Why Natural Attenuation has Emerged as a Viable Remediation Approach
The primary motivation for considering natural attenuation as a potential remediation technology is the lack of efficient, cost-effective remediation technologies that can deal with the large number of contaminated sites across the nation. Pump-and-treat, for example, has proven to be an inefficient and costly technology due to "tailing" and "rebound" phenomena. A study conducted by the National Research Council (NRC) Committee on Groundwater Cleanup Alternatives concluded that groundwater pump-and-treat methods could require untenable operating periods (in some cases centuries) to restore groundwater to drinking water standards at NAPL-impacted sites (NRC, 1994). To evaluate the performance of pump-and-treat systems, the NRC Committee established a four-category site rating system characterizing the relative practicability of groundwater cleanup. The NRC surveyed 77 sites with operating pump-and-treat systems, classified them in one or two of the four categories, and noted the number of sites where cleanup goals had been achieved. The survey results are summarized below (note that some sites were assigned to two different categories, making the total more than 77):


According to the NRC, the key technical reasons for the difficulty in cleanup are:

• Physical heterogeneity, making groundwater migration pathways difficult to predict;
• Migration of contaminants to inaccessible regions, such as clays or small pores in aggregates;
• Sorption of contaminants to subsurface materials;
• Difficulties in characterizing the subsurface, making knowledge of the subsurface incomplete; and
• Presence of NAPLs, creating long-term continuing sources in the subsurface.

Other commonly used remediation technologies such as enhanced in-situ biodegradation and air sparging also are hampered by these and other technical constraints. For example, enhanced in-situ aerobic biodegradation can employ oxygenated water that is circulated through the affected aquifer zone to enhance localized biodegradation of oxidizable organic constituents. Despite the goal of enhancing a pumping system with a biologically destructive process, the mass flux of oxygen that can be delivered is relatively small compared to the potential mass of organics in the source zone. Because enhanced in-situ aerobic biodegradation requires circulating groundwater through the source zone, better performance is achieved in homogeneous aquifers with relatively high permeabilities. These conditions (no NAPL, favorable hydrogeology) are found at relatively few sites.

In the air sparging process, air is injected directly into the saturated zone, forming small, continuous air channels from the sparge well to the water table. Dissolved organics are removed from the groundwater through volatilization, and dissolved oxygen is introduced into the contaminated zone. This technology is generally used to treat fuel hydrocarbon plumes, and is better suited for relatively homogeneous media. The air movement is extremely sensitive to small variations in the permeability of the formation, and may cause large portions of the source zone to be by-passed unless very close well spacing is employed (Johnson et al., 1993). For example, if the channels only contact 1% of the NAPL in the source zone, then the mass removal of 99% of the contaminants in the source is limited by the transport of oxygen through the water.

While there are numerous emerging technologies for groundwater remediation, such as oxygen-releasing materials, surfactants, specialized electron acceptor and electron donor applications, none of these technologies represent a cost breakthrough whereby a large percentage of the sites with contaminated groundwater will be restored to pre-contaminant-release conditions. Therefore scientific and regulatory attention has been focused on technologies and regulatory approaches that stress managing contaminants in place for some sites. Natural attenuation is one such technology.

A new class of studies conducted recently, known as "plume-a-thon" studies, have provided tremendous insights into petroleum hydrocarbon plume behavior and have supported using natural attenuation as a remediation technology at many hydrocarbon release sites. The California Leaking Underground Fuel Tank (LUFT) Historical Case Analysis (Rice et al., 1995), for example, reported that plume lengths at 271 fuel hydrocarbon sites in California "change slowly and stabilize at relatively short distances from the release site" (usually less than 250 ft). Of these 271 plumes, 59% were stable, 33% were shrinking, and only 8% were growing. Rice et al. (1995) also indicated that while active (engineered) remediation may help reduce benzene concentrations, "significant reductions can occur over time, even without remediation".

A Texas Bureau of Economic Geology (BEG) study (Mace et al., 1997), based on 217 sites in Texas, found that most benzene plumes (75%) are less than 250 ft long and have either stabilized or are decreasing in length and concentration. In a manner similar to Rice et al. (1995), Mace et al. (1997) found no statistical difference between sites where groundwater remediation activities have or have not been implemented. A Florida risk-based corrective action (RBCA) study (Groundwater Services, Inc., 1997) determined a median length of 90 ft for 117 leaking underground storage tank (LUST) sites in Florida (based on a 50-ppb benzene limit).

The HydroGeologic DataBase (HGDB) (Newell et al., 1990) was re-analyzed by Newell and Connor (1997), who reported a median benzene/BTEX (benzene, toluene, ethylbenzene, and xylene) plume length from 42 service station sites of 213 ft. Newell and Connor (1997) additionally compared the above-mentioned four studies and concluded that most hydrocarbon plumes associated with leaking fuel tanks at service stations are under 200 ft long.

While not as well documented as fuel-hydrocarbon-contaminated sites, similar studies have been completed or are underway for chlorinated solvent sites. The HGDB (Newell et al., 1990), for example, reported a median length of chlorinated ethene (tetrachloroethene [PCE], TCE, dichloroethene [DCE], or VC) plumes of 1,000 ft based on 88 sites across the country; and a median length of 500 ft for other chlorinated solvent plumes (defined as trichloroethane [TCA] and dichloroethane [DCA] and based on 29 sites).