Surfactant-enhanced Aquifer Remediation Research Articles

Demonstration of pilot‐scale pervaporation systems for volatile organic compound removal from a surfactant enhanced aquifer remediation fluid I. Spiral wound membrane modules

AbstractDuring the summer of 1996, a pilot‐scale demonstration of a surfactant enhanced aquifer remediation (SEAR) process for removal of dense non‐aqueous phase liquids (DNAPLs) from soils was conducted at Hill Air Force Base in Layton, Utah. Five thousand gallons of the extracted DNAPL‐surfactant solution were transported to Cincinnati, Ohio, to be processed at the U.S. EPA's Test & Evaluation Facility, in a pervaporation pilot‐scale unit equipped with spiral wound modules. The main DNAPLs in the extracted SEAR fluid were trichloroethylene (TCE) at 3,000 mg/L, 1,1,1‐trichloroethane (TCA) at 450 mg/L, and tetrachloroethylene (PCE) at 400 mg/L. The solution also contained 2.5 wt % anionic surfactant, 1.5 wt % isopropyl alcohol, and 1.0 wt % salt. This demonstration highlighted complications posed by secondary contaminants to this application of spiral wound modules. Thirteen runs were performed with commercially available spiral wound pervaporation modules at various flow rates and temperatures. Single‐pass VOC removal ranged from 50% to 98%. In all cases, TCE removal was almost identical to TCA removal and both were easier to remove than PCE.

Surfactant enhanced recovery of tetrachloroethylene from a porous medium containing low permeability lenses: 1. Experimental studies

A matrix of batch, column and two-dimensional (2-D) box experiments was conducted to investigate the coupled effects of rate-limited solubilization and layering on the entrapment and subsequent recovery of a representative dense NAPL, tetrachloroethylene (PCE), during surfactant flushing. Batch experiments were performed to determine the equilibrium solubilization capacity of the surfactant, polyoxyethylene (20) sorbitan monooleate (Tween 80), and to measure fluid viscosity, density and interfacial tension. Results of one-dimensional column studies indicated that micellar solubilization of residual PCE was rate-limited at Darcy velocities ranging from 0.8 to 8.2 cm/h and during periods of flow interruption. Effluent concentration data were used to develop effective mass transfer coefficient ( K e) expressions that were dependent upon the Darcy velocity and duration of flow interruption. To simulate subsurface heterogeneity, 2-D boxes were packed with layers of F-70 Ottawa sand and Wurtsmith aquifer material within 20–30 mesh Ottawa sand. A 4% Tween 80 solution was then flushed through PCE-contaminated boxes at several flow velocities, with periods of flow interruption. Effluent concentration data and visual observations indicated that both rate-limited solubilization and pooling of PCE above the fine layers reduced PCE recovery to levels below those anticipated from batch and column measurements. These experimental results demonstrate the potential impact of both mass transfer limitations and subsurface layering on the recovery of PCE during surfactant enhanced aquifer remediation.

High-performance liquid chromatography–diode array detection of trichloroethene and aromatic and aliphatic anionic surfactants used for surfactant-enhanced aquifer remediation

A method utilizing direct aqueous injection with high-performance liquid chromatography and diode array detection (HPLC–DAD) is presented for the quantitation determination of trichloroethene (TCE) in the presence of anionic surfactants that are used to enhance the recovery of dense non-aqueous phase liquids from contaminated groundwater aquifers. The anionic surfactants investigated in this study including alkyl diphenyloxide disulfonate (Dowfax 8390) and dihexylsulfosuccinate (Aerosol MA 80-I) are used to enhance the solubility, and hence recovery, of TCE. In this type of environmental engineering application, the levels of surfactants and TCE encountered are very high (part per million to part per thousand). The anionic surfactants and TCE are quantitatively determined by direct aqueous injection onto a reversed-phase HPLC column with diode array detection. The quantitation limits of the method obtained using 100 μl injections are 0.1 mg/l for alkyl diphenyloxide disulfonates, 20 mg/l for dihexylsulfosuccinate, and 0.05 mg/l for TCE. This approach is advantageous over using gas chromatography for TCE and HPLC for the surfactants because the use of a single analytical instrument reduces sample preparation and analysis times, which increases sample throughput.

Sorption of Nonionic Surfactant Oligomers to Sediment and PCE DNAPL: Effects on PCE Distribution between Water and Sediment

Introduction of surfactant mixtures to the subsurface for the purpose of surfactant-enhanced aquifer remediation requires consideration of the effects of surfactant sorption to sediment and nonaqueous phase liquids. These effects include alteration of the solubilizing properties of the surfactant mixture and alteration of the sorption properties of the stationary phase. Sorption of octylphenol ethoxylate (EO) surfactant oligomers to a low organic carbon content (foc) aquifer sediment and to dense nonaqueous phase liquid (DNAPL) consisting of tetrachloroethene (PCE) was examined in batch experiments. At aqueous surfactant concentrations far below the critical micelle concentration (CMC) of the mixture, sorption to sediment was characterized by an initial steep isotherm for both high and low EO content oligomers, with somewhat greater uptake of high EO content oligomers. This stage of sorption resulted in mild increases in the equilibrium constant, Kd,PCE, for distribution of PCE between solution (including surfactant) and sediment (including sorbed surfactant). As the aqueous surfactant concentration increased, surface aggregation of low EO content oligomers on the sediment commenced, and a dramatic increase in Kd,PCE was observed. At aqueous surfactant concentrations increasing above the CMC, the formation of solution micelles caused the sorbed surfactant concentrations to plateau and then decrease. This decrease in sorbed surfactant, along with competition by micelles for contaminant, likely contributed to the observed rapid decrease in Kd,PCE toward zero. Surfactant sorption to PCE DNAPL was greater relative to sediment by 1−2 orders of magnitude, with much greater uptake of the low EO content oligomers. Sorption of all but the lowest EO content oligomers to the PCE DNAPL was terminated by micellization.

Impact of surfactants for aquifer remediation on physical properties of the aqueous phase

Surfactant-enhanced aquifer remediation (SEAR) has shown great promise in laboratory studies as a technique for removing nonaqueous phase liquids (NAPLs) from soils. Failures of SEAR in the field are often attributed to the reduction of conductivity of the soil. Conductivity may be altered due to changes in liquid properties, in porous medium properties or in both. This study focuses on the changes in liquid properties, specifically viscosity and density, in systems comprising aqueous solutions of surfactants selected for their solubilization potential of NAPLs that are denser than water, such as tetrachloroethylene (PCE). Measurements showed that the viscosity of the surfactant solutions increased with surfactant concentration, with the increase being considerably greater with the addition of ethanol as a co-surfactant. The impact of temperature was also substantial, with the viscosity increasing as the temperature decreased. The solubilization of PCE and the subsequent formation of a dilute emulsion also increased the viscosity, but only slightly. This increase could be predicted using equations developed for calculating the viscosity of emulsions, even at organic liquid concentrations below the point of emulsion formation. The density increased somewhat with the solubilization of PCE in the surfactant solution. This increase could be estimated from a weighted-volume average of the densities of PCE and the surfactant solution. The resultant reductions in conductivity calculated based on the measured changes in viscosity and density ranged from about 26% to 73% at a surfactant concentration of 4% and a PCE concentration of 40–50 g/l, depending on the temperature, the type of surfactant and the ethanol concentration. Since some laboratory studies, particularly those employing fine-grained soils, have reported reductions in conductivity greater than those determined here, changes in porous medium properties must contribute to conductivity reductions as well.

Partitioning of Ethoxylated Nonionic Surfactants into Nonaqueous-Phase Organic Liquids: Influence on Solubilization Behavior

The ability of surfactants to enhance the apparent aqueous solubilities of nonaqueous-phase liquid (NAPL) contaminants has increased interest in surfactant-enhanced aquifer remediation (SEAR) applications. This study examines the effect of surfactant partitioning on surfactant solubilization efficiency by analyzing the solubilization of several chlorinated NAPLs by a homologous series of linear alcohol ethoxylate nonionic surfactants. In general, it is found that the partitioning of surfactants can, especially in the cases of polar NAPLs, lead to substantial losses of surfactant, leaving little surfactant remaining in aqueous solution for solubilization. More importantly, it is found that neglecting surfactant partitioning can produce artifacts in the measured solubilization capacities of surfactants for NAPLs. For example, when surfactant partitioning is neglected, measurements of the solubilization capacity for homologous series of surfactants are often observed to produce a maximum solubilization capacity as a function of the surfactant hydrophile--lipophile balance number (HLB). The work described here demonstrates that the observed maximum solubilization capacity can result from neglecting partitioning of the surfactant into the NAPL. When solubilization capacity is calculated based on the amount of aqueous surfactant present after partitioning losses, no maximum is observed. The implications of this work for aquifer remediation are discussed.

Prediction and minimization of vertical migration of DNAPLS using surfactant enhanced aquifer remediation at neutral buoyancy

Surfactant enhanced aquifer remediation (SEAR) had previously been thought to require a capillary barrier below the contaminated zone to prevent the vertical migration of a microemulsion containing solubilized DNAPL. This paper shows the vertical migration of the dense microemulsion is described and predicted by the value of three dimensionless scaling groups. Embedded within these scaling groups are four design parameters. The value of these parameters can be manipulated in order to reduce the amount of vertical migration anticipated for a given remediation design. Plots have been constructed that illustrate the relationship between vertical migration and the value of the scaling groups; such plots can be used to predict vertical migration and to determine appropriate screen intervals of extraction wells to ensure full capture of the contaminants. This predictive capability has been verified in laboratory experiments. Predicted migration of the microemulsion agreed within about 2% of that observed. Development of the scaling groups is presented, remediation design implications are discussed, and laboratory verification is shown. Additional discussion of the laboratory work is given in a companion paper [Kostarelos, K., Pope, G.A., Rouse, B.A., Shook, G.M., 1998. A new concept: the use of neutrally-buoyant microemulsions for DNAPL remediation, J. Contam. Hydro., this edition].

A new concept: the use of neutrally-buoyant microemulsions for DNAPL remediation

Even in the absence of mobilization of dense nonaqueous phase liquid (DNAPL), the microemulsion that forms when the surfactant solubilizes a dense contaminant such as trichloroethylene will be more dense than water and tends to migrate downward. This paper addresses the issue of migration with a new concept: surfactant enhanced aquifer remediation at neutral buoyancy. Laboratory results of surfactant remediation in two-dimensional model aquifers show that downward migration of microemulsion containing solubilized dense contaminants can be reduced to an acceptable level, even in the absence of capillary barriers in the aquifer. One model experiment was designed to exhibit a small degree of vertical migration and full capture of the microemulsion at the extraction well. The second experiment was designed to demonstrate the effect of large buoyancy forces that lead to excessive downward migration of the microemulsion. Density measurements of aqueous solutions containing sodium dihexyl sulfosuccinate surfactant, isopropanol, trichloroethylene, and sodium chloride are presented. A companion paper presents the results of the flow and transport calculations needed for this approach to surfactant flooding.

Permeability reductions induced by sorption of surfactant

Surfactant enhanced aquifer remediation (SEAR) is an attractive alternative to traditional pump and treat methods for remediating aquifers contaminated with nonaqueous phase liquids. However, initial studies indicate that the application of surfactant can reduce aquifer permeability by more than an order of magnitude, limiting the efficiency of SEAR. A series of column experiments using mixtures of medium sand and montmorillonite clay demonstrate that existing permeability reduction models for biofouling and deep‐bed filtration poorly predict surfactant induced permeability reductions. An alternative permeability reduction model is proposed which is based on the assumption that the sorbed surfactant effectively increases the volume fraction of the clay. The model is shown to reasonably predict observed permeability reductions, particularly for clay fractions less than 20%. A numerical simulation of surfactant transport that incorporates the effective clay fraction model demonstrates that induced permeability reductions significantly influence the transport of surfactant through an aquifer.

A compositional simulator for modeling surfactant enhanced aquifer remediation, 1 formulation

We describe a three-dimensional, multicomponent, multiphase compositional finite-difference simulator for application to the analysis of contaminant transport and surfactant enhanced aquifer remediation (SEAR) of non aqueous- phase liquid (NAPL) pollutants. Mixtures of surfactant, water and NAPL can form many types of micellar and microemulsion phases with a complex and important dependence on many variables of which the dilute aqueous solution typically assumed in SEAR models is just one example. The phase behavior model is central to our approach and allows for the full range of the commonly observed micellar and microemulsion behavior pertinent to SEAR. The other surfactant related properties such as adsorption, interfacial tension, capillary pressure, capillary number and microemulsion viscosity are all dependent on an accurate phase behavior model. This has proven to be a highly successful approach for surfactant enhanced oil recovery modeling, so it was adapted to SEAR modeling. However, there are many significant differences between petroleum and environmental applications of surfactants, so many new features have been added to model contaminant transport and remediation and these are described and illustrated for the first time here.

Simulation of surfactant-enhanced aquifer remediation : C. L. Brown, G. A. Pope, L. M. Abriola & K. Sepehrnoori, Water Resources Research, 30(11), 1994, pp 2959–2977

Simulation of surfactant-enhanced aquifer remediation : C. L. Brown, G. A. Pope, L. M. Abriola & K. Sepehrnoori, Water Resources Research, 30(11), 1994, pp 2959–2977

Simulation of surfactant‐enhanced aquifer remediation

Surfactant‐enhanced aquifer remediation (SEAR) is currently under active investigation as one of the most promising alternatives to conventional pump‐and‐treat remediation for aquifers contaminated by dense nonaqueous phase organic liquids. An existing three‐dimensional finite‐difference enhanced oil recovery simulator is adapted to model the SEAR process. This simulator incorporates the complex chemistry and multiphase transport behavior of surfactant/water/organic mixtures in permeable media. Model governing equations and parameter requirements are discussed, and simulations are employed to illustrate some important issues potentially affecting SEAR performance at the field scale. Simulations suggest that the total time for remediation could be reduced by more than an order of magnitude over conventional remediation approaches by employing SEAR. The assumptions, approximations, and conditions required to achieve such a favorable result are identified, and the importance of modeling as a quantitative tool for the assessment of SEAR is highlighted.

Surfactant‐enhanced remediation of DNAPL zones in granular aquifer systems

AbstractDense nonaqueous phase liquids (DNAPLs), in particular chlorinated solvents such as trichloroethene, pose groundwater contamination problems at hazardous waste sites across North America. The mobility of DNAPLs in the subsurface, their low aqueous solubility, and the heterogeneity of typical aquifer systems combine to create conditions that inhibit the rapid remediation of DNAPL sites by traditional pump‐and‐treat methods. Surfactant‐enhanced methods for DNAPL‐site remediation accelerate the pace of remediation in granular aquifer systems, e.g., alluvium and outwash. The importance of adequate hydraulic conductivity and aquitard conditions is stressed in the application of surfactant‐enhanced aquifer remediation (SEAR).