DNAPL is heavier than water and is sparingly soluble in water (often less than 1 part per thousand), yet the small amounts that do dissolve are harmful to exposed humans and other members of the global ecosystem, even at concentrations nearly a million times lower than solubility. At an aqueous solubility of 1 part per thousand, a little bit of DNAPL can contaminate a lot of water. Probably the most common DNAPL contaminants are chlorinated solvents ubiquitously used in nearly all industrialized countries for degreasing metals and in dry cleaning since the early to mid-twentieth century. Until the 1970s and 1980s the spent solvents were often released directly into or onto the ground.
A persistent question for scientists and engineers is “how far does DNAPL travel away from its source?” The answer to this question significantly affects the choice of treatment technologies and the potential effectiveness of deployed treatment(s). A technology that targets dissolved phase groundwater contamination may perform poorly in the presence of DNAPL and a technology that targets DNAPL (if not actually needed) will be expensive and may create adverse collateral impacts due to the need for aggressive treatment chemicals or energy. Our research confirmed, based on multiple lines of evidence, large (km) scale migration of DNAPL at a case study site. This suggests that larger scale DNAPL transport be considered – along with groundwater flow and diffusion processes -- at sites that had significant historical DNAPL releases.
Like groundwater in an aquifer, once DNAPL is in the subsurface, gravity moves it through complicated pathways in the pores of materials comprising unconsolidated aquifers (sand, silt, and clay) or in fractures within rocks. The hydraulic (potential) head of the dense, contiguous DNAPL can displace water in materials with larger pores like sand and sometimes silt, but generally not in clays where the properties of the water saturated, smaller pores can resist DNAPL entry. When it is unable to penetrate a material, DNAPL can move and spread laterally under the force created by the hydraulic head.
Like groundwater flow, DNAPL moves in tortuous 3-dimensional pathways. Think of a meandering river or stream and then add a third dimension.
It can even move due to gravity in opposition to the prevailing groundwater flow direction. When DNAPL is no longer released and its hydraulic head dissipates, movement in the subsurface mostly stops and dissolution and diffusion in groundwater slowly become the prevailing force affecting the contaminant’s fate. Groundwater flow continues and can isolate or occlude DNAPL in single pores or small contiguous groups of pores.
Because one drop of DNAPL can contaminate more than 16,000 liters of water, it is useful to know where DNAPL is and even more useful to remove or destroy it. Focusing very aggressive remedial efforts near the contaminant release or source area where most of the DNAPL is likely to reside is considered a vital step in cleaning up a site. Such treatments limit future dissolution into groundwater and therefore limits contributions to a “plume” of dissolved contamination that expands away from the source due to groundwater flow. At sites where a lot of DNAPL was released, the groundwater plume is often extensive with varied volumetric configurations – typically attributed to contaminants emanating from the small source area. Most assume that these contaminated groundwater plumes are created purely from groundwater passing through the source area, contacting the DNAPL and carrying dissolved contaminants with them. However, even at sites where the source area near the release location has been remediated, contamination in the distant plume persists longer than would be expected.
For the past two decades, this persistence in the distant groundwater plume has been attributed by most environmental professionals to the slow release of dissolved contaminants that had steadily diffused out of the expanding groundwater plume into the aquifer materials and clay layers. The slow release, or back diffusion, presents a significant problem for remediation because the diffusion process is very slow and limits effectiveness by preventing effective removal by pump and treat and hindering access to contaminants by in situ treatment amendments. If it took decades for the contaminants to diffuse in, it could take approximately that long for an amendment to do the same or for contamination to back diffuse out into pumpable water. This means that contaminant cleanup in many cases, excluding the few places where the aquifer material could actually be removed, may be so costly as to be impractical. Back diffusion is certainly occurring, however we have shown in this work that it is not the only process that can create a persistent contaminant plume far from the source release area.
Although nearly all environmental professionals acknowledge that it is possible for DNAPL to travel very long distances in the subsurface limited only by the balance of forces driving or resisting it, few consider DNAPL moving very far in actual field conditions. There are several reasons for this but probably the main one is that DNAPL is rarely detected more than a few tens of meters from the spot where it was released. In fact, it is rarely detected even in the release area. DNAPL generally occupies a very small portion of the aquifer volume being investigated (<< 1%) and the tools available for subsurface DNAPL detection often don’t have the resolution to directly identify tiny DNAPL rivulets or droplets. But you don’t need much DNAPL to cause problems. In fact, you would need 1,000 times more back diffusion volume than DNAPL to contaminate the same amount of groundwater.
From extensive investigations and experience with a contaminated site in South Carolina, we have long held that DNAPL transported by gravity driven flow along structurally controlled pathways on the surface of a contiguous clay over very long distances has been a significant contributor to the persistent, kilometers long chlorinated solvent plume. With the data from a field test conducted in a remote portion of the plume nearly a kilometer from the source release area, we believe we have proved that theory. By opportunistically using the unique properties of metallic mercury, a proportionally minor co-contaminant originally released with the waste chlorinated solvent DNAPL, plus other lines of evidence, such as the release of excess chloride following addition of oxidants that destroy DNAPL molecules containing chlorine atoms. Our data show compelling evidence of DNAPL far from the release area. Because of the low background concentrations of mercury and chloride at the site and sensitive detection capability, even the small amounts released from the DNAPL are significantly discernible.
If DNAPL is an important contributor to persistent contaminant plumes, it may be possible to target these remote areas using DNAPL centric treatments in the specific intervals or at interfaces where the DNAPL is present. This strategy has the potential to facilitate cleanup at persistent contaminated sites or at least to reduce the timeframe for remediation.
Paraphrasing and horribly contorting the Bobby Russell song (and Roger Miller’s performance in 1968), “God didn’t make the little DNAPLs…” but maybe we can clean them up.
(Figures 1 and 2 are artistic computer generated representations created by Daniel Dubno, of Blowing Things Up, LLC, using DALL.E with prompts “toxic chlorine droplets flow through sand”, and “toxic chlorine droplets make rivulets in sand”, respectively.)