Horizontal and Vertical Well Comparison for In Situ Air Sparging

A laboratory study was conducted to determine the effectiveness ol vertical and horizontal well configurations for ground water remediation using in situ air sparging. A lexan lank was designed and constructed to allow both the visualization of air flow and quantitative measurement of the distribution of air flow. Two media, sand and glass beads. were tested with both Vertical and horizontal air sources. In each case, most of the air traveled through preferential channels as continuous flow rather than as discrete bubbles as reported in other studies. Liven though glass beads were selected to have the same grain-size distribution as the sand, air flow was quite different through the two media. Results show that glass beads are not a suitable material for modeling air flow through natural sediments. In this study, the horizontal well proved to be more effective than the vertical well by impacting more of the media with a uniform distribution of air throughout the media. The vertical well resulted in a nonuniform distribution of air flow with most of the air concentrated directly above the well.     

Monitoring Air Sparging Using Resistivity Tomography

Air sparging is a relatively new technique for the remediation of ground water contaminated with petroleum hydrocarbons. In this technique, air is injected below the water table, beneath the contaminated soil. Remediation occurs by a combination of contaminant partitioning into the vapor phase and enhanced biodegradation. The air is usually removed by vacuum extraction in the vadose zone.
The efficiency of remediation from air sparging is a function of the air flow pattern, although the distribution of the injected air is still poorly understood. Cross-borehole resistivity surveys were performed at a former service station in Florence, Oregon, to address this unknown. The resistivity measurements were made using six wells, one of which was the sparge well. Data were collected over a two-week period during and after several air injections, or sparge events. Resistivity images were calculated between wells using an algorithm that assumes axially symmetric structures. The movement of the injected air through time was defined by regions of large increases in resistivity, greater than 100 percent from the background. During early sparge times, air moved outward and upward from the injection point as it ascended to the unsaturated zone. At later sparge times, the air flow reached a somewhat stable cone-shaped pattern radiating out and up from the injection point. Two days after sparging was discontinued, a residue of entrained air remained in the saturated zone, as indicated by a zone of 60 to 80 percent water saturation.     

Laboratory Study of Air Sparging: Air Flow Visualization

Laboratory flow visualization experiments, using glass beads as the porous medium, were conducted to study air sparging, an innovative technology for subsurface contaminant remediation. The purpose of these experiments was to observe how air flows through saturated porous media and to obtain a basic understanding of air plume formation and medium heterogeneity effects. The experiments indicate that air flow occurring in discrete, stable channels is the most probable flow behavior in medium to fine grained water saturated porous media and that medium heterogeneity plays an important role in the development of air channels. Several simulated scales of heterogeneities, from pore to field, have been studied. The results suggest that air channel formation is sensitive to the various scales of heterogeneities. Site-specific hydrogeologic settings have to be carefully reviewed before air sparging is applied to remediate sites contaminated by volatile organic compounds.     

An Overview of In Situ Air Sparging

In situ air sparging (IAS) is becoming a widely used technology for remediating sites contaminated by volatile organic materials such as petroleum hydrocarbons. Published data indicate that the injection of air into subsurface water saturated areas coupled with soil vapor extraction (SVE) can increase removal rates in comparison to SVE alone for cases where hydrocarbons are distributed within the water saturated zone. However, the technology is still in its infancy and has not been subject to adequate research, nor have adequate monitoring methods been employed or even developed. Consequently, most IAS applications are designed, operated, and monitored based upon the experience of the individual practitioner.
The use of in situ air sparging poses risks not generally associated with most practiced remedial technologies: air injection can enhance the undesirable off-site migration of vapors and ground water contamination plumes. Migration of previously immobile liquid hydrocarbons can also be induced. Thus, there is an added incentive to fully understand this technology prior to application.
This overview of the current state of the practice of air sparging is a review of available published literature, consultation with practitioners, a range of unpublished data reports, as well as theoretical considerations. Potential strengths and weaknesses of the technology are discussed and recommendations for future investigations are given.     

Three-Dimensional Experimental Testing of a Two-Phase Flow-Modeling Approach for Air Sparging

Air sparging has been used for several years as an in situ technique for removing volatile compounds from contaminated ground water, but few studies have been completed to quantify the extent of remediation. To gain knowledge of the air flow and water behavior around air injection wells, laboratory tests and model simulations were completed at three injection flow rates (62, 187, and 283 lpm) in a cylindrical reactor (diameter - 1.2 m, depth = 0.65 m). Measurements of the air flux distribution were made across the surface of the reactor at 24 monitoring locations, six radial positions equally spaced along two orthogonal transects. Simulations using a multiphase flow model called T2VOC were completed for a homogeneous, axisymmetric configuration. Input parameters were independently measured soil properties. In all the experiments, about 75 percent of the flow injected exited the water table within 30 cm of the sparge well. Predictions with T2VOC showed the same. The averages of four flux measurements at a particular distance from the sparge well compare satisfactorily with T2VOC predictions. Measured flux values at a given radius varied by more than a factor of two, but the averages were consistent between experiments and agreed well with T2VOC simulations. The T2VOC prediction of the radial extent of sparging coincided with the distance out to which air flow from the sparge well could not be detected in the reactor. The sparging pattern was relatively unaffected by the air injection rate over the range of conditions studied. Changes in the injection rate resulted in nearly proportional changes in flux rates.     

A Field and Laboratory Investigation of Air Fingering During Air Sparging

Since the rejection of the bubble flow conceptual model for in situ air sparging, most practitioners have adopted the conceptual model of air channeling, which generally implies the development of widely spaced, discreet air channels that bypass large regions of the subsurface. While air channeling clearly develops in response to stratigraphic heterogeneity, the universality of widely spaced air channels in homogeneous media is not supported by available evidence. Air channeling results in low bulk air saturation due to bypassing, and field and laboratory measurements of air saturations and previously published studies were used to evaluate if air channeling is realistic. The results indicated that homogeneous coarse sands are prone to the development of air channeling, and that homogeneous fine sands show higher air saturations and are not prone to air channeling. Breakthrough air saturations, which represent the minimum air saturations, that will conduct air flow, of approximately 0.02 to 0.04 were observed in coarse sands. In contrast, breakthrough air saturations of 0.10 to 0.13 were observed in fine sands and medium sands. The transition between these behaviors falls at about 15 to 20 cm water air entry pressure. These result indicate that, at both the field and laboratory scale, coarse sands are more prone to air channeling and bypassing than fine sands. Additionally, the larger air gradients and capillary pressures in fine sands result in a less buoyancy-dominated flow pattern, with a larger lateral extent of air flow.     

Fundamental Changes in In Situ Air Sparging How Patterns

Two types of gas-phase flow patterns have been discussed and observed in the in situ air sparging (ISAS) literature: bubble flow and air channels. A critical factor affecting the flow pattern at a given location is the grain size of the porous medium. Visualization experiments reported in the literature indicate that a change in the flow pattern occurs around 1 to 2 mm grain diameters, with air channels occurring below the transition size and bubbles above. Analysis of capillary and buoyancy forces suggests that for a given gas-liquid-solid system, there is a critical size that dictates the dominant force, and the dominant force will in turn dictate the flow pattern. The dominant forces, and consequently the two-phase flow patterns, were characterized using a Bond number modified with the porous media aspect ratio (pore throat to pore body ratio). Laboratory experiments were conducted to observe flow patterns as a function of porous media size and air flow rate. The experimental results and the modified Bond number analysis support the relationship of flow patterns to grain size reported in the literature.     

Laboratory and Numerical Investigations of Air Sparging Using MTBE as a Tracer

Air sparging experiments were conducted in a laboratory column to investigate air flow and mass transfer behavior in different types of sand at different air injection rates. Methyl tertiary butyl ether (MTBE) was applied as a tracer, and by measuring the volatilization and the mean air content during the experiments, the air flow pattern and its influence on mass transfer were assessed. The experimental results showed large differences among the sand types. In fine sand, the mean air content was high and the volatilization of MTBE was rapid with total recovery after a few hours. In coarse sand, the mean air content was low and the volatilization of MTBE was limited. The results indicate two different air flow distributions. In fine-grained materials, a uniform air distribution can be expected compared to coarse-grained materials where isolated air channels will limit the mass transfer. Afterwards, the experiments were simulated using the numerical multiphase flow code T2VOC, and the results compared to those obtained in the laboratory. The experiments with fine sand were simulated well, while for coarser sand types the volatilization was highly overestimated. The differences between model and laboratory results were mainly attributed to the nonuniformity of the air saturation and the neglection of kinetics in the mass transfer formulation.     

Analytical Model for Contaminant Mass Removal by Air Sparging

An analytical model was developed lo predict the removal of volatile organic compounds (VOCs) from ground water by air sparging (AS). The model treats the air sparging zone as a completely mixed reactor subject to the removal of dissolved contaminants by volatilization, advection, and first-order decay Nonequilibrium desorption is approximated as a first-order mass transfer process. The model reproduces the tailing and rebound behavior often observed at AS sites, and would normally require the estimation of three site-specific parameters. Dimensional analysis demonstrates that predicted tailing can be interpreted in terms of kinetic desorption or diffusion of aqueous phase contaminants into discrete air channels. Related work is ongoing to test the model against field data.     

Air Channel Formation, Size, Spacing, and Tortuosity During Air Sparging

Characterizing mass transfer during in situ air sparging requires knowledge of the size, shape, and interfacial area of air channels. These characteristics were determined by analysis of digital images of air channels passing through submerged glass beads having particle size in the sand range. Pore-scale channeling occurred in all cases. The analysis showed that the air channels were narrower, more tortuous, more closely spaced, and moved nearly vertically through the coarser media. In the finer media, air channels had larger diameter, were spaced further apart, and passed nearly horizontally through the media. The mean diameter of the channels varied between 2.8 and 8.1 mm, and the mean spacing varied between 8.3 and 19.4 mm. Estimates of the area of the air-water interface per unit volume of soil (a₀), computed using data from the digital images and an assumed arrangement of channels, ranged from 0.02 to 0.2 mm²/mm³. Larger a₀ were obtained for coarser media and uniformly graded media. These estimates of a₀ compare well with published values for common packed-column materials and for unsaturated soils.     

Effect of Temporal Changes in Air Injection Rate on Air Sparging Performance Groundwater Remediation

Air sparging (AS) is a commonly applied method for treating groundwater contaminated with volatile organic compounds (VOCs). When using a constant injection of air (continuous mode), a decline in remediation efficiency is often observed, resulting from insufficient mixing of contaminants at the pore scale. It is well known that turning the injection on and off (pulsed mode) may lead to a better remediation performance. In this article, we investigate groundwater mixing and contaminant removal efficiency in different injection modes (i.e., continuous and pulsed), and compare them to those achieved in a third mode, which we denote as “rate changing.” In this mode, injection is always on, and its rate is varying with time by abrupt changes. For the purpose of this investigation, we conducted two separate sets of experiments in a laboratory tank. In the first set of experiments, we used dye plume tracing to characterize the mixing induced by AS. In the second set of experiments, we contaminated the tank with a VOC and compared the remediation efficiency between the different injection modes. As expected, we observed that time‐variable injection modes led to enhanced mixing and contaminant removal. The decrease in contaminant concentrations during the experiment was found to be double for the “rate changing” and “pulsed” modes compared to the continuous mode, with a slightly preferable performance for the “rate changing” mode. These results highlight the critical role that mixing plays in AS, and support the need for further investigation of the proposed “rate changing” injection mode.     

Pulsed Air Sparging in Aquifers Contaminated with Dense Nonaqueous Phase Liquids

Air sparging was evaluated for remediation of tetrachloroethylene (PCE) present as dense nonaqueous phase liquid (DNAPL) in aquifers. A two-dimensional laboratory tank with a transparent front wall allowed for visual observation of DNAPL mobilization. A DNAPL zone 50 cm high was created, with a PCE pool accumulating on an aquitard. Detailed process control and analysis yielded accurate mass balances and insight into the mass-transfer limitations during air sparging. Initial PCE recovery rates were high, corresponding to fast removal of residual DNAPL within the zone influenced directly by air channels. The vadose zone DNAPL was removed within a few days, and the recovery in the extracted soil vapors decreased to low values. Increasing the sparge rate and pulsing the air injection led to improved mass recovery, as the pulsing induced water circulation and increased the DNAPL dissolution rate. Dissolved PCE concentrations both within and outside the zone of air channels were affected by the pulsing. Inside the sparge zone, aqueous concentrations decreased rapidly, matching the declining effluent PCE flux. Outside the sparge zone, PCE concentrations increased because highly contaminated water was pushed away from the air injection point. This overall circulation of water may lead to limited spreading of the contaminant, but accelerated the time-weighted average mass removal by 40% to 600%, depending on the aggressiveness of the pulsing. For field applications, pulsing with a daily or diurnal cycling time may increase the average mass removal rate, thus reducing the treatment time and saving in the order of 40% to 80% of the energy cost used to run the blowers. However, air sparging will always fail to remove DNAPL pools located below the sparge point because the air will rise upward from the top of a screen, unless very localized geological layers force the air to migrate horizontally. Unrecognized presence of DNAPL at chlorinated solvent sites residual and pools could potentially hamper success of air sparging cleanups, since the presence of small DNAPL pools, ganglia or droplets can greatly extend the treatment time.     

Enhancing Groundwater Remediation in Air Sparging by Changing the Pulse Duration

An important operation parameter in the design of a pulsed air sparging (PAS) system is the pulse duration (PD). To study the effect of the PD on the remediation process, a series of laboratory experiments and numerical simulations were performed. The experimental apparatus was a cylindrical tank, packed with fine sand and partially filled by water contaminated with toluene. Toluene concentrations in water and in effluent air were measured over time during the application of PAS, which was applied with three different PD. Next, the T2VOC model, an extension of the TOUGH2 simulation program, was used to simulate the two-phase flow and transport processes for these cases. The simulation model was calibrated to the experimental results, and then run with a range of PD values. Results showed that there exists an optimal PD which yields the highest remediation efficiency. Next, it was shown that this PD may be obtained by performing a PAS pilot test and measuring the groundwater pressure response in a monitoring well. The characteristic time which describes the exponential decay of the pressure response was shown to provide an adequate estimate for the optimal PD. The estimation improved by taking a number of injection cycles.