Introduction

The failure to consider the effect of nonaqueous phase liquid (NAPL) upon dissolved plume persistence is largely responsible for the inability of many early remediation efforts to achieve groundwater cleanup goals, particularly at sites with dense NAPL (DNAPL). Understanding of the significance of NAPL was not widespread until the early 1990s, but significant removal of DNAPL and residual NAPL was generally considered unfeasible. Consequently, the strategy for many sites with known or suspected DNAPL has consisted solely of plume remediation and/or control, and the thorough characterization and removal of subsurface DNAPL has not been considered. Recent development of several promising characterization and remediation technologies has increased interest in NAPL removal. This paper summarizes various emerging NAPL characterization and remediation techniques are summarized.

NAPLs are compounds that are immiscible in water and represent one of the most common and problematic class of groundwater contaminants. NAPLs provide a continual source of dissolved constituents to the groundwater and greatly increase the complexity and scope of groundwater remediation efforts. Additionally, vapors from the NAPL may spread contamination through the vadose zone to areas beyond the location of the original spill. The volume and duration of the release, the area of infiltration, properties of both the NAPL and subsurface media, and subsurface flow conditions affect NAPL migration in the subsurface.

Light NAPLs (LNAPLs) are less dense than water, and are typically petroleum-related fuels and distillates. Due to their buoyancy, the distribution of LNAPLs in the subsurface is generally limited to the vadose zone and the water table low. Synthetic chlorinated solvents, coal tars, and the heavy fractions of petroleum distillation are generally classified as DNAPLs because they are denser than water. Sites with DNAPL contamination are categorically more difficult to investigate and monitor, and can be exceptionally challenging to remedy. DNAPLs initially migrate as a continuous phase under the influence of capillary, viscous, and gravity forces. If present in sufficient quantity, DNAPL can continue to migrate below the water table. Along the migration path, DNAPLs can dissolve into the groundwater, sorb to soils, become trapped within interstitial pore spaces, and settle onto impermeable surfaces. This complex distribution is controlled by the physical properties of the DNAPL and geological conditions. Specific NAPL architectures are commonly referred to as “pools,” “fingers,” and residual “blobs.”

NAPLs represent a source of groundwater contamination that can persist for decades. By definition, NAPLs exhibit low aqueous solubility (typically from less than 1 g/L to 20 g/L); however, drinking water standards for these compounds may be five to six orders of magnitude lower (typically 2 g/L to 100 g/L). Consequently, relatively small NAPL volumes in the subsurface can act as long-term contaminant source zones producing large and persistent groundwater plumes.

Until relatively recently (early 1990s), the significance of the presence of NAPL was not well understood, or was ignored at most sites. This was particularly true for sites with DNAPL, where detection and accurate characterization of source areas is substantially more difficult than LNAPL sites. Meaningful and cost-effective recovery of DNAPL and immobile, or residual, NAPL was not considered possible. Consequently, the strategy for many sites with known or suspected DNAPL has consisted solely of plume remediation and/or control. Recent development of several promising characterization and remediation technologies has increased interest in NAPL removal. It should be noted, however, that as NAPL characterization and remediation become better understood, the importance of other processes that complicate site remediation (sorption/desorption, rate-limited transport from low-flow zones, etc.) should continue to be investigated and evaluated.

NAPL Detection and Characterization

Generally, successful and cost-efficient site remediation depends on accurate characterization of contaminant mass and distribution. This is particularly true for NAPL sites, since NAPL zones greatly influence plume geometry and represent a majority of contaminant mass. Historically, DNAPL zone characterization has been largely inadequate; however, recent work has greatly improved characterization technique, often allowing estimation of NAPL chemical composition, saturation, total volume, and distribution. However, additional work is necessary to better understand the limitations of these new techniques.

Soil Quality Data

Chemical analysis of soil samples is frequently used to characterize the presence of NAPL, both in the vadose and saturated zones. Several drilling and soil core collection methods are common and soil sampling activity is typically coordinated with other site investigation activity, such as subsurface geology characterization and monitoring well installation. Drilling techniques allow for nearly continuous vertical soil characterization, which can be evaluated in the context of site geology to infer the possible NAPL architecture (i.e., pooling or residual). Methods have been developed to quantitatively determine NAPL composition and saturation based on soil quality data (Mott 1995 and Mariner et al. 1997). The user-friendly software NAPLANAL (based on Mariner et al. 1997) estimates NAPL composition and saturation from soil quality data, and is available free on the Internet at http://www.napl.net. NAPL saturation estimates from soil cores provide only discrete information; therefore, the presence of NAPL between sampling locations is interpolated, rather than measured directly.

Groundwater Quality Data

Analysis of groundwater quality data from monitoring wells is probably the easiest and most frequent technique to observe mobile LNAPL and to infer the nearby presence of DNAPL. Mobile LNAPL floating on the water table can be observed with appropriately screened monitoring wells. DNAPL is infrequently observed in monitoring wells; however, the presence of subsurface NAPL is generally suspected if dissolved contaminants are present in concentrations greater than one percent of their aqueous solubility (common rule-of-thumb). Groundwater concentrations less than this screening threshold do not necessarily confirm the absence of NAPL, since monitoring well location, construction, and sampling affect the degree of contaminant dilution. Additionally, the effective aqueous solubility of a compound is reduced in multi-component NAPLs (as described by Raoult’s Law). It should also be noted that high dissolved concentrations by themselves do not positively identify NAPL, since the original subsurface contaminant source may have been water with elevated constituent concentrations, and not NAPL.

Partitioning Tracer Tests (PTTs)

The partitioning tracer test (PTT) is a promising new NAPL characterization technique that is based on methods developed in the oil industry to quantify residual hydrocarbons in reservoirs. During a PTT, the delayed transport of tracers that partition into NAPL (usually alcohols or dissolved gases) is compared to the transport of conservative, or non-partitioning tracers, to directly measure subsurface NAPL. The volume of NAPL and the average pore-space NAPL saturation is determined from the known equilibrium partition coefficients and the observed retardation of the partitioning tracers. The delayed transport of dissolved helium and neon (partitioning tracers) relative to bromide (conservative tracer) is due to the presence of residual NAPL in the test volume.

PTTs require careful design and are generally more costly than soil or groundwater analysis; however, unlike these more common NAPL characterization techniques, the PTT directly measures NAPL over a relatively large volume of the subsurface. Successful PTT implementation requires effective hydraulic control, frequent tracer collection and analysis, and appropriate tracer selection. Additional work is needed to further understand the importance of tracer detection limits, spatial variations in NAPL composition, and geologic heterogeneity upon PTT accuracy and reliability. For examples and discussion about the PTT technique, see Jin et al. (1995) and Divine (2000).

Additional Characterization Techniques

Several additional NAPL characterization techniques that have been used with varying success include (see Pankow and Cherry 1996 and Semprini et al. 1998):

• Visual identification of NAPL in soil samples using UV light or a hydrophobic dye (i.e., SUDAN IV);
• Soil gas surveys;
• Subsurface geophysics, including electromagnetic resistivity and ground penetrating radar; and
• Radon flux analysis.

NAPL Remediation

As mentioned previously, recent work in NAPL removal or destruction is encouraging and has greatly increased interest in DNAPL zone remediation. Some results from laboratory and pilot-scale data are very promising, and cases where NAPL mass removal greater than 90% have been reported. However, significant challenges remain, particularly in the complete remediation of trapped residual NAPL, and at sites with fine-grained soils or fractured clay/bedrock. NAPL dissolution rates are largely independent of NAPL saturation; consequently, meaningful reductions in dissolved plume size and longevity for recalcitrant DNAPL compounds may require 90% to 99% DNAPL remediation efficiencies. Furthermore, Sale (1998) notes that the restoration of groundwater quality to drinking water standards within a DNAPL zone has yet to be documented.

Soil Vapor Extraction

Soil vapor extraction (SVE) is a proven method for removing volatile NAPL components from the vadose zone and near the water table by inducing significant subsurface airflow in the contaminated zone. Constituents with sufficiently high vapor pressures (greater than 1 to 2 mmHg) will volatize and be transported by the soil gas to the surface. Because SVE is based on phase partitioning, it can effectively remediate NAPL, contaminated pore water in the unsaturated zone, and sorbed contaminants. SVE design may be based on experience and empirical data for simple projects, or may require pilot-scale testing and/or system modeling. Simple analytical models (see Johnson et al. 1988) as well as complex numerical models are available for SVE design and optimization.

Surfactants and Cosolvents

Recent advances in flood-based NAPL remediation strategies include the use of surfactants and cosolvents to increase NAPL mobility and enhance contaminant recovery in pumping wells. Surfactant-enhanced NAPL removal is based on two processes: 1) micellar solubilization of the entrapped NAPL and 2) displacement or mobilization of the NAPL due to interfacial tension reductions. Cosolvent flushing, which is based on methods developed in petroleum engineering, involves the use of cosolvents, such as water-miscible alcohols, ketones, and sugars, to solubilize and mobilize the NAPL. Surfactants and cosolvents have been tested at the bench scale in laboratories, at the pilot scale in controlled and contained environments, and at full scale in actual field studies (see Broholm and Cherry 1994, Jin et al. 1995, Annable et al. 1996, and McCray et al. 2000).

Promising results have been documented from several field-scale surfactant and cosolvent flushing efforts; however, these technologies require intensive site characterization and careful design. In particular, hydrologic control, surfactant/cosolvent selection, and the risk for uncontrolled NAPL remobilization need to be thoroughly assessed prior to remediation activity through laboratory bench-scale experiments and numerical modeling. Additionally, potential regulatory concerns related to the injection of surfactants/cosolvents may need to be addressed

Thermal Technologies

Thermal, or heat-based, remediation technologies essentially consist of applying heat to the subsurface to remove volatile contaminants from NAPL, soil, and water. A significant advantage with thermal technologies is that contaminant mass-transfer limitations of traditional remediation techniques are not encountered, and laboratory-scale research indicates that heating the subsurface to steam-temperature may allow nearly complete removal of DNAPL (Udell 1996). Most frequently, steam is injected into the subsurface and extraction wells are used to remove volatized contaminants. Dorrler et al. (2000) describe a successful DNAPL remediation effort that consisted of groundwater dewatering, the application of high vacuum to establish cross-hole flow paths, and forced hot-air injection with an in-well heating system.

Six-phase heating (SPH) involves the conduction of electricity through the subsurface via a neutral electrode surrounded by six charged electrodes. The surrounding electrodes are charged sequentially and the resistance of flow of electrical current generates heat and steam within the subsurface. For an example of this technology applied at the field scale to remedy DNAPL, see Beyke et al. (2000). The successful application of a thermal remediation technology may require significant design effort including an assessment of subsurface response to heating, contaminant behavior, vacuum extraction design, and groundwater control and dewatering.

Chemical Oxidation

This technique uses strong oxidizing agents such as hydrogen peroxide and iron (Fenton’s reaction), potassium permanganate, and ozone to chemically transform contaminants into carbon dioxide, water, and other byproducts. Chemical oxidation can be used to treat a variety of organic contaminants, is largely unaffected by contaminant concentrations, and is relatively inexpensive. Effective delivery of the oxidizing agent to contaminants is a primary remediation design consideration since naturally occurring organics will be oxidized in the absence of contaminants, minimizing remediation efficiencies. Strong oxidizing agents may also have the capacity to precipitate ferric iron and manganese within the aquifer. These agents may also cause the oxidation of immobile metals such as chromium III (Cr3+) to more mobile and toxic forms (i.e., chromium VI [Cr6+]). Additionally, significant health and safety risks exist for workers handling these hazardous oxidizing chemicals.

Plume Control and Treatment

While NAPL removal is an important component to long-term site remediation, management and treatment of the groundwater plume is usually necessary. Dissolved plume control and treatment efforts may be independent of NAPL remediation activity, or designed as part of the overall site remediation strategy. Pump-and-treat is a standard method for plume control, however, new techniques including reactive barriers and enhanced biodegradation offer additional options for dissolved plume management and treatment.

Pump-and-Treat

The effectiveness of traditional pump-and-treat systems in remediating NAPL and their associated groundwater plumes is limited due to the low solubility, higher viscosity, and higher retardation coefficients typically associated with NAPLs. However, pump-and-treat can be an effective method to control dissolved plume migration and is often used in tandem at a site with other remediation and plume treatment techniques. Procedures for pump-and-treat system design and performance assessment are well established, and generally involve pilot-scale testing, modeling, and optimization analysis.

Reactive Barriers

The use of reactive barriers to treat groundwater is a relatively new approach that shows much potential. In essence, a permeable ‘wall’ of reactive material is installed in the groundwater flowpath, which treats contaminated groundwater as it passes through. Commonly, a funnel-and-gate design is used to maximize groundwater capture. A variety of contaminant remediation techniques have been incorporated into reactive barriers, including:

• Transformation via zero-valent iron or other media;
• Sorption via replaceable granular activated carbon (GAC);
• Enhanced biodegradation via nutrient addition; and,
• Volatilization via air sparging curtains.

Accurate hydrogeological site characterization is necessary for successful reactive barrier design. Additionally, bench-scale experiments and modeling is generally required to assess contaminant degradation rates, reactive material life, and the capacity for plugging and bypass flow.

Natural Attenuation and Enhanced Bioremediation

Natural attenuation of contaminants is a combination of nondestructive (dilution, sorption, volatilization) and destructive (biodegradation, oxidation, hydrolysis) processes. Biodegradation can be a significant mechanism for dissolved contaminant destruction; additionally, naturally occurring biodegradation activity can be enhanced by the addition of electron acceptors and/or donors and manipulation of subsurface conditions. The biodegradation of NAPL phase is generally considered negligible, but natural attenuation processes may be sufficient to control plume migration and eliminate the need for NAPL removal at some sites.

Most petroleum fuels are readily metabolized by naturally-occurring microorganisms under aerobic conditions where oxygen is the electron acceptor. Although they are not as thermodynamically favorable, nitrate, sulfate, iron, and manganese can also serve as electron acceptors if oxygen concentrations decrease in the subsurface. The biodegradation of most fuel hydrocarbons is generally limited by electron acceptor availability, and can be enhanced by providing additional oxygen via air sparge wells or Oxygen Release Compound (ORC), or by nitrate injection.

Many chlorinated solvents are recalcitrant to aerobic biodegradation, but may be destroyed under reducing conditions via reductive dechlorination. This process can be limited both by the availability of electron donors and electron acceptors. Reductive dechlorination occurs when indigenous anaerobic microbes (such as acetogens) metabolize organic carbon producing hydrogen. The hydrogen is then used by other subsurface microbes (reductive dehalogenators) to strip the solvent molecules of their chlorine atoms. Reductive dechlorination can be enhanced by the supply of additional carbon sources, such as molasses, vegetable oil, landfill leachate, or coalescing hydrocarbon fuel plumes.

Natural attenuation as a site remediation strategy typically requires the collection of additional groundwater quality data (dissolved oxygen, redox potential, etc) and long-term monitoring. Simple spreadsheet-based screening models (such as BIOCHLOR and BIOSCREEN which are available free at http://www.epa.gov/ada/csmos/models.html) or more complex numerical transport models are commonly used to assess natural attenuation processes and predict future plume behavior.

Summary

The failure to consider the effect of NAPL upon dissolved plume persistence is largely responsible for the inability of many early remediation efforts to achieve groundwater cleanup goals, particularly at sites with DNAPL. Improved NAPL characterization techniques and promising new NAPL remediation technologies have recently increased interest in NAPL removal and source-zone treatment. It is important to note that, despite the recent encouraging results for new NAPL recovery techniques, the restoration of groundwater to drinking water standards in relatively short times requires nearly complete NAPL removal efficiency. Other rate-limited processes such as desorption from aquifer materials and diffusion from non-flowing zones will complicate system design and lengthen remediation times, and must be considered when developing site remediation strategies and goals.

References

Annable, M. D., P. S. C. Rao, R. K. Sillian, K. Hatfield, W. D. Graham, A. L. Wood, and C. G. Enfield, 1996, “Field-scale Application of In-situ Cosolvent Flushing,” In Proceedings of American Society of Civil Engineers Specialty Conference Nonaqueous Phase Liquids (NAPLs) in Subsurface Environment: Assessment and Remediation, Washington D.C., pp. 212-220.

Beyke, G., G. Smith, and V. Jurka, 2000, “DNAPL Remediation Closure with Six-Phase Heating,” In Proceedings of 2nd International Conference of the Remediation of Chlorinated and Recalcitrant Compounds, Monterey, CA, C2-5, pp. 183-189.

Broholm, K. and J. A. Cherry, 1994, “Enhanced Dissolution of Heterogeneously Distributed Solvents Residuals by Methanol Flushing: A Field Experiment,” In Transport and Reactive Processes in Aquifers, Balkema, Rotterdam, pp. 563-568.

Divine, C. E., 2000, The Applicability of Dissolved Helium and Neon as Dense Nonaqueous Phase Liquid (DNAPL) Partitioning Tracers, M. S. Thesis, Colorado State University, Fort Collins, CO.

Dorrler, R. C., S. S. Suthersan, and R. E. Komoski, 2000, “Crosshole Flow/Thermal Technology Used in Remediation of PCE DNAPL,” In Proceedings of 2nd International Conference of the Remediation of Chlorinated and Recalcitrant Compounds, Monterey, CA, C2-5, pp. 141-156.

Jin, M., M. Delshad, V. Dwarakanath, D. C. McKinney, G. A. Pope, K. Sepehrnoori, C. E. Tilberg, and R. E. Jackson. 1995. “Partitioning Tracer Test for Detection, Estimation, and Remediation Performance Assessment of Subsurface Nonaqueous Phase Liquids,” Water Resources Research, Vol. 31, pp. 1201-1211.

Johnson, P. C., M. B. Kemblowski, and J. D. Colthart, 1988, “Practical Screening Models for Soil Venting Applications,” In Proceedings of the Conference on Petroleum Hydrocarbons and Organic Chemicals in Groundwater, Houston, TX, pp. 521-546.

Mariner, P. E., M. Jin, and R. E. Jackson, 1997, “An Algorithm for the Estimation of NAPL Saturation and Composition from Typical Soil Chemical Analysis,” Groundwater Monitoring and Remediation, Vol. 17, No. 2, pp. 122-129.

McCray, J.E., Boving, T., Brusseau, M.L., 2000. “Cyclodextrin-enhanced Solubilization of Hydrophobic Organic Compounds With Implications for Aquifer Remediation,” Groundwater Monitoring and Remediation, Vol. 20, No 1, pp. 94-103.

Mott, M. V., 1995, “A Model for Determination of the Phase Distribution of Petroleum Hydrocarbons at Release Sites,” Groundwater Monitoring and Remediation, Vol. 15, No. 3, pp. 157-167.

Pankow, J. F. and J. A. Cherry, 1996, Dense Chlorinated Solvents and Other DNAPLs in Groundwater, Waterloo Press, Portland, OR.

Sale, T, 1998, Interphase Mass Transfer From Single Component DNAPLs, Ph.D. Dissertation, Colorado State University, Fort Collins, CO.

Semprini, L, M. Cantaloub, S. Gottipatti, O. Hopkins, and J. Istok, 1998, “Radon-22 as a Tracer for Quantifying and Monitoring NAPL Remediation,” In Proceedings from the 1st International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterey, CA, C1-2, pp. 137-142.

Udell, K. S., 1996, “Heat and Mass transfer in Clean-up of Underground Toxic Wastes,” In Annual Reviews of Heat Transfer, Vol 7, Bengell House, Inc., New York, pp. 333-405.
   

Top