applicable test sites for future studies and improving NAPL and source identification
• Performing more demonstrations
• Further analyzing this technology's potential niche in the market by evaluating:

• Who is going to need this product
• How many sites will have conditions that are conducive to this technology
• The status of competing technologies

• Identifying funding sources for further development because funding for subsurface technology research is diminishing rapidly
• Clearly defining the benefits of partial mass removal and determining whether other processes, such as natural attenuation or bioremediation, can be used to achieve cleanup levels after the in situ flushing has been performed. (One member suggested that it would be a natural next step to collaborate with the Bioremediation Consortium of the RTDF.)

Attachment A

In Situ Flushing Action Team Meeting

The Ray Close Conference Room - Building 100
Hill Air Force Base, UT
May 8-9, 1997

Final Attendee List

*Speaker

Attachment B