The Dakota Technologies team conducted its first field testing of the Dye-LIF system in Columbus, Ohio, from January 31 through February 2. The crew visited the site of a former commercial dry cleaner where, like so many other former dry cleaning sites, chlorinated solvents had leaked. The site had already been well characterized and shallow (10-15 ft) DNAPL had been confirmed, so it gave us the confidence that our developmental probe would be sure to encounter DNAPL. Confirmation of actual DNAPL (not simply inferred from dissolve phase levels) was important since testing at a site with only the strong suspicion of DNAPL would lead to a lot of "head-scratching" if for some reason we weren’t observing the expected response. We had previously tested the site’s DNAPL in the lab and it tested positive for fluorescence with TarGOST. We suspect that this site's DNAPL had picked up oil, grease, fabric brighteners or other fluorophores sometime in its time at the site making it fluorescent.
First, let's remind ourselves of how Dye-LIF is supposed to work. The concept is based on the fact that clean/pure chlorinated solvents don't fluoresce naturally, but they will fluoresce if they've solvated an indicator dye that goes into solution in them (think Sudan or Oil Red dye shake test). Chlorinated DNAPLs also fluoresce if they have solvated other fluorophores (petroleum, textile brighteners, etc) such as our test site DNAPL. Dye-LIF is designed to take advantage of either/both scenarios by outfitting a traditional LIF tool (TarGOST, TarGOST HD, or UVOST) with the ability to inject an indicator dye into the soil from a port located six inches below the sapphire window of the LIF probe. The dye is injected continuously so that as the LIF sensor is pushed down it arrives in the same soil about 6 seconds later by which time the DNAPL is now fluorescent, allowing the Dye-LIF instrument to "see":
- any DNAPL that has fluorophores already dissolved
- any DNAPL that has been exposed to the indicator dye
- the indicator dye itself - which fluoresces less intensely and with different waveforms than DNAPL-solvated dye
The photos show the systems (TarGOST, TarGOST HD, and Dye-LIF) being deployed on site. As you can see, there were some very tight quarters at this indoor/outdoor location! Our team of Steve Adamek and Tom Rudolph was able to log 33 pushes across three different areas of the site (one outdoors) over a three day period for a grand total of 536 feet of testing! Various flavors of LIF logging tested included:
- 6 traditional TarGOST (LIF only)
- 7 TarGOST HD (LIF only)
- 20 Dye-LIF (mix of TarGOST and TarGOST HD sensing)
So what kind of results did we see? The butterfly style plot for two co-located TarGOST pushes (no dye flowing) in Figure 1 shows the inherently fluorescent tetrachloroethene (aka "perc" or PCE) showing up between 11-13 ft on both logs, with some heterogeneously distributed ganglia (say that 5 times really fast) also showing up at 14-15ft – but on only one of the two co-located logs. This small scale heterogeneity was common across the site (not exactly unexpected) making interpretation of any individual log more difficult than a site with a consistent "layer" of LNAPL at the groundwater surface. Of course, we're accustomed to seeing this heterogeneous distribution after 47 miles of TarGOSTing. Remember that coal tar is a DNAPL, and we know our coal tar!
When we switched to Dye-LIF mode (by simply turning the dye-delivery pump on) the TarGOST logs revealed the non-solvated indicator dye's weak fluorescence almost continuously as the Dye-LIF probe was delivered at a normal LIF speed of ~1 inch per second. This is apparent in Figure 2 where you can see the non-solvated dye (magenta color in the Signal plot). Along with the dye you can see a very narrow brighter chartreuse-colored ganglia at 9.42 ft, indicating the dye had solvated into a narrow feature of DNAPL. In many of the other Dye-LIF logs the fluorescence signals were a mixture of the inherent fluorescence of the DNAPL and the fluorescence of the solvated dye (as expected).
Notice the pressure and flow of the injected dye solution that was also monitored continuously with depth as well (at right side of Figure 2). Both flow and pressure appeared to be responding to changes in hydraulic conductivity in a manner similar to commercial profiling systems. We did not have an opportunity to compare the system's hydraulic behavior against other methods to see just how accurately the system might be measuring hydraulic conductivity.
The most intriguing logs came from a location where the Dye-LIF strongly indicated DNAPL at a depth that none of the previous logs (direct LIF or Dye-LIF) had encountered DNAPL. To confirm the Dye-LIF detected what appeared to be dye-indicated DNAPL (response was a dye-solvation phenomenon) we turned off the indicator dye flow and conducted a second log just one foot away. We did this co-located LIF-only log to see if the fluorescence at depth was 1) innately fluorescent DNAPL, 2) "clean" non-fluorescent DNAPL, or 3) a false positive (non-NAPL).
The butterfly in Figure 3 shows the Dye-LIF log on the left (DL-10) with the dye (magenta) and solvated dye (yellow/chartreuse) responses. On the right side of Figure 3 is the log from the same location logged without dye flow (DL-10 No Dye). Notice that without dye flowing that the dye signal (magenta) is missing, and there is also no response from the DNAPL we suspect was residing at approximately 15-18.5 ft bgs. There also was a horizon of modest natural fluorescence at 11-14 feet (source unknown) in both logs. We feel that this is our best evidence of the Dye-LIF performing exactly as we‘d hoped - by making what was invisible DNAPL now visible to LIF. Of course, lateral heterogeneity always lurks in our mind as having played a role (was the suspected "clean DNAPL" simply not at the second location?). We attempted to return to the site later to recover continuous cores at this location to confirm our interpretation – but it simply wasn't possible to work it out with stakeholders.
So, in summary, our first field-test of the Dye-LIF led to the following conclusions:
- The site's DNAPL was fluorescent enough to be logged with standard TarGOST, giving us excellent "controls" on where at least some of the DNAPL was [note that we're finding a LOT of field DNAPLs fluoresce well, especially with TarGOST – we think high UV absorbance "hurts" UV's performance]
- Localized DNAPL heterogeneity was assessed using co-located LIF logs, allowing us to "temper" our interpretations with the knowledge we gained
- The Dye-LIF appears to have worked as designed, and our conclusion is that we most likely mapped a non-fluorescing DNAPL at one location at least
- Waveform differences between solvated and non-solvated dye allow continuous confirmation of successful dye injection
- No hardware failures (sapphire windows, dye ports, fiber optics, tubing, etc.) occurred during the tests
- Dye solution flow/pressure appears to be influenced by hydraulic conductivity of the formation (as one would predict)
So where do we go from here? Next, we'll be field testing our CPT-tooling version, and we'll be continuing to examine field samples and locate new test sites. Remember that if you have a DNAPL site, we'll examine them free of charge. Just send them in using our shipping instructions. You never know, maybe you'll get some free Dye-LIF characterization data should your site qualify as a good testing location!
Dakota would like to thank the following people and their organizations for their valued assistance on this test project:
- Adam Clements and Vince Malone, Antea Group
- Steve Gross, Hull and Associates
- Andrew Kirsch and Jim Dzubay, Matrix Environmental LLC
- Adrian Fure and Murray Einarson of AMEC (both now with Haley & Aldrich)