VZJ
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 27 May 2008
Published in Vadose Zone J 7:732-740 (2008)
DOI: 10.2136/vzj2007.0093
© 2008 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hodges, R. A.
Right arrow Articles by Falta, R. W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Hodges, R. A.
Right arrow Articles by Falta, R. W.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hodges, R. A.
Right arrow Articles by Falta, R. W.
Related Collections
Right arrow Chlorinated Hydrocarbons
Right arrow Volatile Organic Compounds
Right arrow Heat Transport

SPECIAL SECTION: VADOSE ZONE MODELING

Vertical Confinement of Injected Steam in the Vadose Zone Using Cold Air Injection

Rex A. Hodges and Ronald W. Falta*

Dep. of Environmental Engineering and Earth Sciences, Clemson Univ., Clemson, SC 29634-0919. R.A. Hodges, current address, S.M. Stoller, 2597 B3/4 Rd., Grand Junction, CO 81503
* Corresponding author (faltar{at}clemson.edu).

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


Received 15 May 2007.



   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Steam Flood Modeling
 Results
 Discussion and Conclusions
 REFERENCES
 
Steam flooding is used for removal of volatile organic contaminants (VOCs) from concentrated source zones. The application of steam flooding to shallow vadose zone sites can be problematic if the steam or contaminants vent to the ground surface. Cold air injection holds promise as a technique for controlling the vertical movement of the injected steam. A series of two-dimensional numerical simulations using a radially symmetric grid show the sensitivity of this process to geological and operational parameters. Air control of vertical steam movement appears to be a useful technique, using approximately equal air and steam volumetric flow rates. Air injection is shown to be more effective than a low-permeability cap at the ground surface for keeping high temperatures away from the surface. The most effective design appears to be a combination of a low-permeability surface cap and air injection above the steam injector. The cap mainly serves to direct the air flow, preventing it from exiting at the ground surface near the air injection point.

Abbreviations: NAPL, nonaqueous phase liquid • VOC, volatile organic contaminant


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Steam Flood Modeling
 Results
 Discussion and Conclusions
 REFERENCES
 
STEAM FLOODING is an aggressive remediation technology that can be used to remove a large percentage of a volatile contaminant from a high concentration source zone. It is typically used on sites with a persistent source of VOC or nonaqueous phase liquids (NAPLs), such as hydrocarbon fuels or chlorinated solvents. The benefits of an aggressive remediation effort include accelerated land usability and value in the vicinity of the source zone, decreased downgradient concentrations, and reduced long-term management costs.

Sites in the vadose zone or saturated zone with a well-defined, localized contaminant source in a stratified lithologic environment are the most conducive to steam flooding. Steam is injected into wells positioned in a pattern designed to surround a contaminated zone and drive contaminants to a centralized collection system in a coordinated effort. Steam is injected into the more permeable layers to maximize flow and velocity for a given pressure and to allow the steam front to expand rapidly. Interbedded low-permeability layers limit the tendency of the steam to rise vertically and enhance the desired horizontal spread. Buoyancy effects of steam are most pronounced under saturated conditions.

The primary means by which steam improves the removal of NAPL from a source zone is through an increased rate of contaminant evaporation due to an increase in vapor pressure and Henry's constant with increasing temperature. The rate of evaporation can increase by more than a factor of 50 (Hunt et al., 1988; Basel, 1991) over the temperature range of interest, 10 to 100°C. Other contributing factors are the high steam velocity (Falta et al., 1992b), the development of a high-saturation NAPL bank at the steam condensation front, higher displacing pressures due to the velocity of the steam front, and a reduction in NAPL viscosity and surface tension at elevated temperatures. During a steam flood, hundreds of pore volumes of steam are injected (due to the large volume reduction as steam condenses back to water), creating a stripping effect that makes it unlikely that much dissolved chemical will remain behind the steam front (Udell, 1994).

A zone of steam spreads outward from each injection point until it reaches a distance at which it is cooled by the surrounding formation, forming a condensation steam front. The steam zone volume grows with continued injection heating more of the formation, pushing the steam front farther from the injection point. The advancement of the steam front is inherently stable as steam fingering into small high-permeability layers is stabilized by large, lateral heat conduction losses. Only large-scale heterogeneities, such as high-permeability channel sands, are expected to significantly influence steam zone growth (Udell, 1994; Lake, 1989).

Several field-scale steam flood operations have been performed with promising results (Udell and Stewart, 1989; Newmark, 1994; Gildea et al., 1997; Stewart et al., 1998; Newmark et al., 2000; Heron, 2000; Integrated Water Resources and IT Corp., 2000; Heron et al., 2005). Most of these efforts have been performed in strongly stratified geological settings, with steam injection points located some distance below the ground surface.

The potential exists for undesirable horizontal and vertical migration of both steam and contaminants during steam flooding. The outward spread of the steam front could encroach on areas that would be adversely affected by heating, such as below-ground building structures or utilities. Care must be taken to design an injection pattern that completely surrounds the source since the outward spread of the steam front can mobilize contaminants external to the pattern and concentrate them away from the collection system. Collection wells located outside the pattern will have limited effectiveness as they will be trying to capture contaminants from a spreading (diverging) front rather than a front that is converging toward a central location. When the contaminant is a dense NAPL, adverse vertical migration is of concern if no low-permeability units exist below the target zone to inhibit downward movement. However, special flood designs, can lower the risk of adverse vertical migration (Farber, 1997; Emmert, 1997; Helmig et al., 1998; Kaslusky and Udell, 2005).

In the absence of strong horizontal layering, steam tends to travel toward the ground surface due to the large pressure gradients typical of steam injection operations and the negative impact of buoyancy. This upward vertical movement can be problematic if the steam or contaminants vent through the ground surface.

The problem of effectively controlling steam and volatized contaminants has limited the use of steam flooding for remediation of shallow, nonstratified sites. However, it has been shown through tank experiments, a field experiment, and numerical modeling that cold air injection wells placed opposite steam injection wells can control the outward horizontal spread of steam. This idea was originally developed and applied by L. Stewart of Praxis Environmental Technologies, Inc., in the mid-1990s and was further developed in Hodges et al. (2004) for the case of horizontal confinement of the steam. Cold air injection controls steam front movement primarily through the creation of gas pressure reflection boundary and a counter-convective mechanism to prevent a steam zone from growing by conduction toward the air injection wells. Conceptually, this is similar to groundwater flow from an injection well with an identical image well located some distance away. The presence of the image well results in a linear no-flow boundary between the wells. With cold air injection, a similar effect is seen when air is injected at the same volumetric (not mass) rate as the steam. The interaction between the steam zone and air injection is somewhat more complicated, however, because thermodynamic and heat transfer effects such as evaporative cooling near the air–steam boundary and thermal conduction effects.

The goals of the current study are to show that it is feasible to control the vertical rise of steam by injecting cold air above the steam injection points (Fig. 1 ) and to show the effect of factors that vary from site to site (e.g., permeability, boundaries, and injection rates). The vertical air injection method will allow sites to be evaluated to determine what effort is required to keep injected steam and contaminants from reaching the ground surface. This could make steam flooding a viable technology for shallow sites previously considered to be poor candidates.


Figure 1
View larger version (28K):
[in this window]
[in a new window]

  		FIG. 1. Schematic cross-section showing the growing zones of injected air (dotted lines) and steam (solid lines) to prevent upward steam migration.

 

   Steam Flood Modeling
TOP
 ABSTRACT
 INTRODUCTION
 Steam Flood Modeling
Results
 Discussion and Conclusions
 REFERENCES
 
This study uses the T2VOC module of the TOUGH2 multiphase flow numerical simulator developed at Lawrence Berkeley National Laboratory (Pruess et al., 1999). TOUGH2 is a modular integral finite difference method code that solves the coupled, nonlinear heat and mass balance equations using a residual-based Newton Raphson iteration method. The TOUGH2 code includes 11 or more different equation-of-state modules designed for simulating different mass components. The T2VOC module (Falta et al., 1995, 1992a) models three-phase (gas, liquid water, NAPL) flow with three mass components (air, water, chemical) and heat. The present study mainly considered air and steam movement, so the NAPL phase and chemical transport capability of T2VOC were largely not needed, and the simulations mostly could have been conducted using one of the two-phase TOUGH2 equation of state modules that consider air and water mass components (EOS3, EOS4, EOS7, EOS7R, or EWASG). It is also worth noting that a related code, TMVOC (Pruess and Battistelli, 2002), is capable of simulating three-phase flow with multicomponent NAPLs and multiple noncondensible gases. The multicomponent NAPL capability in TMVOC would be needed for simulating thermal processes where distillation of the NAPL was occurring due to different NAPL component vapor pressures.

All of the TOUGH codes simulate compositional multiphase flow in three-dimensional, anisotropic, heterogeneous porous or fractured media while allowing for phase appearance or disappearance and equilibrium mass transfer between phases. Several simplifying assumptions are made in the T2VOC formulation. Air is treated as a single pseudo-component with average properties, the usually small solubility of water in the NAPL phase has been neglected, and no allowance is made for hydrodynamic dispersion. The Henry's constants for air dissolution in aqueous and NAPL phases are small and for simplicity have been assumed to be constant. It is assumed that the three phases are in local chemical equilibrium and that no chemical reactions are taking place other than (i) interphase mass transfer, (ii) linear equilibrium adsorption of the chemical compound to the solid phase, and (iii) decay of VOC by first-order biodegradation (Falta et al., 1995).

Fluid phases in the TOUGH codes flow in response to pressure and gravitational forces according to the multiphase extension of Darcy's law, including the effects of relative permeability and capillary pressure. Transport of the mass components also occurs by multicomponent diffusion in each phase. Mechanisms of interphase mass transfer for the components include evaporation and boiling of liquids, dissolution of the NAPL components and noncondensable gases into the aqueous phase (T2VOC and TMVOC), condensation of the components from the gas phase into the aqueous or NAPL phases, and equilibrium phase partitioning between the gas, aqueous, NAPL, and solid phases.

Heat transfer results from conduction and multiphase convection. The heat transfer effects of phase transitions between NAPL, aqueous and gas phases are fully taken into account by considering the transport of both latent and sensible heat. The overall porous media thermal conductivity is calculated as a function of water saturation.

Steam Migration Simulations
Idealized models were designed to simulate the air control of injected steam to examine the effects of porous medium, boundary, and operational factors that vary at individual sites and to determine its feasibility for field applications. The model parameters for the base case simulations are shown in Table 1 . Several of these parameters were varied in the simulations to determine how porous medium properties influence the control of steam migration with injected air. Porous medium factors varied include permeability, anisotropy (ratio of horizontal and vertical permeability), water saturation (by altering the capillary pressure function), and the presence of less-permeable layers. Boundary factors include the upper model boundary conditions. Operational factors include the steam injection rate and the ratio of injected air to injected steam. In each case, steam was injected until approximate steady-state conditions were achieved. The cases simulated are shown in Table 2 .


View this table:
[in this window]
[in a new window]

  		TABLE 1. Base case porous media and thermal properties.{dagger}

 

View this table:
[in this window]
[in a new window]

  		TABLE 2. Simulation cases.

 
The cases were simulated using a two-dimensional cylindrical (r,z) configuration with 1995 elements. Two advantages of this type of numerical grid are that it is perfectly suited to simulate radially symmetric flow to or from nested wells and that it uses far fewer elements than a true three-dimensional model, greatly reducing computation time. However, a fully three-dimensional model is needed to simulate the typical steam flood with multiple injection wells surrounding a contaminated source area. A number of field-scale three-dimensional simulations of cold air injection for vertical steam control were given by Hodges (2006); they are not shown here.

Model Domain and Initial Conditions
The numerical mesh consists of 57 radial elements extending a total of 99.5 m from the centerline. The width of the elements increases outwardly, from 0.5 m for the two inner cylinders, to 1 m out to 20 m, to 2 m out to 70 m, and 3 m to 99.5 m. The lateral dimensions are large enough to minimize the effects of boundary conditions at the model edge that are set at constant gravity-static pressure, saturation (based on gravity–capillary equilibrium), and uniform and constant temperature.

The vertical mesh extends from the ground surface (set at an elevation of 0.0) to a depth of 51.0 m. The thickness of the elements increase downward, with 1 m thickness from 0 to 20 m, 2 m from 20 to 39.5 m, and 3 m to 51.0 m. The bottom of the grid, a no-flow boundary, is sufficiently deep to minimize lower boundary effects. Constant atmospheric conditions (pressure, temperature, and relative humidity) are simulated at the ground surface, which forms the top boundary.

The initial gas and water saturation conditions are calculated assuming that the water phase pressure is hydrostatic both above and below the water table. Above the water table, the water saturation depends on the water pressure (compared to the static gas pressure) through the capillary pressure curve. A typical initial gas saturation distribution (water table at –40.5 m) and the numerical mesh are shown in Fig. 2 .


Figure 2
View larger version (34K):
[in this window]
[in a new window]

  		FIG. 2. Two-dimensional axi-symmetric mesh with initial gas saturation distribution (water table at 40.5 m).

 
Model Parameters and Source Terms
The base case model simulations use a uniform horizontal permeability of 10 darcys, (typical of sands that would be good steam flood candidates), no anisotropy (horizontal and vertical permeability ratio = 1), and a porosity of 30%. Base case capillary pressure and relative permeability functions were estimated based on those that represent typical medium grain sands (Table 1). The steam flood results were found to be relatively insensitive to the small changes in capillary pressure and relative permeability parameters (Hodges, 2006).

The steam injection depth (10 m), injection rate (0.010 m3 s–1 = 0.006 kg s–1 = 50 lbs h–1), steam specific enthalpy (2440,000 J kg–1), and steam quality (0.90) were the same for all simulations. The base steam injection rate corresponds to an energy input of about 15 kW. Air was injected at a depth of 5 m and at the same volumetric rate (0.010 m3 s–1) as steam injection for the base case. Base case operational parameters are shown in Table 3 . The simulations were run until steam propagation reached near steady state conditions which typically occurred within 20 yr (Fig. 3 ). For perspective, environmental steam floods are operated for time scales of days to months.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Operational parameters, base case simulations.

 

Figure 3
View larger version (28K):
[in this window]
[in a new window]

  		FIG. 3. Time sequence comparison of steam injection with (blue) and without (black) air control. Isotherm (70°C) times at 10, 30, 60 (thick line), 100 d and 1, 2, 5, 10, 20 (thick line), 50, 100, and 200 yr radiating from the steam injection point at 10 m (air injection at 5 m). Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 

   Results
TOP
 ABSTRACT
 INTRODUCTION
 Steam Flood Modeling
 Results
Discussion and Conclusions
 REFERENCES
 
The simulation results for the base case show that injecting air is a viable method for controlling the vertical rise of high temperatures associated with steam injection (Fig. 3). Heat expands outward and upward toward the ground surface from the steam injection point unless a controlling mechanism is used. Injected cold air effectively controls the high temperature zone, although it is less effective at controlling the location of the larger warm zone away from the injection point.

A series of contaminant transport simulations were run using the base case parameters (Fig. 4 ). In these simulations, a cylindrical contaminated zone containing tetrachloroethylene at a soil mass fraction of 100 mg kg–1 was emplaced from a depth of 6 to 9 m, starting at a radius of 5 m. This is a highly idealized configuration, but it illustrates the dramatic effect that air control can have on contaminant movement. In this case, injecting steam alone causes almost no displacement of the contaminant after 2 wk (Fig. 4, center plot) because the contaminant is located outside of the steam radius of influence. However, by adding an equal amount of air injection (Fig. 4, right-hand plot), the contaminant is rapidly driven outward from the injection location. It is also evident that some of this contaminant would vent from the ground surface in this case because there was no vapor extraction. In a real field situation, vapor extraction wells would be a critical part of the operating system. Simulations of these systems requires a three-dimensional geometry, however (see, e.g., Hodges, 2006).


Figure 4
View larger version (10K):
[in this window]
[in a new window]

  		FIG. 4. Contaminant transport with and without air control. Left hand plot shows the initial distribution of tetrachloroethylene contamination (volatile organic contaminant gas mass concentration in g L–1 chlorinated volatile organic contaminant [CVOC] gas). The center plot shows the contaminant distribution after 14 d of steam injection with no air control, while the right hand plot shows the contaminant distribution after 14 d of steam injection with air control. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
The basic element of a vertical steam control system is a nested pair of air and steam injection wells. Because the present study focuses mainly on design and parameter sensitivity of a single nested pair of air and steam wells, it is much more convenient to use the two-dimensional cylindrical geometry. For this reason, the remaining simulations focus on the control of the high temperature zone created during steam injection, rather than on optimizing the contaminant recovery, which requires a more complex simulation geometry.

The results of the simulations performed to investigate the influence of porous medium, boundary, and operational factors on the temperature profile for a nested pair of air and steam injection wells follow. These factors are expected to influence steam flood design, implementation, and feasibility.

Porous Medium Influences on Air Control of Steam Floods
Porous medium factors that could affect migration of steam in the subsurface include porosity, permeability, anisotropy, capillary pressure, and the presence of low-permeability layers. Porosity was found to have the least effect, with only small differences for cases ranging in values from 0.05 to 0.40. Varying capillary pressure had a larger impact than porosity, though much less than permeability, anisotropy, and the presence of a low-permeability layer.

Permeability
Candidate zones for steam flooding must be permeable enough to allow steam injection at a rate sufficient to heat the targeted area within a reasonable period of time and to allow the steam front to spread far enough to drive contaminants to extraction wells (Ochs et al., 2003). To test the effect of varying permeability, the medium was assumed to be homogeneous with no anisotropy; all other factors were held constant. High-permeability media allow the steam front to spread farther and treat a larger volume for a given injection rate because of the lower injection pressure and temperature. Vertical steam rise is greater in high-permeability material and is also more difficult to control.

Temperature distributions from simulations with permeability values of 0.1, 1, 10, 30, and 100 darcys were compared. The results show the greatest effect for the higher-permeability values. In Fig. 5 , there is almost no difference in the temperature distribution (left, 80°C; right, 40°C) for 0.1 and 1 darcy simulations. The effect of permeability becomes significant at 10 darcys and above, with the 40°C isotherm nearly reaching the surface for the 100 darcy simulation at "steady state." Another pronounced effect is that for high-permeability material, the downward spread of steam is limited by its buoyant rise. This could require an additional injection point in high-permeability material if the contamination is more vertically distributed. The results indicate that injected air can prevent the high temperatures associated with a shallow steam flood from reaching the surface at permeability values up to 100 darcys.


Figure 5
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 5. Temperature distributions (left, 80°C; right, 40°C) comparing the effects of permeability on the air control of steam propagation. Simulations are at "steady state" for permeability values 0.1, 1, 10, 30, and 100 darcys. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
Anisotropy
Anisotropy is typical of many systems with horizontal permeability (kh) being greater than vertical permeability (kv). Stratified systems with sufficient horizontal permeability are excellent steam flood candidates since the rise of steam is inhibited by the relatively lower kv. Cases were simulated with a constant kh of 10 darcys for kh/kv ratios of 2, 5, 10, and 100. The simulations with no air control (Fig. 6 ) show the effect of increasing the kh/kv ratio, progressing from a spherical shaped hot zone (kh/kv of 1) to a compressed ellipsoidal shaped hot zone for a kh/kv of 10. Even within highly stratified systems, high temperatures can approach the surface at large times due to conductive heating, but this is not expected to be a problem for floods of normal duration. The chart on the right of Fig. 6 shows the 40°C distribution after 100 d of injection. The ability of injected air to control upward heat migration is enhanced with increasing anisotropy (Fig. 7 ). There is almost no rise in high temperatures and the moderate temperatures are kept well away from the surface.


Figure 6
View larger version (12K):
[in this window]
[in a new window]

  		FIG. 6. Temperature distributions (left, 80°C; center, 40°C) comparing the effects of anisotropy on steam propagation. Anisotropy ratios at 1, 2, 5, 10, and 100x are shown. The left and center graphs are at "steady state"; the right graph shows the 40°C distribution after 100 d of injection. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 

Figure 7
View larger version (12K):
[in this window]
[in a new window]

  		FIG. 7. Temperature distributions (left, 80°C; right, 40°C) at "steady state" for steam injection (10 m) into anisotropic media (2, 5, 10, 100x) with air control (5 m). The isotropic case is shown in gray. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
Water Saturation
The effect of water saturation (achieved by varying the capillary pressure function) on the temperature distribution only becomes significant when the difference in water saturation is considerable. Two variations from the base case were examined: one that more readily drains to a lower saturation and one that retains more water. The capillary pressure function parameters are given in Table 1, and plots of the functions are shown in Fig. 8 . Results (Fig. 9 ) show that injected air is effective in limiting vertical steam rise for these cases, but the high saturation conditions can significantly limit the lateral spread of the steam front (as indicated by the heating pattern).


Figure 8
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 8. Capillary pressure (Pc) curves for the high and low water saturation (Sw) cases.

 

Figure 9
View larger version (10K):
[in this window]
[in a new window]

  		FIG. 9. Temperature distributions (left, no air control; right, air injected at 5 m) for steam injection (10 m) into media with varying water saturation due to capillary pressure differences. The solid black lines indicate the high water saturation case. The dashed black lines indicate the low water saturation case. Contours are shown for 7, 30, and 90 d and "steady state." The base case "steady state" contour is shown in gray. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
Low-Permeability Layer
The presence of a low-permeability layer above steam injection inhibits steam rise. Cases for a 0.1 darcy layer (1 m thick at 5 m and 3 m thick from 5 to 8 m) were simulated and compared to the case of air injection at 5 m. Results (Fig. 10 ) show that although the low-permeability (with relatively high capillary pressure) layer keeps high temperatures from reaching the surface, it is less effective than injected air.


Figure 10
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 10. Low-permeability layer simulated temperature distribution (left, 80°C; right, 40°C) at "steady state" for steam injection (10 m). The isotropic case (steam only and with air injection) is shown in gray. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
Top Boundary: Cap
Low-permeability caps have been used to keep steam from venting to the surface. An installed layer of clay or other very low permeability material provides a barrier for safety and helps keep energy in the subsurface. A cap or partial cap, such as asphalt in a parking lot or the cement pad of a loading dock, may already be present at some sites. Caps can limit steam from venting to the surface, but they are still subject to conductive heating and can get uncomfortably hot. Depending on the design of the flood and the composition of the cap, it could also be physically altered from prolonged heating.

Simulations were performed with a very low permeability (0.0001 darcys), low-porosity (0.02) 1-m-thick cap that extends to a radius of 20 m from the injection point at the surface. Results show that a cap does not significantly impair heat (steam) rise until steam (heat) reaches the cap. It prevents the high temperatures (>80°C) from reaching the surface, although they are near (Fig. 11 ). Lower temperatures (40°C) do reach the surface but only at late times. A comparison case shows that injected air prevents even the low temperature (40°C) isotherm from nearing the surface.


Figure 11
View larger version (12K):
[in this window]
[in a new window]

  		FIG. 11. Temperature distribution ("steady state") with low-permeability cap to control steam rise to the surface (black line). No cap and air controlled with no cap shown in gray. Cap shown. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
A more demanding problem with an increased steam injection rate (10 times higher) was simulated. The cap keeps the 80°C temperature just below the surface (it would vent to the surface with no cap), although not as effectively as air injection. The 40°C temperature isotherm reaches the surface for both of these cases (Fig. 12 ).


Figure 12
View larger version (16K):
[in this window]
[in a new window]

  		FIG. 12. Comparison of low-permeability cap (solid line) to injected air (dash) for controlling steam rise (10x injection rate) at "steady state." Cap not shown. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
A hybrid design that combines both a cap and injected air was simulated, and it resulted in improved temperature rise control for the high steam injection rate. Cases were simulated with air injected at the same ("full") and half the steam volumetric rate with a cap present. This design significantly lowers the 80°C temperature isotherm and keeps the 40°C isotherm below the surface for the 10x steam injection rate (Fig. 13 ). Vertical control is significantly improved when the air injection rate is only half the steam rate with the 40°C isotherm only a meter above the case where the air injection rate is the same as the steam rate. These results suggest that a cap in combination with injected air not only improves vertical control but could be used to better guide steam front propagation and prevent cap alteration due to heating.


Figure 13
View larger version (21K):
[in this window]
[in a new window]

  		FIG. 13. Comparison (at "steady state") of a cap-only design to a hybrid design that uses the low-permeability cap (solid) in combination with injected air (dash) for controlling steam rise (10x injection rate). Cap only (solid line), cap plus air injected at the steam volumetric rate (dash), and cap plus air injected at half the steam rate (dot dash). Cap not shown. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
Operational Influences on Air Control of Steam Floods
Operational factors (steam injection rate, ratio of air and steam rates) were varied to test their effect on the temperature distribution of a vertically air controlled steam flood. The operational factors (Table 3) were constant for the simulations that tested porous medium and boundary effects, except for the previous case that used high injection rates to test a top-boundary low-permeability cap. Simulated cases are listed in Table 2.

Steam Injection Rate
The clear benefit to higher injection rates is that a given target zone can be heated in a shorter time. Simulated temperature distributions of steam and air injected at 0.01 (base case), 0.02, 0.04, and 0.10 m3 s–1 were compared to assess the effect of increasing injection rates on the ability of injected air (at the same volumetric rate as the steam) to limit the rise of high temperatures associated with a steam flood. Injected air keeps the 80°C isotherm from reaching the surface at all rates (Fig. 14 ). Results at both 100 d and at "steady state" are shown on Fig. 14. However, the 40°C isotherm reaches the surface for the highest injection rate (10x the base rate) and nears the surface for other high rates. This suggests that at high rates, additional air injectors or a hybrid design that includes some sort of cap may be needed.


Figure 14
View larger version (20K):
[in this window]
[in a new window]

  		FIG. 14. Temperature distribution comparison for simulations with increasing injection rates at 100 d (fine line) and "steady state" (SS) (thick line). Injection rates (m3 s–1) are 0.01 (base case, solid gray), 0.02 (dash), 0.04 (dot dash), and 0.10 (long dash). Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
Ratio of Air and Steam Injection Rates
The ability to control steam rise with less air injection would be more economical, with smaller compressors and lower energy costs. Additionally, the ratio of air injected could be decreased to allow steam to rise higher and affect a more vertically expansive target zone while maintaining control of the rise. Higher air/steam injection ratios could be used to ensure that steam (heat) does not reach the surface. Simulated temperature distributions of different air/steam injection ratios were compared to determine the effect of varying air/steam injection ratios.

Cases were run with air injected at half and double the base volumetric rate of steam and at double the 10x steam injection rate (20x base air rate). The temperatures rise only about a meter more than the base configuration when air is injected at half the steam rate (Fig. 15 ). This suggests that a lower air injection ratio, at least initially, would adequately control steam rise for many sites. If subsurface temperatures increased to undesired levels nearing the surface, the air rate could be increased as necessary.


Figure 15
View larger version (13K):
[in this window]
[in a new window]

  		FIG. 15. Temperature distribution comparison for simulations with air injection at half the rate of steam (stm) (100 d, fine line; "steady state" [SS], thick line). The base case is shown in solid gray, and the half air case in dashed black. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
For sites and designs where high temperatures might approach the surface, the air injection rate can be higher than the steam rate. Simulations were run with the air injection rate double that of the steam rate. Doubling the air injection rate keeps both the 40°C and 80°C isotherms about 1 to 2 m deeper than the base case (Fig. 16 ). For the more demanding problem where steam was injected at 10 times the base rate and the 40°C isotherm reached the surface (Fig. 14), a simulation was run with air injected at 20 times the base air rate to determine if doubling the amount of injected air (at high steam injection rates) could keep the 40°C isotherm below the surface. The 40°C isotherm did not reach the surface but was less than a meter away (Fig. 17 ). The shallowmost rise of the 80°C isotherm was kept several meters lower than the 40°C isotherm.


Figure 16
View larger version (13K):
[in this window]
[in a new window]

  		FIG. 16. Temperature distribution comparison for simulations with air injection at double the rate of steam (stm) (100 d, fine line; "steady state" [SS], thick line). The base case is shown in solid gray, and the double air case in dashed black. Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 

Figure 17
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 17. Temperature distribution comparison of 10x base steam (stm) injection rate with 10x base air injection rate (dash line) and 10x base steam injection rate with 20x base air injection rate, double air (solid). Thin line is for 100 d, and thick line is for "steady state" (SS). Vertical axis, depth below ground (m); horizontal axis, radial distance (m).

 
These simulations suggest that varying the air/steam ratio can be used to control how shallow high temperatures rise to some degree. However, ratios ranging from half to double the steam rate only affected the height of rise by several meters. For the high steam injection case, additional air injectors could be located radially surrounding the steam injector or the steam injection rate could simply be decreased. Injection rates for air and steam should be optimized for the goals of a particular site.

Full-scale steam floods would not have a single steam injector but would consist of a series of wells surrounding a target zone that would interact with each other in three dimensions. Simulations of these types of full three-dimensional patterns were given by Hodges (2006). The vertical rise of steam would only need to be controlled if the flood design targeted a shallow zone, in a poorly layered high-permeability material with an extended steam injection rate and/or duration.


   Discussion and Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Steam Flood Modeling
 Results
 Discussion and Conclusions
REFERENCES
 
Injecting cold air above steam injectors appears to be a viable technology for controlling the vertical rise of high temperatures associated with shallow steam floods based on simulation results. It can be more effective than a low-permeability cap placed over a site by keeping high temperatures several meters below the surface. A cap by itself allows high temperatures to rise to just below the cap, subjecting shallow buried objects (e.g., utilities) to high heat and subjecting the cap to heating and possible alteration that could allow steam to escape.

Controlling the rise of steam by injecting air has the advantages of allowing higher steam injection rates that give a wider lateral spread while limiting the rise of high temperatures, pushing the steam front and higher temperatures deeper. Injecting air keeps the near surface cooler, allowing shallow source zones in poorly stratified permeable material to be targeted. For some designs, injecting air can improve the spread of the highest temperatures within the steam pattern when injecting only steam can leave some areas below the highest temperatures. For other designs, injected air can disperse energy to the point of preventing the temperature at the center of the steam pattern from reaching the highest temperatures. Injected air can also be used to improve the spread of high temperatures within a target zone for deeper steam floods.

Steam injection in the vadose zone for source treatment is typically accompanied by soil vapor extraction. The primary disadvantage of using air injection to control steam zone growth is the need to treat a larger volume of contaminated air. The injected air is likely to become contaminated and therefore should be extracted and treated. The transport and fate of contaminants during steam injection, air injection and vapor extraction were not explored in the current study but could significantly impact field design and operations.

The most effective design, based on the two-dimensional simulation results, combines air control with a cap. The cap accentuates the effect of injected air by preventing it from escaping at the surface and spreading its effects laterally, allowing lower air injection rates. Also, during a shallow steam flood, the possibility of "fast" paths that could allow steam to vent to the surface is more of a threat than deeper floods where any fingering is collapsed by temperature losses to the surrounding media.

The simulation results suggest that air injection is effective even in poorly stratified high-permeability material, although the buoyant rise limits the downward spread of steam at the bottom of the steam front even in the vadose zone. This makes the depth of steam injection relative to a target zone a more crucial variable for high-permeability sites than for lower-permeability sites. The buoyancy effect is decreased in lower-permeability material, allowing the steam front to grow laterally and downward more than in high-permeability material.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Steam Flood Modeling
 Results
 Discussion and Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 Vadose Zone JHome page
J. Simunek and S. A. Bradford
Vadose Zone Modeling: Introduction and Importance
Vadose Zone J., May 27, 2008; 7(2): 581 - 586.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hodges, R. A.
Right arrow Articles by Falta, R. W.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Hodges, R. A.
Right arrow Articles by Falta, R. W.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hodges, R. A.
Right arrow Articles by Falta, R. W.
Related Collections
Right arrow Chlorinated Hydrocarbons
Right arrow Volatile Organic Compounds
Right arrow Heat Transport

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome