J. Rossabi, B. D. Riha, C. A. Eddy-Dilek, B. B. Looney, and W. K. Hyde
Westinghouse Savannah River Company
Aiken, SC 29808

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/ordering.htm

Available electronically at  http://www.osti.gov/bridge/

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

One of the most important technological developments for characterization of unconsolidated sediments in the past twenty years is the direct push method for accessing and probing the subsurface. The cone penetrometer and related direct push technologies, such as the Geoprobe^â, have been increasingly used for geologic and chemical characterization at sites throughout the United States and abroad. In addition to its standard suite of sensors (i.e., tip pressure, sleeve friction, capillary pressure) to determine soil type, the cone penetrometer has been used with innovative sensors and samplers to perform contaminated site assessments and has also been used to install wells, sampling points, and geophysical tools and arrays. By integrating geologic information from the standard cone penetrometer sensor with the depth-discrete chemical and physical information obtained from innovative cone penetrometer-based samplers and sensors, an accurate, rapid and cost-effective characterization can be accomplished. Cone penetrometer tests (CPT) provide quality, high-resolution data. The tests are minimally invasive, rapid, and produce a minimum of investigation derived waste. These attributes are critical to investigative and cleanup operations at large hazardous waste sites with heterogeneous sediments.

Subsurface characterization methods have traditionally focused on the transport of dissolved phase chemical contaminants in groundwater or on the deposition of heterogeneous media. The disposition of dense non aqueous phase liquids (DNAPL) in the subsurface at a given time is the result of physical and chemical driving forces, initial release conditions, and path-sequential, spatio-temporal conditions. The heterogeneous spatial conditions create the most difficulty in finding the DNAPL and predicting its movement. Figure 1 illustrates an idealization of a vadose zone DNAPL release scenario. In this figure, DNAPL is released from or near the surface and travels primarily downward through the open pores of the coarse grain material with some slight lateral spreading. All of the sediments are assumed to have a residual water saturation and remain water-wet. Initially the DNAPL pools at the surface of the finer grain material. Since the fine grain material is not fully water saturated, DNAPL is drawn into the available pore space by capillary forces and moves vertically and laterally into the sediments. Some pores are not invaded by the DNAPL because they are water saturated or occluded and the DNAPL head pressure cannot in these case cases overcome the pore entry pressure. After the DNAPL release has ceased, the coarse grain sediments will not retain significant amounts of DNAPL as a result of advection and volatilization. A gas plume will remain in the coarse grain sediments and DNAPL will permeate the available pore space in the fine grain sediments. Figure 2 illustrates the conceptual model at this point. The pore water in contact with the DNAPL will approach the aqueous solubility for the compound and diffusion through the water will occur. Subsequent water infiltration events and further volatilization will then chemically and physically alter the configuration of the DNAPL as shown in Figure 3. This will result in discontinuous, discrete residual DNAPL distributed through portions of the fine grain materials and pore water in these areas at or near aqueous solubility.

These figures are cartoons that may represent ideal behavior at a very small scale but the well-sorted materials and properties are extreme simplifications of actual subsurface systems. Variable energy depositional and post depositional processes will dramatically increase the heterogeneity at all scales and complicate the behavior of the system.

Figure 4 shows a photograph of soil core typical of coastal plain sediments with fluvial influences consisting primarily of poorly sorted sand and clay. The amount and scale of low permeability materials (white, kaolin clay) mixed with more permeable sand (tan) is daunting when trying to determine non aqueous phase liquid (NAPL) pathways. The red blobs in the figure are gross conceptualizations of where thin, discrete DNAPLs may be located in such a system. As heterogeneous as this system is, there are some recognizable patterns in the core. A horizontal layering, however discontinuous, is clearly evident. Since gravity and permeability contrasts primarily control the flow of DNAPL, general horizontal layering of materials with contrasting permeability suggests strategies to exploit this feature. Specifically, a high resolution, vertical cross section through a volume would logically offer the highest probability of detecting DNAPL or the evidence of it in comparison with sampling or sensing parallel to the layering.

Because of the large heterogeneity of the formation and the small disparate blobs of contaminant, finding DNAPL in a large volume by accessing the subsurface in a vertical line still has a low probability of success. Ideally a method that could probe the entire volume without permanently disturbing it would be desirable. The ability to pervasively penetrate the subsurface is required for this and the acoustic or electromagnetic geophysical methods may have that potential capability (Olhoeft, 1992; Brewster et al., 1995). Unfortunately their resolution is often not fine enough to consistently detect small, dispersed DNAPL blobs. This is primarily because of the heterogeneity of materials and their physical and chemical properties that produce a large noise from which to detect a small signal.

Another promising technology for probing large volumes is the injection and recovery of fluids that interact with the DNAPL. The partitioning tracer test is an example of this and can be an excellent indicator of DNAPL that is in the flow path of the injectate (Jin, et al., 1995). It has limited ability, however, to locate DNAPLs that have settled in zones outside of the advective path of the tracer. Below the water table, this is less of an issue, as DNAPL will tend to reside in more permeable media above low permeability zones. In the vadose zone, however, fine grain materials will draw in and retain DNAPLs. Tracer gas will preferentially flow in the more permeable coarse grain material and not directly into the fine grain zones.

Other common methods that can help focus characterization for DNAPL include sampling and analysis of fluids in the subsurface. Carefully targeted conventional water samples can point to likely source areas by tracing the concentration gradient. Gas samples can do the same, however more care must be taken when collecting these samples to ensure they are representative of a relatively undisturbed plume. Samples collected of both of these media are generally representative of a diffusion profile from dissolution or direct volatilization from near the source area. It is as rare to find samples at the aqueous solubility (or vapor pressure) limit as it is to find separate phase liquid.

Soil sample collection and analysis provides some advantage over fluid collection. With this method very small blobs of residual DNAPL retained in low permeability material (inaccessible to advective sampling methods) can be detected. Given the conceptual model described previously, sampling the low permeability materials in the vadose zone is vital to successfully characterizing for residual DNAPL. Determining the existence of DNAPL using this method is generally made by first quantifying the mass fraction of contaminant in the soil sample. Equilibrium partitioning assumptions are then used to assess whether the amount of contaminant mass exceeds the capacity of the aqueous, gas, and sorbed phases in the sample volume. The soil sampling and analysis method, like the gas and aqueous phase collection and analysis, provides quantitative chemical information that can be used to determine mass distribution and focus future source zone efforts even if separate phase liquids are not clearly found in the sample.

The obvious limitation of the soil sampling method is that only a very small portion of the subsurface, the recovered volume of the core barrel is sampled. A further complication is that conventional soil sampling is rarely designed to increase the likelihood of finding DNAPL. Conventional soil sampling methods generally specify that samples be taken at predetermined intervals (e.g., 5 or 10 feet) often with little regard for subsurface material. Generally, only the sediment very near the predetermined interval is recovered so selection of a sampling interval is limited although occasionally a screening instrument (e.g., Photo Ionization Detector [PID]) is used to guide the sample location. Given the conceptual model for DNAPL, it is probable that DNAPL will be missed using this strategy.

A better strategy is to analyze the core over the complete vertical interval. Since this is often economically infeasible, collecting the complete core, screening for hot spots and collecting sample from those locations would provide the next best level of data. Collecting the complete core also provides valuable information on the lithology at the site and the ability to revisit the samples for further examination and testing as the conceptual model of the site evolves or as future funding becomes available. Figure 5 shows a plot of soil samples collected using this strategy at Cape Canaveral Air Station (Eddy-Dilek et al., 1999) in comparison with two nearby deployments of a continuous sample collection and separate phase organic liquid sensing method (Ribbon NAPL Sampler). Soil samples from a hollow stem auger drill with a split spoon sampler were collected based on observations of lithology and PID screening at nominal one-foot intervals from the surface to the target depth of 40 feet. From this figure it is clear that contaminant concentrations vary widely with depth. Although the one-foot resolution matches the continuous sampling fairly well, some details are lost even with this small sampling interval. Predetermined sampling intervals at conventional spacing (5-10 feet) rarely captures the structure of the contamination and might miss the DNAPL entirely, although at this site, high concentrations would be detected. High-resolution data are required to accurately assess the residual mass at this location. Because of the limitations on the number of holes that would be practical at a site, ideally a combination of a volume interrogation method and high-resolution vertical method should be used to conduct a cost effective DNAPL characterization program. Direct push technologies may provide the most cost effective platform for high resolution vertical sensing and sampling.

One of the least invasive ways of achieving subsurface access is by using the small diameter (< 3 inch) direct penetration methods. One technique, called a cone penetrometer test (CPT), is performed by pushing an instrumented steel rod into the ground to determine the properties of the penetrated subsurface materials. The standard array of instruments on a cone penetrometer includes tip pressure, sleeve friction and pore pressure sensors. This ensemble is commonly called a piezo-cone configuration. Geotechnical properties, stratigraphy, and soil type of the subsurface materials can be estimated using the data from these sensors. (Lunne et al., 1997).

The CPT is performed using a hydraulic pressure system to deploy the rods and a heavyweight truck to supply the inertial mass. Sensor data are collected and processed electronically with a typical resolution of approximately one-inch. Depths of penetration vary dependent on the subsurface materials. For example, at the Savannah River Site (coastal plain sediments), depths greater than 100 feet are routine with occasional pushes beyond 250 feet, but refusal of penetration has also occurred at depths of less than 30 feet.

In addition to the standard suite of sensors for soil type characterization, many other sensors and samplers have been developed to use the subsurface access platform that the direct penetration system provides (Looney and Falta, 2000). Among these are several tools that focus on NAPL characterization. Optical probing through a hardened window in the CPT push rod has been used to detect DNAPL either directly by Raman spectroscopy (Rossabi et al., 2000a) or inferentially using induced fluorescence measurements of co-constituents dissolved in the DNAPL (Kram, 1998; Rossabi and Nave, 1998). The induced fluorescence method has already gained wide acceptance for subsurface characterization of light NAPL. Visual detection of DNAPL has also been accomplished with a high-resolution video microscope mounted in a CPT probe (Lieberman, 1998). This technology resolves features the size of fine sand grains and can image colored NAPLs in pores. Another important tool developed for a Geoprobe^â system and modified for CPT use is the Membrane Interface Probe (MIP). Using a selectively permeable membrane permitting entry of only volatile organic compounds (VOCs), the tool provides a continuous profile of VOCs through depth (Christy, 1998). Finally, the Ribbon NAPL Sampler (RNS) by FLUTe can be used in conjunction with direct push or drilling methods (Riha et al., 1999; Keller et al., 2000). It consists of a hydrophobic sorbent liner with an impregnated indicator dye that is deployed in a borehole and directly contacts the formation. DNAPL is wicked into the liner and causes a color change in the dye. The liner is retrieved from the borehole and colored spots and their depths are logged as DNAPL detect locations. The RNS samples for DNAPL over the complete circumference and vertical extent of the borehole. This has been one of the most consistently robust DNAPL characterization techniques to date.

Other sensors for the cone penetrometer can also be used to help refine and enhance DNAPL characterization efforts. These tools may not sense DNAPL but the information they provide contributes to the weight of evidence suggesting the presence or absence of DNAPL when it is not directly detected. They also can help focus subsequent DNAPL characterization activities by characterizing the other physical or chemical properties of the sediments at discrete depths. Besides the basic soil classification sensors, electrical conductivity, soil moisture, and permeability sensors are examples of tools that significantly contribute to the development of a conceptual model of a site (Rossabi et al., 2000b) and to the choice of tools to further refine that model. For example, the transport of gases and nonaqueous phase liquids (NAPLs) will be significantly reduced when they encounter a nearly water saturated stratum of fine grain sediments. A less water-saturated zone will allow free transport of the gas and often increased receptivity of the NAPL by imbibition through capillary suction. This information can be incorporated to identify zones likely to harbor DNAPL and help select strategies and depths for sampling, probing, and focusing tracer tests. Other uses of cone penetrometer data might include the specification of boundary conditions and degree of heterogeneity for incorporation in numerical models. Overall the cone penetrometer’s ability to collect many different types of corroborating or complementary information make it an ideal match to a weight of evidence or toolbox approach to DNAPL characterization.

The most compelling aspect of cone penetrometer data is their vertical resolution. The standard penetration rate for cone penetrometers is 1 inch/second and data acquisition systems accompanying these tools acquire data at a rate of approximately 1 Hz. The 1-inch resolution contrasts with conventional borehole logging tools whose resolution is often on the order of three feet. Even recovered core is rarely described at a resolution of less than 1 foot. As the flow and transport of subsurface contaminants is often strongly influenced by thin, discontinuous layers, the identification of these layers is critical. A 1-inch layer of nearly saturated clay may prevent the downward migration of a non-aqueous phase fluid.

With the exception of the Membrane Interface Probe, the NAPL specific CPT sensors are essentially binary sensors that determine presence or absence of DNAPL that is in direct contact with the sensing element on the direct penetration tool. Gas, liquid, and soil samplers for direct penetration are available but until the recent development of the CPT wireline sampling system, only the gas and liquid samplers could be used to obtain multiple samples cost effectively.

One of the difficulties of direct push sampling is that in order to remove a sample the entire cone rod section must be removed, an operation that takes time and increases cost. The cone rods do not require removal when using the wireline CPT sampling system. Like other direct push soil sampling systems (e.g., MosTap^â, Vertek^â, etc.) the wireline tool provides a significant advantage over conventional vadose zone soil sampling methods when waste minimization is an issue. The small diameter and lack of soil cuttings of the direct push tools eliminates most of the waste associated with subsurface investigations. The wireline offers a further advantage when several samples are desired by dramatically saving time and energy. Applied Research Associates, Inc. under a grant from the Department of Energy’s National Energy Technology Laboratory (NETL) program developed the CPT wireline system (Farrington, et al., 1999).

Direct push soil samples are generally obtained by advancing a core barrel at the penetrating end of the direct push rods. At the desired sampling depth, the tip preventing soil from moving into the rods is unlocked and the rods are advanced the length of the core barrel to collect the soil sample. The tip is displaced up the inside of the rods by the collected soil column. All of the rods are then removed from the hole to recover the core barrel and then reinserted to collect the next sample. As the core barrels are usually less than 2 feet long and the depth of penetration can be greater than 100 feet, large amounts of time, energy and money are consumed using this method. The wireline soil-sampling tool uses a removable core barrel that fits inside the push rods. With this system, the cone rods are pushed to the desired depth (either using a piezocone or blank tip at the end of the rods). The tip is removed through the rods using a tether wire, which releases a locking mechanism that holds the tip in place during penetration. The core barrel is then dropped through the rods and locked in place to collect the soil at that depth. After advancing the rods to collect soil in the barrel, the core barrel is removed by its wire tether and rapidly brought to the surface. Another core barrel, a blank tip, or piezocone tip is then re-deployed at the end of the push rods and the rods are advanced until the next sample collection depth.

Because the push rods are not removed and reinserted during this process, soil collection is much faster and less energy intensive then the conventional direct push collection method. Using this method, consecutive one-foot long core samples were consistently collected at the Savannah River Site at the rate of one every two minutes in pushes to depths of 60 feet. Approximately 50 to 100 samples were collected per day.

The wireline core barrel is one foot long with an inner diameter of 1-inch (o.d. = 1 1/8 inch). The small diameter of the core barrel helps ensure good recovery of the soil samples yet is large enough for adequate geologic description and soil sampling for chemical analyses of contaminants. The sampler requires the use of push rods slightly larger in diameter (o.d. = 2 inch) than the more common 1.75 inch diameter rods but most penetration systems can adapt to this size.

The wireline system can also be used to deploy a small diameter piezocone and a grout module. Using these tools in concert, geologic strata can be precisely determined, several soil samples of the desired interval(s) can be collected and the hole grouted in a single push, preventing the possibility of downward migration during site characterization. Figure 6 shows a version of the CPT wireline system with the small diameter piezocone installed through the cutting mouth. Current limitations of wireline design preclude the use of the wireline system below the water table where sediments have the opportunity to flow into the open core barrel during removal and reinstallation of a wireline tool. The principal advantage of this method for sampling in the vadose zone is the ability to collect a continuous core for chemical analysis rapidly, cost effectively, and with essentially no waste.

Since no single tool has been developed capable of completely characterizing a DNAPL contaminated subsurface, several tools should be used to contribute data to the conceptual understanding of the site. The noninvasive geophysical tools provide an excellent picture of the larger geologic features but often don’t have the resolution to discriminate discrete DNAPL. Partitioning tracer methods detect NAPL in and near the tracer’s advective flow path but cannot easily access residual NAPL in low permeability zones. In situ sensing methods determine NAPL presence or absence at discrete points, but these methods only provide information about sediments immediately contacting the sensor. Sample analysis is also ostensibly limited to the relatively small zone where the sample is collected. However, this method provides more information on the disposition of the contaminant at the sample point than simple presence or absence of DNAPL.

The analysis of samples without obvious separate phase product indicate the effects of advection, diffusion, and reaction on the source contaminant as it is realized at a particular point in space and time. When properly constructed and with appropriate attention to the geology, these measurements can legitimately be used in contour projections of the site. Contour maps are useful in the development of the site conceptual model and are used in nearly all site investigations. The traditional drawback of soil samples is the cost of collection of these samples. With the ability to cost effectively collect high vertical resolution soil samples it is beneficial to revisit the analysis of these samples with regard to DNAPL characterization.

Although quantitative data on the chemical mass fraction is most important in understanding the contamination at a site, there is a need to determine the presence or absence of DNAPL based on these measured values. One useful tool for assessing whether a site has separate phase contamination based on measured contaminant mass fractions from soil samples is presented below. The measured mass fraction is compared with the theoretical threshold mass fraction for which DNAPL must exist based on equilibrium partitioning assumptions (i.e., pore water at aqueous solubility for the compound and gas at the compound’s vapor pressure). This technique is clearly described by Feenstra et al. (1991) and Cohen and Mercer (1993). The equation describing equilibrium partitioning for a single chemical component in a multiphase system is given by:

,

where C_Total is the total concentration of the chemical in the bulk sample, f is the porosity, S_gas, S_water and S_NAPL are the volumetric pore saturations of gas, water, and NAPL, r _bulk and r _NAPL are the bulk and NAPL densities, K_d is the partitioning coefficient of the chemical to the sediments, and H is the Henry’s Law constant for the chemical. C_water is the concentration of the chemical in the aqueous phase and is equal to the aqueous solubility when the sample contains a separate phase liquid.

Figures 7 through 10 show threshold soil concentrations for which tetrachloroethylene (PCE) and trichloroethylene (TCE) separate phase liquid must be present given the properties of the contaminants and various soil parameters. Table 1 lists the physical and chemical parameters for PCE and TCE used in the equilibrium partitioning calculations.

Finally, based on equilibrium partitioning assumptions, Figures 11 and 12 show the relationship between NAPL saturation and mass fractions of PCE and TCE in soil samples. These curves are useful for roughly assessing residual saturation based on soil sample analysis.

Although a single, robust detection system does not exist for DNAPL, we feel that DNAPL sites can be efficiently characterized using a strategy emphasizing high-resolution vertical sampling. Apart from high-resolution vertical data, the fast and inexpensive access to the subsurface provided by the direct penetration tools and minimal production of investigation derived waste makes it an ideal tool for dynamic site investigations. In this role direct penetration methods have hastened the development of new protocols that substantially reduce characterization times and increase characterization accuracy.

Wireline methods substantially enhance the utility of direct push technologies for fast and efficient collection of samples. The cone penetrometer and other direct push methods provide a unique advantage over conventional drilling in the acquisition of soil and soil gas samples while producing a minimum of investigation derived waste. In many cases, boreholes can be grout sealed upon removal of the rods further reducing the chances of cross contamination and exposure. Wells (currently up to 2 inches in diameter) can also be installed directly into the formation using the direct push methods. Another use of direct push technologies is the injection of fluids for reaction with contaminants or creating fractures in low permeability soils. A more complete list of direct push tools can be found on pages 186-201 of Vadose Zone: Science and Technology Solutions (Looney and Falta, 2000).

1. Brewster, M.L., Annan, A.P., Greenhouse, J.P., Kueper, B.H., Olhoeft, G.R., Redman, J.D., and Sander, K.A., 1995, Observed migration of a controlled DNAPL release by geophysical methods: Journal of Ground Water, Vol. 33, No. 6, pp. 977-987.
2. Christy, T., 1998, A Permeable Membrane Sensor for the Detection of Volatile Compounds in Soil. In Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems, Chicago, IL.
3. Cohen, R. M., and J. W. Mercer, 1993, DNAPL Site Evaluation: C. K. Smoley-CRC Press, Boca Raton, Florida.
4. Eddy-Dilek, C. A., B. D. Riha, D. Jackson, and J. Rossabi. 1999. DNAPL Source Zone Characterization of Launch Complex 34, Cape Canaveral Air Station, Florida. WSRC-TR-99-00024, Westinghouse Savannah River Co., Aiken, SC 29808.
5. Farrington, S. P., M. L. Gildea, and J. C. Bianchi, 1999, Development of a wireline CPT system for multiple tool usage. In Proceedings of the 1999 Industry Partnerships to Deploy Environmental Technology Conference, Morgantown, WV, Oct. 12-14. National Technical Information Service: Springfield, VA.
6. Feenstra, S. D.M. Mackay, and J.A. Cherry, 1991. A method for assessing residual NAPL based on organic chemical concentrations in soil samples: Groundwater Monitoring Review, 11(2):128-136.
7. Jin, M., M. Delshad, V. Dwarakanath, D. C. McKinney, G. A. Pope, K. Sepehrnoori, C. Tilburg, and R. E. Jackson, 1995, Partitioning tracer test for detection, estimation and remediation performance assessment of subsurface nonaqueous phase liquids: Water Resour. Res. 31(5):1201-1211.
8. Keller, C., J. Rossabi, B. D. Riha, J. Walker, B. Nashold, and P. Werenko, 2000,Utility of Flexible Liners Installed via Push Rods. In Proceedings of the The Third International Symposium on Integrated Approaches to Site Characterization, Chicago, IL, September 12-14, 2000.
9. Kram, M., 1998, Use of SCAPS petroleum hydrocarbon sensor technology for real-time indirect DNAPL detection:. Journal of Soil Contamination, 17(1), 73-86.
10. Lieberman, S.H., and D.S. Knowles, 1998, Cone penetrometer deployable in situ video microscope for characterizing sub-surface soil properties: Field Analytical Chemistry and Technology 2(2):127-132.
11. Looney, B. B. and Falta, R. W., eds., 2000, Vadose Zone, Science and Technology Solutions: Battelle Press, Columbus, OH, 1540 p.
12. Lunne, T., P. K. Robertson, and J. J. M. Powell, Cone Penetration Testing in Geotechnical Practice: Blackie Academic & Professional, London, 1997.
13. Olhoeft, G. R., 1992, Geophysical Detection of Hydrocarbon and Organic Chemical Contamination. In: Proc. Symposium on the Application of Geophysics to Engineering and Environmental Problems, SAGEEP, April 26-29, Chicago, IL.
14. Riha, B. D., J. Rossabi, C. A. Eddy-Dilek, and D. Jackson, 2000, DNAPL Characterization Using the Ribbon NAPL Sampler: Methods and Results. In Proceedings of the Second International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterey, CA, May 22-25, 2000.
15. Rossabi, J. and S. E. Nave, 1998, Characterization of DNAPL using Fluorescence Techniques (U), WSRC-TR-98-00125, Westinghouse Savannah River Company, Aiken, SC 29808.
16. Rossabi, J., B. D. Riha, J. W. Haas III, C. A. Eddy-Dilek, A. G. Lustig Kreeger, M. Carrabba, W. K. Hyde, and J. Bello, 2000a, Field tests of a DNAPL characterization system using cone penetrometer-based Raman spectroscopy, Ground Water Monitoring and Remediation, 20 (4), pp 72-81.
17. Rossabi, J., B. B. Looney, C. A. Eddy-Dilek, B. D. Riha, and D. G. Jackson, 2000b, DNAPL site characterization: The Evolving conceptual model and toolbox approach. In Treating Dense Nonaqueous-Phase Liquids (DNAPLs): Remediation of Chlorinated and Recalcitrant Compounds- Monterey 2000, pp 41-48, Monterey, CA, May 22-25, 2000, Battelle Press, Columbus, Ohio.
    
    
    

Figure 8. PCE/soil mass fraction for which separate phase liquid must occur (f_oc = 0.001).

Figure 9. TCE/soil mass fraction for which separate phase liquid must occur (f_oc = 0.0001).

Figure 10. TCE/soil mass fraction for which separate phase liquid must occur (f_oc = 0.001).

Figure 11. PCE/soil mass fraction and corresponding NAPL saturations (f_oc = 0.0001).

Figure 12. TCE/soil mass fraction and corresponding NAPL saturations (f_oc = 0.0001).