HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Oostrom, M. Articles by Wietsma, T. W. Search for Related Content PubMed Articles by Oostrom, M. Articles by Wietsma, T. W. GeoRef GeoRef Citation Agricola Articles by Oostrom, M. Articles by Wietsma, T. W. Related Collections Volatile Organic Compounds Experiment Design

ORIGINAL RESEARCH

Removal of Carbon Tetrachloride from a Layered Porous Medium by Means of Soil Vapor Extraction Enhanced by Desiccation and Water Table Reduction

^a Environmental Technology Division, Pacific Northwest National Lab., P.O. Box 999, MS K9-33, Richland, WA 99354
^b Dep. of Agronomy and Soils, Auburn Univ., Auburn, AL 36849-5412
^c Environmental Molecular Sciences Lab., Pacific Northwest National Lab., P.O. Box 999, MS K8-96, Richland, WA 99354
^* Corresponding author (mart.oostrom{at}pnl.gov)

Received 3 December 2004.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
A two-dimensional flow cell experiment was conducted to study the removal of the carbon tetrachloride component of a DNAPL mixture from a layered porous medium through soil vapor extraction (SVE) with moist and dry air. A dual-energy radiation system was used at various times to non-intrusively determine fluid saturations. The mixture, which contained the volatile organic carbon tetrachloride, mimics the DNAPL disposed at the Hanford Site in Washington State. The flow cell, which is 100 cm long, 75 cm high and 5.5 cm wide, was packed with two sloped coarse sand and two sloped silt layers in an otherwise uniform matrix of medium-grained sand. A V-shaped fine sand layer was placed at the bottom of the flow cell to prevent DNAPL from exiting the flow cell. The water table was located 2 cm from the bottom, creating variably saturated conditions. A 500-mL spill was introduced at the top of the flow cell from a small source area. It was observed that the DNAPL largely by-passed the silt layers but easily moved into the coarse sand layers. Residual DNAPL was formed in the medium-grained sand matrix. The DNAPL caused a distinct reduction of the capillary fringe. Most of the DNAPL ended up in a pool on top of the V-shaped fine sand. Through four treatments with moist air soil vapor extraction, most residual carbon tetrachloride was removed from the medium-grained matrix and the coarse sand layers. However, soil vapor extraction with moist air was not able to remove the carbon tetrachloride from the silt layers and the pool. Through a water table reduction and subsequent soil vapor extraction with dry air, the carbon tetrachloride in the silt layers and the pool was effectively removed. Based on measurements and carbon tetrachloride vapor concentration data, it was estimated that after the final remediation treatment, almost 90% of the total mass was removed.

Abbreviations: DBBP, dibutyl butyl phosphonate • DNAPL, dense non-aqueous phase liquid • DSVE, dry soil vapor extraction • PCE, perchloroethylene • SVE, soil vapor extraction • TBP, tributyl phosphate • TCE, trichloroethylene • VOC, volatile organic compound • WSVE, wet soil vapor extraction

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
SOIL VAPOR EXTRACTION is an in situ soil-cleaning process designed to remove volatile organic compounds (VOCs) from the unsaturated (vadose) zone of soil (Mercer and Cohen, 1990; Stamnes and Blanchard, 1997). The process has been used at many contaminated sites with considerable success. For example, at the Hanford Site in Washington State, more than 70 000 kg of carbon tetrachloride have been removed through a SVE campaign using up to 46 wells (Rohay, 2002). At this site, an estimated 580 m³ of a DNAPL mixture, containing about 70% carbon tetrachloride, was disposed to waste sites. Although SVE applications have yielded considerable amounts of VOCs in the field, high initial recovery rates are usually followed by long periods of low rates, commonly referred to as tailing. Yoon et al. (2003) listed several reasons for observed tailing behavior including preferential by-passing of DNAPL in unswept regions of low permeability, limited access of the vapor phase to trapped DNAPL in water, slow desorption from soils, rate-limited dissolution and volatilization, and retardation of volatile components in low moisture soils.

Some of the more commonly occurring VOCs that have contaminated soils and groundwater are the dense non-aqueous phase liquids (DNAPLs) trichloroethylene (TCE), perchloroethylene (PCE) and carbon tetrachloride (Travis and MacInnis, 1992). These contaminants can occur in the liquid phase as free, residual, and/or trapped (Lenhard et al., 2004); in the solid phase by means of sorption, especially to organic matter; in the aqueous phase when dissolved; and/or in the gas phase (Armstrong et al., 1994). Mass transfer between the different phases has been described based on the local equilibrium assumption or by a first-order kinetic approach (Fischer et al., 1996).

Substantial tailing in the gaseous concentration has been observed when free liquid DNAPL is no longer present (Grathwohl et al., 1990; Gierke et al., 1992). Fischer et al. (1996) attributed tailing solely to diffusion in inter-particle water which suggested that local non-equilibrium conditions prevailed. Because of local non-equilibrium conditions and the reduced efficiency in the removal of VOCs during tailing, it has been suggested that pulsed pumping be considered (Crow et al., 1987; Gierke et al., 1992). However, Armstrong et al. (1994) showed that for their experimental conditions pulsed pumping was less efficient than continuous pumping, because the latter maximizes the diffusive transfer of the contaminant from the water to the gas phase by maintaining the concentration gradient at a steady maximum value.

For homogeneous porous media, and as long as the VOCs only occur in the unsaturated part of the soil, vapor extraction can remove close to 100%. For heterogeneous systems, Liang and Udell (1999) showed that the process of through-flow venting is responsible for removal of contaminants in the high-permeability regions of the vadose zone. In that process, air preferentially moves through the high-permeability zones and the removal rates of VOCs are primarily determined by their volatilities. Evaporation rates are relatively high and mass transfer resistances low. On the other hand, VOCs located in low-permeability zones are bypassed by the moving air. Recovery rates in this case are governed by diffusion, through the aqueous or gas phase, from low to high permeability zones. Liang and Udell (1999) refer to this process, where evaporation rates are small, as bypass-flow venting. In support of these findings by Liang and Udell (1999), Chu et al. (2004) demonstrated through magnetic resonance imaging of NAPL during SVE that diffusion limits removal or residual NAPL from low-permeability zones. Evaporation rates of NAPLs from a layered system with coarse- and fine-grained dry sands were conducted by Ho and Udell (1992). They found that, although gas flowed preferentially through the coarse sand, NAPL movement as a result of capillary action occurred from the fine to the coarse sand. This result showed that NAPL flow could influence mass removal during SVE in dry porous media. A general discussion of the effects on soil heterogeneity on SVE can be found in Suthersan (1997).

Remaining challenges in the removal of DNAPLs from porous media include the removal of trapped DNAPL from the vadose zone and the removal of DNAPL present below the capillary fringe, that is, the region where only DNAPL and water are present. Solutions suggested for the removal below the capillary fringe and trapped contaminants in the vadose zone include the use of surfactants to mobilize trapped globules and to increase the dissolution into the aqueous phase (Walker et al., 1998; Oostrom et al., 1999); air sparging, which increases aerobic biodegradation and promotes the physical removal of organics by direct volatilization (Stamnes and Blanchard, 1997); biodegradation, in which enhanced microbial action is used to degrade organic contaminants into harmless metabolic products (Brady and Weil, 2002); chemical reactions to destroy the NAPL (Schroth et al., 2001); and thermal treatment (Gudbjerg et al., 2004). Sometimes combinations of the above are being used (Zhang and Miller, 1992; Stamnes and Blanchard, 1997).

All of the above recommendations still suffer from technical problems. The use of surfactants often results in the creation of micro-emulsions and dense aqueous phase liquids, both of which tend to move downward because their density is greater than the ambient fresh water. Consequently the contaminated region will be enlarged. The delivery of surfactant solutions, or needed bacteria and the required nutrients, to the proper locations remains a difficult problem. The much higher flow rates that can be established with air compared to water seem to favor the use of air. As with water, preferential flow channels may need to be closed off to let the air flow through those parts where it is needed. Because both DNAPLs and air are non-wetting fluids, this may, however, be less of a problem than with water flow. When studying extraction of VOCs, it is important to realize that the vapors often differ in density from the ambient air. This may result in density-driven advection and hence in VOC distributions that differ from those if only diffusion is considered (Lenhard et al., 1995).

For the removal of medium- to high volatile contaminants in the vadose zone, isothermal SVE is considered to be sufficient since the vapor pressures at ambient temperatures are high enough for recovery (Kaluarachchi and Mebah-Ul Islam, 1995). For removal of less-volatile organics or volatile organics in low permeability zones or pools, thermal treatments such as steam injection (Falta et al., 1992; Gudjberg et al., 2004; Schmidt et al., 2002) and thermal heating (Kaluarachchi and Mebah-Ul Islam, 1995) have been used. Thermal treatments may be regarded as modifications to SVE since removal of contaminants occurs, at least, partly through the gas phase. Enhanced remediation might also be obtained when dry air is injected during SVE operations instead of moist air. Injection of dry air at ambient temperature leads to desiccation (Rossi and Nimmo, 1994; Webb, 2000) resulting in a potential increased NAPL accessibility for volatilization. Yoon et al. (2003) demonstrated through detailed one-dimensional modeling exercises that, even when dry air is injected at ambient conditions, a temperature decrease can be expected due to the heat exchange needed to volatilize NAPL and liquid water. However, it is likely that in multi-dimensional systems, water vapor and heat transport from adjacent zones will reduce the cooling that is observed in one-dimensional systems.

Although SVE is one of the most applied remediation techniques for removal of VOCs from the vadose zone, to our knowledge no intermediate-scale flow cell experiment has been conducted to demonstrate the features of the technique in a heterogeneous porous medium contaminated with a DNAPL mixture. The experimental work with liquid VOCs related to SVE so far involved one-dimensional columns (Baehr et al., 1989; Lingineni and Dhir, 1992; Ho and Udell, 1995). A two-dimensional experiment discussed by Fischer et al. (1996) did not involve liquid VOCs in the system.

Based on some of our knowledge gaps and desirability to enhance our understanding of DNAPL fate in the environment and subsequent cleanup using SVE, we formulated the following objectives for the intermediate-scale experiment: (i) investigate the infiltration and redistribution of a multicomponent DNAPL in a variably saturated, layered porous medium, (ii) verify whether the expected through-flow and bypass-flow principles of SVE would occur, (iii) investigate the effectiveness of residual carbon tetrachloride removal from the vadose zone by means of SVE with moist air; (iv) investigate whether DNAPL located in the vadose zone and pooled DNAPL present in the saturated zone can be removed by the combined effects of water table reduction and SVE with dry air; and (v) develop a data set that can be used to test multifluid flow numerical simulators.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
The experiment was performed in a 100-cm long by 80-cm high by 5.5-cm wide flow cell. It was packed with four porous media, viz., 40/50 mesh Accusand, 70 mesh Accusand, and a silt and coarse sand obtained at the Hanford site. Relevant properties of the sands and soil materials are presented in Table 1. The porosity was determined with the system as part of the calibration procedure. The permeability was obtained using a constant-head method (Klute and Dirksen, 1986). The Brooks and Corey (1964) retention parameters were obtained using a saturation-pressure apparatus technique (Lenhard and Parker, 1988) followed by an analysis with the RETC program (Van Genuchten, 1985). The flow cell and the packing arrangement are shown diagrammatically in Fig. 1 . It should be noted that the V-shaped bottom layer consisted of a fine sand (70 Accusand) to prevent DNAPL from reaching the bottom of the flow cell. The bulk of the remaining material consisted of a medium sand (40/50 Accusand) in which two coarse and two fine sloped layers were embedded. All porous materials were either packed under water or under a slight suction to allow the proper slopes to be implemented. Before the cover was placed on the flow cell, a 2-cm layer of kaolinite paste was placed on top of the porous medium to ensure that no preferential air flow could occur between the porous medium and the flow cell cover. The flow cell itself was constructed from glass, stainless steel, and Teflon gaskets. The water table level in the flow cell was controlled by two end chambers, each of which was connected to a constant head reservoir. Each reservoir contained an overflow. Excess water spilled into a buffer tank from which water was returned to each reservoir at a regulated flow rate. The end chambers were separated from the porous medium chamber by fine-meshed screens. The porous medium chamber contained two drains which were connected to pressure transducers to determine the height of the water table.

View larger version (20K): [in this window] [in a new window]   Fig. 1. Overview of intermediate-scale flow cell (i.l. = injection location; e.l. = extraction location).

The spill was introduced under a constant head from a source area located on top of the 40/50 mesh Accusand (Fig. 1). The liquid was poured into a rectangular 5.5 by 2.5-cm stainless steel sleeve, which covered the 5.5-cm inside width of the flow cell, while maintaining a head of approximately 3 cm. The DNAPL mixture used in the experiment is representative of the approximately 750 m³ of DNAPL disposed at the Hanford site between 1955 and 1972 (Rohay, 2002). The mixture consisted of 70% (by volume) carbon tetrachloride, 8% tributyl phosphate (TBP), 14% dibutyl butyl phosphonate (DBBP), 3% Peacock lard oil, and 5% iodo-heptane. All components of the mixture, expect for carbon tetrachloride, are considered to be nonvolatile. Fluid properties of this mixture are listed in Table 2. For observational purposes, the mixture was dyed with 0.01 g L^–1 Sudan IV dye. Measured properties of the DNAPL are listed in Table 2. The properties were obtained at a laboratory temperature of 22°C. The density was obtained by weighing controlled volumes. Viscosity measurements were made using a low shear falling ball viscometer (Gilmont 100). The pendant drop method was used for both the surface and interfacial tension measurements.

It was assumed that the nonvolatile DNAPL saturation (i.e., 30% of the total DNAPL saturation) remained constant in the vadose zone for the duration of the experiment. The fluid properties of the nonvolatile component mixture, consisting of 27% TBP, 46% DBBP, 10% Peacock lard oil, and 17% iodo-heptane, are listed in Table 2. The density value indicates that the nonvolatile component mixture is still a DNAPL. In terms of DNAPL saturations, the total DNAPL saturation, s^t_n, is the sum of the nonvolatile saturation, s^nv_n, and the carbon tetrachloride saturation, s^c_n. Before the SVE processes remove the volatile carbon tetrachloride, the following attenuation equation was used to compute total DNAPL and water saturations:

[1]

where I_j is the observed count rate (T^–1) for source j, I^d_j is the computed count rate through the flow cell containing dry porous medium only, U_jw and U_jn are the volumetric attenuation coefficients (L^–1) for water and the DNAPL mixture with 70% carbon tetrachloride, and x (L) is the path length through the flow cell. During stages were carbon tetrachloride evaporation reduces the DNAPL saturation, the following equation was used:

[2]

U^nv_jn and U^c_jn are the volumetric attenuation coefficients (L^–1) of the nonvolatile DNAPL mixture and carbon tetrachloride, respectively. Assuming a constant nonvolatile DNAPL saturation, Eq. [2] can be reduced to:

[3]

where I^d,nv_j is the computed count rate through the porous medium containing nonvolatile DNAPL only.

The concentration of the carbon tetrachloride in the gaseous effluent was determined as a function of time during the SVE treatments. Gas samples were collected with the use of a syringe pump. The 3-mL samples were trapped on charcoal sorbent tubes (Anasorb CSC Sample Tubes, SKC Inc., Eighty Four, PA). The sorbent tubes were extracted following the procedure in the National Institute for Occupational Safety and Health (NIOSH) Method No. 1003. The extracted samples were analyzed by a gas chromatograph equipped with a capillary column and a flame ionization detector.

Finally, a network of 15 gas extractors (perforated needles; Fig. 1) was available to determine carbon tetrachloride vapor distributions after the spill and during remediation. Only those extractors (#1, 3, 4, 5, 6, 9, and 10) that were present in the unsaturated parts were actually used to obtain vapor samples, that is, extractors located in the saturated silt layers and below the capillary fringe could not be used. During the desiccation part of the experiment additional extractors became available to extract gas samples.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
DNAPL Infiltration and Redistribution
A picture of the aqueous phase saturation conditions before the spill, including the position of the water table, is shown in Fig. 2 . The picture shows that the capillary fringe in the 40/50 mesh Accusand is approximately 20 to 25 cm. This corresponds well with the value of 19.5 cm found by Schroth et al. (1996) for the same sand. A scan for this situation is presented in Fig. 3 . The scan indicates that the coarse sand layers have drained to saturations close to irreducible saturation and that the silt layers are, except for the edges, still fully saturated. The partial drainage at the edges may be a result of the packing procedures. To avoid smearing of the interfaces between the silt and the 40 to 50 mesh sand, the sides and the upper interface of each silt layer was packed with slightly less intensity than the internal part and the bottom interface. As a result, larger pores might have been present near the sides and upper interface, compared to the interior and lower interface, allowing some desaturation at the prevalent air-water capillary pressure. A series of five photos displaying the infiltration and redistribution process of the DNAPL is presented in Fig. 4a to 4e . The first picture (Fig. 4a) clearly demonstrates the effect of capillary action in the unsaturated zone. Initially, the DNAPL infiltrated very much as a sphere, similarly to the infiltration of water from a point source into a dry soil. A similar wetting behavior for the infiltration of PCE into an unsaturated zone was reported by Hofstee et al. (1998). Once the DNAPL reached the upper silt layer it started to spread to the left and to the right and edged over both ends. Only some minor infiltration into the upper part of the silt layer was observed at locations that were unsaturated. The lack of entry into the silt is the result of its relatively large nonwetting-fluid entry pressure. Apparently the capillary forces were still dominant as the DNAPL moved over the higher end of the fine layer. During the subsequent redistribution, the DNAPL continued to move more or less straight downward on the left side of the flow cell (high ends of the layers), but followed a more intriguing path at the lower ends. The first coarse sand layer, which was unsaturated, accepted the DNAPL after some spreading occurred at the boundary. The spreading occurred because of the larger pores in the coarse sand causing the interface to act somewhat as a barrier. After sufficient DNAPL collected on top of the interface, DNAPL was able to infiltrate into the coarse sand. The phenomena observed at this interface again indicate that the DNAPL moved as an intermediate wetting fluid in the unsaturated zone.

View larger version (103K): [in this window] [in a new window]   Fig. 2. Photograph of initial water distribution.

View larger version (38K): [in this window] [in a new window]   Fig. 3. Measured initial water saturations.

View larger version (96K): [in this window] [in a new window]   Fig. 4. Photographs of DNAPL infiltration and redistribution.

The final DNAPL distribution, for which we assumed static equilibrium, is shown in Fig. 4e. It clearly shows the collapse of the capillary fringe, which was attributed to a reduction of the surface tension of water caused by carbon tetrachloride diffusion in the liquid phase and by advective vapor movement in the gas phase. This picture also shows the accumulation of DNAPL in the V-shaped region below the capillary fringe, where two phases exist, and the distribution of the DNAPL at higher elevations, where three phases exist. This distribution is the initial condition for the remediation process using SVE. A scan depicting the equilibrium, pre-remediation DNAPL saturations is shown in Fig. 5a . The corresponding carbon tetrachloride saturations are shown in Fig. 5b. The latter figure was obtained from Fig. 5a by multiplying the total DNAPL saturation by 0.7, representing the volume fraction of carbon tetrachloride in the mixture. The DNAPL saturations shown in Fig. 5a are in the form of residual DNAPL in the vadose zone, entrapped (by the aqueous phase) DNAPL in the capillary fringe, and as free DNAPL in a pool on top of the 70-mesh Accusand.

View larger version (28K): [in this window] [in a new window]   Fig. 5. Measured (a) total DNAPL and (b) carbon tetrachloride saturations for equilibrium distribution following the spill.

Soil Vapor Extraction with Moist Air
A total of four WSVE treatments were included with extraction times of 1, 4, 8, and 48 h, respectively (Table 3). The carbon tetrachloride saturation distributions after each treatment are shown in Fig. 6a–d . A comparison of Fig. 6a, obtained after WSVE I, with the equilibrium distribution depicted in Fig. 5b, shows that some of the carbon tetrachloride was removed from the left side of the flow cell. The saturations in the rest of the flow cell were not affected. As can be seen in Fig. 6b, all carbon tetrachloride to the left of the four layers and considerable amounts within the two coarse sand layers had been removed during WSVE II, which lasted 4 h. Most carbon tetrachloride located in the silt layers had not been volatilized. After this remediation period, there were still no through-flow venting pathways (Liang and Udell, 1999) in which the carbon tetrachloride had been completely removed. Air, moving from the injection to the extraction ports, still encountered liquid carbon tetrachloride at some points in its path. After WSVE III, an 8-h long extraction treatment, several clean through-flow venting pathways were present (Fig. 6c). In the unsaturated zone, a few pockets of carbon tetrachloride remained in the 40/50 mesh Accusand. Most of the carbon tetrachloride in the finer-grained silt layers had not yet been affected during this treatment. During WSVE IV, which lasted for 48 h, the last remaining carbon tetrachloride was removed from the unsaturated 40/50 mesh Accusand, but the carbon tetrachloride saturations in the silt layers decreased only slightly (Fig. 6d).

View larger version (36K): [in this window] [in a new window]   Fig. 6. Measured carbon tetrachloride saturations after (a) wet soil vapor extracton (WSVE) I, (b) WSVE II, (c) WSVE III, and (d) WSVE IV.

View larger version (20K): [in this window] [in a new window]   Fig. 7. Outflow carbon tetrachloride concentrations after (a) wet soil vapor extracton (WSVE) I, (b) WSVE II, (c) WSVE III, and (d) WSVE IV.

Water Table Reduction and Soil Vapor Extraction with Dry Air
After it became apparent that the WSVE was no longer effective in removing the remaining carbon tetrachloride in the capillary fringe and the silt layers, additional remediation was conducted using DSVE after a lowering of the water table from z = 2 cm to z = –10 cm. Water table lowering was a viable option in this experiment because the fine sand was located in the capillary fringe and no downward movement was expected. The carbon tetrachloride and aqueous phase saturations after the water table lowering are shown in Fig. 8a and 8b , respectively. The DSVE consisted of three treatments of 1 wk each. After each treatment a scan was obtained. Photographs of the flow cell after DSVE I, II, and III are shown in Fig. 9 . Carbon tetrachloride and water saturations after the treatments I, II, III, are shown in Fig. 10, 11, and 12 , respectively.

View larger version (31K): [in this window] [in a new window]   Fig. 8. Measured (a) carbon tetrachloride and (b) water saturations after water table reduction.

View larger version (60K): [in this window] [in a new window]   Fig. 9. Photographs of the flow cell after (a) dry soil vapor extracton (DSVE) I, (b) DSVE II, and (c) DSVE III.

View larger version (25K): [in this window] [in a new window]   Fig. 10. Measured (a) carbon tetrachloride and (b) water saturations after dry soil vapor extracton (DSVE) I.

View larger version (39K): [in this window] [in a new window]   Fig. 11. Measured (a) carbon tetrachloride and (b) water saturations after dry soil vapor extracton (DSVE) II.

View larger version (18K): [in this window] [in a new window]   Fig. 12. Measured (a) carbon tetrachloride and (b) water saturations after dry soil vapor extracton (DSVE) III.

During DSVE II, the desiccation zone had developed further to the right (Fig. 9b). Both coarse sand layers were now dry, while the upper silt layer appeared, based on the color change, to have undergone some drying too. During this treatment, the capillary fringe was reduced due to evaporation. The scan of the carbon tetrachloride saturation (Fig. 11a) shows removal (compared to Fig. 10a) at the edges of the pool and a reduction of the remaining saturations in the silt layers. During this treatment, the removed mass of carbon tetrachloride was 67 g, which was considerably more than during treatment DSVE I (Table 3). The scan of the aqueous phase (Fig. 11b) clearly shows the expanded desiccated zone, the reduction of the capillary fringe, and some drying of both the silt layers (compared to Fig. 10b). The drying of the silt layers does not show the sharp fronts observed in the sands. Instead, the rather uniform saturation reduction in the 10 to 20% range suggests that drying may have occurred between the silt and the glass walls as result of preferential gas flow. If preferential drying of the silt layers along the glass occurred, the accessible surface area was artificially increased and might have caused water movement in the direction of the glass walls due to capillary action. A volume integration of the water saturations yielded a mass removal of 463 g of water during this period. The computed mass compares favorably with the expected mass removal of 491 g based on flow rates and water vapor density. The observation that average mass removal rates during DSVE I and II are similar to the rate during the WSVE III period (Table 3), giving the impression that DSVE treatments have not led to improvements. However, it is important to note that the carbon tetrachloride removed during the DSVE periods would not have been easily removed using WSVE treatment. The WSVE treatments showed a continuous decline in recovery rates, indicating that carbon tetrachloride availability in the vadose zone for direct volatilization was minimal at the end of WSVE IV.

During the final DSVE period, dry air breakthrough was observed on the downstream end of the flow cell (Fig. 9c). The carbon tetrachloride scan (Fig. 12a) shows additional removal of the pooled carbon tetrachloride and complete removal in the silt layers. The carbon tetrachloride removal during this treatment was computed to be 146 g, corresponding to an average rate of 0.9 g h^–1, which is more than double the rate observed during DSVE II (Table 3). Some DNAPL infiltration into the fine sand was observed. The reason for this infiltration is the water desaturation of the fine sand, which was also measured with the radiation system (Fig. 12b). The movement of DNAPL into the fine sand can be viewed as an undesirable effect of the desiccation process. The total mass of carbon tetrachloride present in the flow cell after all SVE remediation treatments was, according to the final scan (Fig. 12a), 61 g, or 11% of the original mass.

	SUMMARY AND CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
A two-dimensional flow cell experiment was conducted to study the removal of the carbon tetrachloride component of a DNAPL mixture from a layered porous medium through SVE with moist and dry air. To our knowledge, intermediate-scale experiments addressing SVE remediation techniques have not yet been conducted. The DNAPL mixture mimics the DNAPL disposed at the Hanford Site in Washington State.

The flow cell was packed with two sloped coarse sand and two sloped silt layers in an otherwise matrix of medium-grained sand. A V-shaped fine sand layer was placed at the bottom of the flow cell to prevent DNAPL from exiting the flow cell. The water table was located 2 cm from the bottom, creating variably saturated conditions in the flow cell. A dual-energy radiation system was used to determine fluid saturations at several times during the experiment.

A 500-mL spill was introduced at the top of the flow cell from a small source area. It was observed that the DNAPL largely by-passed the silt layers due to their relatively large nonwetting-fluid entry pressure and low permeability, but easily moved in the coarse sand layers. Residual DNAPL was formed in the medium-grained sand matrix. The DNAPL caused a distinct reduction of the capillary fringe. Most of the DNAPL ended up in a pool on top of the V-shaped fine sand. Through four treatments with moist air soil vapor extraction, most residual carbon tetrachloride was removed from the medium-grained matrix and the coarse sand layers. The carbon tetrachloride in these high-permeability zones is in direct contact with the moving air and its volatility, in combination with the air flow velocity, determined the initial removal rate. However, the soil vapor extraction with moist air was not able to remove the carbon tetrachloride present in the silt layers and in the pool. This behavior was consistent with the through-flow venting in permeable zones and the bypass-flow venting in lower permeable zones discussed by Liang and Udell (1999). The WSVE was not able to remove carbon tetrachloride from the inundated pool located on top of the fine-grained sand in the capillary fringe. The WSVE treatments demonstrate clearly that the technique can be effectively used to remove volatile DNAPL components from permeable materials but that the method is not effective in removing DNAPL located in fine-grained materials and from inundated pools.

Through a water table reduction and subsequent soil vapor extraction with dry air, most of the carbon tetrachloride in the silt layers and in the pool could be removed. Over time, water removal through desiccation allowed increased direct contact between the flowing air and the DNAPL, allowing the removal of carbon tetrachloride through volatilization. Although most of the carbon tetrachloride located in the silt layers and the pool could be removed using DSVE, the removal process, even in a relatively simple flow cell, was not efficient. In order for the moving air to access the carbon tetrachloride, relatively large amounts of water had to be removed which was rather time consuming. In natural environments, increasing the accessibility of NAPL through water table lowering followed by desiccation might not be feasible. If the zone of interest cannot be isolated from its surroundings, water vapor movement in the gas phases and water flow due to induced capillary pressure differences will provide an almost infinite source of water.

During the final stages of the desiccation, the underlying fine sand started to desaturate, allowing some of the DNAPL to infiltrate. The infiltration of DNAPL into a fine-grained porous medium is considered to be unfavorable (Mercer and Cohen, 1990) and needs to be avoided in practical applications. Overall, almost 90% of the carbon tetrachloride was removed from the flow cell.

In a future paper, simulations of this experiment using the STOMP simulator (White and Oostrom, 2004) will be described. The extensive data set discussed in this paper will be used to test and verify the simulator. We are particularly interested in how well the simulator will be able to predict the observed NAPL infiltration and redistribution and the reduction of the capillary fringe. In addition, to which extent equilibrium mass transfer can be used to explain the removal of carbon tetrachloride during WSVE and DSVE will be investigated.

	ACKNOWLEDGMENTS

 
Pacific Northwest National Laboratory (PNNL) is operated by the Battelle Memorial Institute for the Department of Energy (DOE) under Contract DE-AC06-76RLO 1830. This research is part of the Groundwater/Vadose Zone Integration Project funded through the DOE's Richland Operations Office. The experiment described in this paper was performed in the Environmental Molecular Sciences Laboratory, a national scientific user facility sponsored by the DOE's Office of Biological and Environmental Research and located at PNNL.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 

• Armstrong, J.E., E.O. Frind, and R.D. McClellan. 1994. Nonequilibrium mass transfer between the vapor, aqueous, and solid phases in unsaturated soils during vapor extraction. Water Resour. Res. 30:355–368.[CrossRef]
• Brady, N.C., and R.R. Weil. 2002. The nature and properties of soils. p. 796–839. Soils and chemical pollution. Prentice Hall. Upper Saddle River, NJ.
• Brooks, R.H., and A.T. Corey. 1964. Hydraulic properties of porous media. Hydrology Papers 3. Colorado State Univ., Fort Collins.
• Chu, Y., C.J. Werth, A.J. Valacchi, H. Yoon, and A.G. Webber. 2004. Magnetic resonance imaging of nonaqueous phase liquid during soil vapor extraction in heterogeneous porous media. J. Contam. Hydrol. 73:15–37.[CrossRef][ISI][Medline]
• Crow, W.L., E.P. Anderson, and E.M. Minugh. 1987. Subsurface venting of vapors emanating from hydrocarbon products in the ground water. Ground Water Monit. Rev. 7:54–57.
• Dane, J.H., M. Oostrom, and B.C. Missildine. 1992. An improved method for the determination of capillary pressure-saturation curves involving TCE, water and air. J. Contam. Hydrol. 11:69–81.[CrossRef]
• Falta, R.W., K. Pruess, I. Javandel, and P.A. Witherspoon. 1992. Numerical modeling of steam injection for the removal of nonaqueous phase liquids from the subsurface. 1. Numerical formulation. Water Resour. Res. 28:433–449.[CrossRef]
• Fischer, U., R. Schulin, and M. Keller. 1996. Experimental and numerical investigation of soil vapor extraction. Water Resour. Res. 32:3413–3427.[CrossRef]
• Gierke, J.S., N.J. Hutzler, and D.B. McKenzie. 1992. Vapor transport in columns of unsaturated soil and implications for vapor extraction. Water Resour. Res. 28:323–335.
• Grathwohl, P., J. Farrell, and M. Reinhard. 1990. Desorption kinetics of volatile organic contaminants from aquifer materials. p. 343–350. In F. Arendt et al. (ed.) Contaminated soil '90. Kluwer Academic, Norwell, MA.
• Gudbjerg, J., T.O. Sonnenborg, and K.H. Jensen. 2004. Remediation of NAPL below the water table by steam-induced heat conduction. J. Contam. Hydrol. 72:207–225.[CrossRef][ISI][Medline]
• Ho, C.K., and K.S. Udell. 1992. An experimental investigation of air venting of volatile liquid hydrocarbon mixtures from homogeneous and heterogeneous porous media. J. Contam. Hydrol. 11:291–316.
• Ho, C.K., and K.S. Udell. 1995. Mass transfer limited drying of porous media containing an immobile binary liquid mixture. Int. J. Heat Mass Transfer 38:339–350.[CrossRef]
• Hofstee, C., M. Oostrom, J.H. Dane, and R.C. Walker. 1998. Infiltration and redistribution of perchloroethylene in partially saturated, stratified porous media. J. Contam. Hydrol. 34:293–313.[CrossRef]
• Kaluarachchi, J.J., and K.M. Mesbah-Ul Islam. 1995. Thermal venting to recover less-volatile hydrocarbons from the unsaturated zone. 1. Theory. J. Contam. Hydrol. 17:293–311.[CrossRef]
• Klute, A., and C. Dirksen. 1986. Hydraulic conductivity and diffusivity: Laboratory methods. p. 687–734. In A. Klute (ed.) Methods of soil analysis. Part 1. ASA and SSSA, Madison, WI.
• Lenhard, R.J., and J.C. Parker. 1988. Experimental validation of the theory of extending two-phase saturation-pressure relations to three-fluid phase systems for monotonic drainage paths. Water Resour. Res. 24:373–380.[CrossRef]
• Lenhard, R.J., M. Oostrom, C.S. Simmons, and M.D. White. 1995. Investigation of density-dependent gas advection of trichloroethylene: Experiment and a model validation exercise. J. Contam. Hydrol. 19:47–67.
• Lenhard, R.J., M. Oostrom, and J.H. Dane. 2004. A constitutive model for air-NAPL-water flow in the vadose zone accounting for immobile, non-occluded (residual) NAPL in strongly water-wet porous media. J. Contam. Hydrol. 71:262–282.
• Liang, H.-C., and K.S. Udell. 1999. Experimental and theoretical investigation of vaporization of liquid hydrocarbon mixtures in water-wetted porous media. Water Resour. Res. 35:635–649.[CrossRef]
• Lingineni, S., and V.K. Dhir. 1992. Modeling of soil venting processes to remediate unsaturated soils. J. Environ. Eng. 118:135–152.
• Mercer, J.W., and R.M. Cohen. 1990. A review of immiscible fluids in the subsurface: Properties, models, characterization and remediation. J. Contam. Hydrol. 6:107–163.
• Oostrom, M., J.H. Dane, B.C. Missildine, and R.J. Lenhard. 1995. Error analysis of dual-energy gamma radiation measurements. Soil Sci. 160:28–42.
• Oostrom, M. C. Hofstee, J.H. Dane, and R.J. Lenhard. 1998. Single-source gamma radiation procedures for improved calibration and measurements in porous media. Soil Sci. 163:646–656.[CrossRef]
• Oostrom, M., C. Hofstee, R.C. Walker, and J.H. Dane. 1999. Movement and remediation of trichloroethylene in a saturated, heterogeneous porous medium. 2. Pump-and-treat and surfactant flushing. J. Contam. Hydrol. 37:179–197.[CrossRef]
• Rohay, V.J. 2002. Performance evaluation report for soil vapor extraction operations at the carbon tetrachloride site. Feb. 1991–Sept. 2001. BHI-00720, Rev. 6. Bechtel Hanford, Inc., Washington.
• Rossi, C., and J.R. Nimmo. 1994. Modeling of soil water retention from saturation to oven dryness. Water Resour. Res. 30:701–708.[CrossRef]
• Schmidt, R. J. Gudbjerg, T.O. Sonnenborg, and K.H. Jensen. 2002. Removal of NAPLs from the unsaturated zone using steam: Prevention of downward migration by injecting mixtures of steam and air. J. Contam. Hydrol. 55:233–260.[CrossRef][ISI][Medline]
• Schroth, M.H., S.J. Ahearn, J.S. Selker, and J.D. Istok. 1996. Characterization of Miller-similar sands for laboratory hydrologic studies. Soil Sci. Soc. Am. J. 60:1331–1339.[Abstract/Free Full Text]
• Schroth, M.H., M. Oostrom, T.W. Wietsma, and J.D. Istok. 2001. In situ oxidation of trichloroethene by permanganate: Effects on porous medium hydraulic properties. J. Contam. Hydrol. 50:79–98.[CrossRef][ISI][Medline]
• Stamnes. R., and J. Blanchard. 1997. Engineering forum issue paper: Soil vapor extraction implementation experiences. Publ. 9200.5–223FS. USEPA, Washington, DC.
• Suthersan, S.S. 1997. Remediation engineering: Design concepts. Geraghty & Miller Environ. Sci. and Eng. Series. CRC Press, Boca Raton, FL.
• Travis, C.C., and J.M. MacInnis. 1992. Vapor extraction of organics from subsurface soils. Is it effective? Environ. Sci. Technol. 26:1885–1887.[CrossRef]
• van Genuchten, M.Th. 1985. RETC.F77. Program to analyze observed soil water tension and hydraulic conductivity data. U.S. Salinity Lab. Spec. Rep. U.S. Salinity Lab., Riverside, CA.
• Walker, R.C., C. Hofstee, J.H. Dane, and W.E. Hill. 1998. Surfactant enhanced removal of PCE in a partially saturated, stratified porous medium. J. Contam. Hydrol. 34:31–46.
• Webb, S.W. 2000. A simple extension of two-phase characteristic curves to include the dry region. Water Resour. Res. 36:1425–1430.[CrossRef]
• White, M.D., and M. Oostrom. 2004. STOMP subsurface transport over multiple phases. Version 3.1. User's guide, PNNL-14478. Pacific Northwest Natl. Lab., Richland, Washington.
• Yoon, H., A.J. Valocchi, and C.J. Werth. 2003. Modeling the influence of water content on soil vapor extraction. Available at www.vadosezonejournal.org. Vadose Zone J. 2:368–381.[Abstract/Free Full Text]
• Zhang, Y., and R.M. Miller. 1992. Enhanced octadecane dispersion and biodegradation by a pseudomonas rhamnolipid surfactant (biosurfactant). Appl. Environ. Microbiol. 58:3276–3282.[Abstract/Free Full Text]

This article has been cited by other articles:

				M. Oostrom, J. H. Dane, and T. W. Wietsma A Review of Multidimensional, Multifluid, Intermediate-Scale Experiments: Flow Behavior, Saturation Imaging, and Tracer Detection and Quantification Vadose Zone J., August 23, 2007; 6(3): 610 - 637. [Abstract] [Full Text] [PDF]

				M. Oostrom, J. H. Dane, and T. W. Wietsma A Review of Multidimensional, Multifluid Intermediate-Scale Experiments: Nonaqueous Phase Liquid Dissolution and Enhanced Remediation Vadose Zone J., May 26, 2006; 5(2): 570 - 598. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Oostrom, M. Articles by Wietsma, T. W. Search for Related Content PubMed Articles by Oostrom, M. Articles by Wietsma, T. W. GeoRef GeoRef Citation Agricola