Environ. Sci. Technol. 2004, 38, 1170-1175
Surfactant-Enhanced Air Sparging in Saturated Sand H E O N K I K I M , * ,† H Y O - E U N S O H , † MICHAEL D. ANNABLE,‡ AND DONG-JIN KIM† Department of Environmental System Engineering, Hallym University, Chuncheon, Gangwon-do, 200-702, Korea, and Department of Environmental Engineering Sciences, University of Florida, Gainesville, Florida 32611-6450
Air sparging as a subsurface remedial technique can be enhanced by the addition of a surfactant. The effect of reduced surface tension of water on the extent of air intrusion and air saturation during air sparging in porous media was investigated. A sand column and a two-dimensional sand box were used for the experiments. The surface tension was controlled using an anionic surfactant, sodium dodecyl benzene sulfonate, and the concentration used was below the critical micelle concentration. Using the sand column, the air saturation was measured at different surface tensions and at different airflow rates. Initially watersaturated, the air saturation achieved in the column by air sparging at a surface tension of 3.42 × 10-2 N/m was up to 5 times larger than that of water with no surfactant. At the same time, the rate at which the air saturation increased as a function of airflow rate was greater at reduced surface tensions. For box experiments with homogeneous sand, reduction of the surface tension caused a dramatic increase in the sparging area up to 5.2 times of that generated using water with no surfactant. A sand box experiment containing a vertical channel produced preferential flow of the air phase injected at the bottom of the channel when the surfactant was not applied. However, reducing the surface tension was found to promote airflow through the preferential channel and the finer sand surrounding the channel. These observations support the use of low concentration surfactants to improve air sparging swept zones.
Introduction Air sparging has been developed for removal of volatile organic compounds (VOCs) present either in the groundwater as in dissolved form or as nonaqueous phase liquids (NAPLs). In situ air sparging has attracted broad attention due to its advantages over conventional pump-and-treat. Advantages include the simplicity of implementation, little wastewater produced, reasonable capital and operational costs, and biological stimulation effect by providing oxygen into the groundwater (1-4). This technique, in fact, was found to accelerate the remediation process for NAPL-contaminated aquifers (1, 4). Soil vapor extraction (SVE) is often coimplemented with air sparging when vapor must be recovered or when the vadose zone requires treatment. * Corresponding author e-mail:
[email protected]; phone: +82-33-248-2155; fax: +82-33-256-3420. † Hallym University. ‡ University of Florida. 1170
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 4, 2004
Air sparging, however, has its own limitations that should be considered for a field application as indicated by McCray and Falta (5). When NAPL is present in pools on lenses of low hydraulic conductivity, the mobile air phase may be located too far from the NAPL for effective volatilization. This is because the injected air flows upward due to pressure gradients and buoyancy, producing poor contact with the NAPL. In a geological formation that contains a laterally extensive, horizontal layer of low hydraulic conductivity, the aquitard may prevent the injected air from flowing upward, resulting in an undesirable horizontal flow pattern potentially further spreading contaminants. In other cases, injected air may flow through preferential flow paths not impacting the target zone. This is particularly true at sites where the aquifer contains a fractured structure. A number of physical parameters that affect the efficiency of air sparging have been investigated. The type of media, airflow rate, and air channel spacing were found to affect the removal rate of dissolved VOCs (6-8). Pulsing of air injection was also examined to improve the efficiency of air sparging (9). Many field-scale applications of air sparging have been reported for varying hydrogeological conditions (10-12). Airflow patterns during air sparging were also studied using a lab-scale physical model (13). Theoretical models for airflow and distribution during air sparging are also available (1419). As described earlier, applying air sparging at a field site can pose many engineering challenges due to undesirable airflow patterns caused by hydrogeological heterogeneities. Since the physical characteristics of a site, layered structure, and heterogeneity of soil grain size cannot be altered, engineering options to overcome problems with applying air sparging in natural media are very limited. Burns and Zhang (20) focused on this issue and proposed manipulating surface tension of water for increasing air saturation and producing a stable and uniform air distribution in the air sparging influence zone. Using a single orifice for air injection, the reduction of surface tension using a nonionic surfactant was found to increase the interfacial area between air bubbles and water, to increase the number of bubbles, and to reduce the bubble size distribution (20). Although this result was obtained from experiments using larger glass beads (diameters >1 cm), a similar result is expected for smaller media such as soils. For soils with particle diameter of sub-millimeter order, the injected air may not flow in discrete bubbles but through established air channels instead. If the interfacial tension between injected air and water is reduced, more channels will likely form resulting in a larger unsaturated sparging area. With reduced surface tension, the threshold pressure required for air to get into the pore throat is reduced, so that more soil pores will be filled with air. The effect of reducing surface tension on the local air saturation is expected to be very useful, particularly for the removal of nonaqueous phase liquids (NAPLs) from an aquifer by volatilization using air sparging. Increasing the number of air channels and the air saturation with the help of a surfactant will decrease the distance between NAPL and air, resulting in faster mass transfer to the flowing air phase. Increased air saturation might even cause direct exposure of NAPL to the flowing air, allowing for rapid volatilization. For a DNAPL (dense NAPL), the potential to open local cavities below the water table, either by using two wells or pulse pumping in a single well, is greatly enhanced with the presence of a surfactant. The objective of this study is to investigate the effect of surface tension reduction on the airflow pattern during air 10.1021/es030547o CCC: $27.50
2004 American Chemical Society Published on Web 01/08/2004
TABLE 1. Summary of Experimental Conditions Used in This Study model
SDBS surface concn tension (mg/L) (×10-2 N/m)
airflow rate (cm3/min)
1-D Column no sand with sand
0 0 80 100 200 300
homogeneous
0 45 70 100 0 150 400
7.16 7.16 4.73 4.67 3.89 3.42
5.1, 15.0, 23.6, 31.2 4.0, 9.4, 27.8 4.0, 12.3, 20.1, 36.1 3.2, 10.5, 16.6, 25.7, 34.6 1.0, 2.3, 6.2, 13.8, 19.7 3.0, 7.6, 12.2, 22.9
2-D Box
heterogeneous (with preferential air channel) a
FIGURE 1. Schematic diagrams of the experimental setups: (a) air sparging system with a sand column installed and (b) configuration of the two-dimensional sand box. The sand packing in the box shown here is for the experiments with a preferential air channel. sparging using one- and two-dimensional physical models. The relationship between the degree of air saturation and the surface tension of water in the sparging zone was explored. The effect of reduced surface tension on the flow pattern of air and the extent of desaturation in a two-dimensional sand domain was examined. Airflow characteristics for a soil system with a vertical, preferential air channel was also investigated at varying surface tensions.
Methods and Materials Materials. Two types of sand were used in the column and box experiments. A uniform sand with a diameter range of 200-500 µm was used as the fine sand. A coarse sand, diameter 1-2 mm, was used to establish a preferential airflow channel for the two-dimensional box experiments. Reagentgrade sodium dodecyl benzene sulfonate (SDBS) was purchased from Tokyo Kasei Kogyo Co. Ltd. and used as received. The concentration of SDBS in the aqueous solutions used in this study was kept below the critical micelle concentration of SDBS (414 mg/L) (21). Double-distilled water was used throughout the experiments. Schematic diagrams of the experimental system are shown in Figure 1. The Plexiglas column (5 cm i.d., 100 cm length) was packed with the coarse sand in the bottom 5 cm of the column. The coarse sand layer was used to prevent sand particles from plugging the orifice where air was introduced
7.16 5.68 5.24 4.60 7.20 3.84 2.89
400a
400a
Airflow rate was fixed at 400 cm3/min for all of the box experiments.
and provide better distribution of the injected air. Fine sand was packed above the coarse sand up to 73 cm from the bottom. The bulk density of the sand was approximately 1.66 g/cm3, and the porosity was 0.36. The sand was packed in the column filled with surfactant solution to avoid entrapped air. Following packing, the water table was set at the surface of the sand so that the volume of displaced water during air sparging could be measured at the top of the sand. Air was injected through a stainless steel tubing (0.10 cm i.d., 0.318 cm o.d.) with the opening located at 0.5 cm above the bottom of the column. Two series of box experiments were performed. The first set of experiments used a box packed homogeneously with the fine sand. The box used for the second set of experiments was packed with the fine sand and a coarse sand strip embedded at the center of the box providing a vertical channel for a preferential flow of the injected air (Figure 1b). For all of the box experiments, the sand was packed under water using the surfactant solution with a predetermined concentration to preclude any air entrapment after packing was completed. The thickness of the box between walls was 2.0 cm, and the width of the coarse sand strip was 3.2 cm. A stainless steel diffuser (Alltech. Co., 40 µm pore size, 1.2 cm o.d., 2.5 cm length) was installed at the bottom center of the box. The saturated zone was set at 6 cm below the sand surface. Procedures. Compressed air was injected into the initially water-saturated sand column by opening the control valves (Figure 1). The water, displaced from the sand layer due to air sparging, was allowed to accumulate on top of the sand. The volume of displaced water and the air pressure were measured after airflow was stabilized, and then the airflow rate was increased for the next experimental step. Three to five different airflow rates were applied at each surface tension. After a set of experiments with different flow rates at a fixed surface tension was completed, the column was dismantled and repacked using water with a different surface tension for the next set of experiments. Experimental conditions are listed in Table 1. Box experiments were conducted in the same manner as the column experiments. Compressed air was introduced through a diffuser, and the airflow rate was controlled using a needle valve (Figure 1b). Instead of measuring the air pressure and amount of water displaced during air sparging, the extent of the unsaturated zone generated due to air sparging was monitored at a fixed airflow rate of 400 cm3/ min. After the airflow and sparged zone were stabilized, the boundary between saturated and unsaturated media was VOL. 38, NO. 4, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1171
visually delineated on transparent films, and the images were computer processed. After completion of an experiment at one level of surface tension, the box was dismantled and repacked with the same sand, using water with a different surface tension for next level experiment (Table 1). Tempe pressure cells (Soilmoisture Equipment Co.) were used to measure the water retention curves of the fine sand (22). Pressure cell experiments were performed at three different interfacial tensions: one with deionized water containing no surfactant and the others with aqueous solutions of SDBS at reduced surface tensions. The surface tension of the aqueous solutions used in this study was measured using the drop weight method (23). The outer diameter of the glass tube used for the surface tension measurement was 3.0 mm. Water droplets from the tip of the glass tube were collected and weighed. Surface tension was calculated based on the weight of single droplet and the radius of the glass tube. Droplet-forming rate at the tip of the glass tube was controlled not to exceed 30 s/drop using a syringe infusion pump (KD Scientific, model 200). The water samples used for surface tension measurement were taken from the water displaced from the sand (column experiments), or by withdrawing water from the box using a syringe after the packing was completed. This was necessary in order to measure the actual surface tensions at equilibrium surfactant concentrations following surfactant sorption to the sand.
Water Retention Curves. Water retention characteristics of the aquifer media were needed to determine the capillary pressure required for air entry into the porous media and for quantifying the resulting air saturation at the applied pressure. In porous media, the capillary pressure and surface tension are related as follows (24, 25);
2σ 2σ ) cos R r′ r
(1)
where pc is the capillary pressure (N/m2); σ is the surface tension (N/m); r and r′ are the radii (m) of the pore and the curvature formed at the air-water interface in the pores, respectively; and R is the contact angle between air and water at the point of contact on the soil surface. Also, pc ) pnw pw where pnw is the pressure of the nonwetting phase (air) and pw is the pressure of the wetting phase (water). The capillary pressure head is defined as hc ) pc/(∆Fg), where ∆F is the density difference between fluids and g is gravitational acceleration. In porous media, the water saturation is expressed as a function of capillary pressure (26, 27):
()
S ) (1 - Sr)
pd pc
λ
+ Sr
for pc > pd
(2)
where S is the water saturation (volume of water/soil void volume), Sr is the residual water saturation measured at high capillary pressure [usually 1 × 104 N/m2 (or hc ) 100 cmH2O) or more for sand], pd is the air entry capillary pressure (N/ m2), and λ is the pore-size distribution index (26). The water retention curves for the fine sand used in this study are shown in Figure 2. The air entry capillary pressure head (hd) of the fine sand initially saturated with deionized water (surface tension 7.16 × 10-2 N/m) was measured at 24.4 cmH2O using a method proposed by Brooks and Corey (26). The hd values measured for the same sand initially filled with SDBS solutions with surface tensions of 5.03 × 10-2 and 2.99 × 10-2 N/m were 16.8 and 9.4 cmH2O, respectively. These values were in very good agreement with the values of 17.1 and 10.2 cmH2O, respectively, predicted using the air entry capillary pressure of the fine sand with surfactant-free 1172
9
water, the surface tensions, and eq 1. With reduced surface tension, the curve shifted to the lower capillary pressure head region (or lower suction). Thus, the capillary pressure (or suction) change required to withdraw the same quantity of water from the sand was smaller at lower surface tensions, since the air entry pressure of the sand decreases by the ratio of surface tensions (28):
σ′ pd′ ) pd σ
Results and Discussion
pc )
FIGURE 2. Water retention curves of the fine sand used in this study. Solid lines are the predicted curves for reduced surface tensions based on the curve measured at 7.16 × 10-2 N/m and eqs 1-3.
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 4, 2004
(3)
where σ′ is the reduced surface tension and pd′ is a new air entry pressure predicted from the original air entry pressure (pd) and the ratio of surface tensions (σ′/σ). The predicted water retention curves for the fine sand at the surface tensions of 5.03 × 10-2 and 2.99 × 10-2 N/m were overlaid on the measured water retention curves with good agreement (Figure 2). This observation indicates that the reduction of surface tension decreased the water-holding capacity of sand. Although the physical structure of the soil system did not change, the change in surface tension greatly affected the water (thus air) saturation. Since the level of air saturation is a critical parameter for the removal of VOCs or NAPL components from aquifer during an air sparging process, this observation and supporting principles provide an approach for enhancing the efficiency of air sparging technique. Sand Column Experiments. The objective of the sand column experiments was to investigate the air saturation change within the sparging zone at varying surface tensions. Since the air injected through the bottom of the column using a metal tube was confined within the column of relatively small cross-sectional area, the entire sand packed in the column was considered to be within the influence zone. During sparging experiments, it was possible to observe the air in the sand layer throughout the circumference of the Plexiglas column, supporting the assumption that the air displaced water across the cross-section of the column. The average air saturation of the sand column was estimated based on the volume of water displaced by air. At a fixed surface tension of water, the air saturation was determined as a function of the applied airflow rate (Figure 3a). As the airflow rate increased, the air saturation increased. The rate of increasing air saturation with increasing airflow rate was more significant at reduced surface tensions. The air saturation seemed to approach a constant value at each surface tension. The airflow rate at which the air saturation reached a maximum decreased as the surface tension decreased. At a fixed airflow rate, the air saturation was found to be inversely proportional to the surface tension (Figure
FIGURE 3. Results of sand column experiments: (a) effects of airflow rate and surface tension of water on the air saturations and (b) measured air pressures (pg) during sparging experiments. G, density of water; g, gravitational acceleration.
FIGURE 5. Extent of the unsaturated (sparging) zone at different surface tension of water measured after the boundary between saturated and unsaturated zones was stabilized. Solid lines represent the air channels visible through the Plexiglas wall of the box. Airflow rate was set at 400 cm3/min. Surface tension (N/m): (a) 7.16 × 10-2, (b) 5.68 × 10-2, (c) 5.24 × 10-2, and (d) 4.60 × 10-2.
FIGURE 4. Functional relationship between surface tension of water and air saturation estimated at an airflow rate of 15 cm3/min. Air saturation was estimated by fitting the curves shown in Figure 3a. 4). This effect of reducing surface tension on the air saturation was already demonstrated in the soil water retention curves, although the Tempe cell experiments were conducted under static conditions. For the air to be injected into a water-saturated porous medium and to form a continuous air phase, the injection pressure (pg) has to overcome the static pressure of water (pws), the pressure associated with the capillarity of the medium (pca), and the pressure losses within the system due to airflow through the system:
pg ) pws + pca + pl + pf + ps
(4)
where pl, pf, and ps are pressure losses due to airflow through the tubing, the air diffuser, and the air channels in the medium, respectively. The air injection pressures measured during the air sparging experiments are shown in Figure 3b. Since the height of the water-saturated sand was 73 cm, pws was 73 cmH2O pressure head and is shown as a horizontal line at the bottom of the graph. For the water column with no sand (73 cm water column), the required air pressure head for air to flow through the column was 88 cmH2O, which is the sum of static pressure (pws), the pressure losses due to
airflow through air tubing, and the diffuser (tubing tip). With no sand, the pressure head (88 cmH2O) was not sensitive to the flow rate change. All the measured air pressure heads for sand columns regardless of the presence of surfactant were higher than 88 cmH2O. At a given surface tension, a slight increase of the air pressure was observed with increasing airflow rate. However, the effect of surface tension on the air pressure for the sand columns was minimal. This is because two opposing effects on the air pressure operated at the same time as the surface tension changed. With decreasing surface tension, pca and pf decreased, while the ps increased due to newly formed air channels in the sand. Thus, decreasing surface tension does not significantly increase the pressure loss but increases the air saturation significantly. Sand Box Experiments. Two sets of box experiments were performed with different sand packings. The first set up was for homogeneous media using the fine sand. With decreasing surface tension, a dramatic change in the sparging area was observed (Figure 5). When no surfactant was applied, the air channels appeared clearly between the diffuser and the water table. However, it was somewhat difficult to delineate the water-unsaturated zone, implying that most of the injected airflow was through narrow air channels without entering small pores in the sand. When the surface tension decreased to 5.68 × 10-2, 5.24 × 10-2, and 4.60 × 10-2 N/m, the sparging area increased 2.0, 4.1, and 5.2 times, respectively, over that observed with surfactant-free water. At low surface tensions, air channels were not observable in the vicinity of the diffuser due to apparently high air saturation, while fine air channels were formed in greater numbers at the upper portion of the sand. At reduced surface tension, the unsaturated zone was VOL. 38, NO. 4, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1173
FIGURE 6. Expansion of unsaturated zone by air sparging in sand box with a preferential air channel presents at different surface tensions.
water flow rate. Calculated values of NT can be used to estimate the range of conditions under which NAPL mobilization can be avoided as the interfacial (surface) tension is to be modified. At the same time care must be taken in selecting the airflow rate applied to a NAPL-contaminated aquifer not to generate significant changes in soil structure that might cause different patterns in the airflow reducing the effectiveness of this technique. If the air-water interfacial tension in an aquifer can be successfully reduced without causing NAPL migration at a NAPL-contaminated site, the air injected into the contaminated zone will desaturate a much larger area. Increased air saturation will enhance the mass transfer rate of volatilization from NAPL to the mobile air phase at a microscopic scale, while the expansion of unsaturated zone means a larger influence zone of air sparging at a macroscopic scale. Under hydrogeological conditions where preferential air channels are present within the sparging zone, reducing surface tension might be particularly helpful to expand airflow further in the aquifer accessing more of the NAPL-contaminated zone. This technique can be easily implemented in the field by injecting a predetermined volume of aqueous surfactant solution into a NAPL-contaminated aquifer through an injection well prior to air sparging.
Acknowledgments clearly observable, which is shown in white color in Figure 5 particularly in the lower side of the sand layer. For the second set of box experiments, a vertical strip of coarse sand was installed at the center of the sand box, which was expected to serve as a preferential flow path for the injected air. With no surfactant applied, injected air traveled only through the vertical layer of coarse sand, indicating that this strip indeed was a preferential flow path of air during air sparging. However, the injected air was found to intrude the fine sand domain at decreased surface tensions, resulting in much larger sparging area (Figure 6). The sparging area was about 4.5 times larger at the surface tension 3.84 × 10-2 N/m and an estimated 11 times larger at 2.89 × 10-2 N/m as compared to the sparging area with no surfactant. The observations presented here provide useful implications for the remediation of NAPL-contaminated aquifers using air injection techniques. Injection of an aqueous solution of surfactant to displace the resident groundwater could greatly enhance the removal of dissolved VOCs by air sparging. Greater desaturation of water in the NAPLcontaminated region will provide better contact between mobile air and NAPL. Removing aqueous barriers to vapor extraction could extend the method for more complex sites and increase the efficiency of mass extraction and the effectiveness of the technique. In the case of DNAPLs, the surfactant solution can help establish or expand desaturated cavities generated deep below the water table by injection and extraction of air at adjacent wells or by pulses of air injected and extracted at a single location. During surfactant enhanced air sparging, NAPLs entrapped in soil pores could migrate with the injected surfactant solution when the interfacial tension between NAPL and the surfactant solution became low enough for water flow or gravity to force the blobs or ganglia of NAPL to move (29). To minimize expansion of NAPL-contaminated zone, the concentration of surfactant should be carefully selected. The mobilization of residual NAPL in porous media may require low interfacial tension between NAPL and the aqueous solution generally associated with high surfactant concentration. A parameter called the Trapping number (NT) was introduced by Pennell et al. (29) to successfully correlate NAPL mobilization with aquifer conditions including the NAPL-water interfacial tension, fluid viscosities, and pore 1174
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 4, 2004
The authors thank Se-Hoon Park, Sue-Ah Hong, and KangHyun Kim for conducting part of the experiments. This research was supported by the Korea Science and Engineering Foundation (Grant R01-2000-000-00057-0). This work was also supported by a research grant from Hallym University, Korea.
Literature Cited (1) Loden, M. E. A technology assessment of soil vapor extraction and air sparging; Office of Research and Development, U.S. Environmental Protection Agency: Washington, DC, 1992; EPA/ 600/R-92/173. (2) U.S. EPA. A citizen’s guide to air sparging; Office of Solid Waste Emergency Response, U.S. Environmental Protection Agency: Washington, DC, 1992; EPA/542/F-92/010. (3) Hinchee, R. E. In Air Sparging for Site Remediation; Hinchee, R. E., Ed.; Lewis Publishers: Boca Raton, FL, 1994; pp 113. (4) Johnson, R. L.; Johnson, P. C.; McWhorter, D. B.; Hinchee, R. E.; Goodman, I. Ground Water Monit. Rem. 1993, 13, 127135. (5) McCray, J. E.; Falta, R. W. J. Contam. Hydrol. 1996, 24, 2552. (6) Rogers, S. W.; Ong, S. K. Environ. Sci. Technol. 2000, 34, 764770. (7) Reddy, K. R.; Adams, J. A. J. Environ. Eng. 1998, 124, 288299. (8) Chao, K. P.; Ong, S. K.; Protopapas, A. J. Environ. Eng. 1998, 124, 1054-1060. (9) Johnson, P. C.; Das, A.; Bruce, C. Environ. Sci. Technol. 1999, 33, 1726-1731. (10) Johnson, P. C.; Johnson, R. L.; Neaville, C.; Hansen, E. E., Stearns, S. M.; Dortch, I. J. Ground Water 1997, 35, 765-774. (11) Aelion, C. M.; Kirtland, B. C. Environ. Sci. Technol. 2000, 34, 3167-3173. (12) Johnston, C. D.; Rayner, J. L.; Briegel, D. J. Contam. Hydrol. 2002, 59, 87-111. (13) Ji, W.; Dahmani, A.; Ahlfeld, D. P.; Lin, J. D.; Hill, E. Ground Water Monit. Rem. 1993, 13, 115-127. (14) Unger, A. J. A.; Sudicky, E. A.; Forsyth, P. A. Water Resour. Res. 1995, 31, 1913-1925. (15) Hein, G. L.; Gierke, J. S.; Hutzler, N. J.; Falta, R. W. Ground Water Monit. Rem. 1997, 17, 222-230. (16) McCray, J. E.; Falta, R. W. Ground Water 1997, 35, 99-110. (17) Philip, J. R. J. Contam. Hydrol. 1998, 22, 337-345. (18) Van Dyke, M. I.; van der Zee, E. A. Water Resour. Res. 1998, 34, 341-354.
(19) Elder, C. R.; Benson, C. H.; Eykholt, G. R. J. Geotech. Geoenviron. Eng. 1999, 125, 947-958. (20) Burns, S. E.; Zhang, M. Environ. Sci. Technol. 2001, 35, 204208. (21) Anwar, A. H. M. F.; Bettahar, M.; Matsubayashi, U. J. Contam. Hydrol. 2000, 43, 129-146. (22) Dane, J. H.; Hopmans, J. W. In Methods of Soil Analysis, Part 4, Physical Methods; Dane, J. H., Topp, G. C., Eds.; SSSA Book Series 5; Soil Science Society of America: Madison, WI, 2002; pp 684-688. (23) Harkins, W. D.; Brown, F. E. J. Am. Chem. Soc. 1919, 41, 499524. (24) Corey, A. T. Mechanics of immiscible fluids in porous media; Water Resources Publications: Highland Ranch, CO, 1994.
(25) Fetter, C. W. Contaminant Hydrology; Prentice Hall: Upper Saddle River, NJ, 1999; pp 210-212. (26) Brooks, R. H.; Corey, A. T. J. Irrig. Drain. 1966, 92, 6168. (27) van Genuchten, M. Th. Soil Sci. Soc. Am. J. 1980, 44, 892898. (28) Lenhard, R. J.; Parker, J. C. J. Contam. Hydrol. 1987, 1, 407424. (29) Pennell, K. D.; Pope, G. A.; Abriola, L. M. Environ. Sci. Technol. 1996, 30, 1328-1335.
Received for review July 11, 2003. Revised manuscript received November 10, 2003. Accepted November 12, 2003. ES030547O
VOL. 38, NO. 4, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1175
