Tracking Injections with Laser-Induced Fluorescence

Discussion
There is increasing interest in injectable remedies including in-situ chemical oxidants (ISCO), solvents to enhance mobility, emulsified vegetable oil (EVO), groundwater tracers, and a host of others. Dakota Technologies has been developing methods of tracking these "injectables" with laser-induced fluorescence (LIF). Labeling an injectable remedy with an LIF-responsive dye provides the ability to map exactly where the injected fluid(s) went during and after injection using standard LIF systems such as the Tar-Specific Green Optical Screening Tool (TarGOST®) or Ultra-Violet Optical Screening Tool (UVOST®). For example, during injection pilot studies many practitioners wonder whether the injectate will be dispersed in a homogeneous "radius of influence" around the injection point as was assumed (or advertised). Or did it instead veer off in some yet-undiscovered geologic feature, reaching out like a lightning bolt in a particular direction by following a highly preferential pathway? Questions such as "Why didn't the injection reach those receiver wells that we assumed it would?" are asked quite often. With LIF tracing such questions could be readily (and accurately) answered.

All that's needed to make an injectable "LIF-trackable" is to label it with dye that:

• Responds well to either UVOST or TarGOST
• Has a different spectral and temporal fluorescence response (waveform) than false positives or fluorescent fuels/oils/tars at the site (hopefully VERY different)
• Is stable over space/time
• Doesn't interfere with the active remedy process
• Is affordable
• Is non-toxic (at least relative to the injectable and/or the contaminants)
• Stays with the injectate (isn't oxidized or sorbed to soil)

There are a wide variety of dyes available for use as tracers and "labels" of injectable remedies. Water tracers are extremely water-soluble, making them perfect for tracing flow paths using straightforward injection of dye water into wells or by direct push. Others are soluble only in solvents, fats, or solvents and cleansers - making them suitable for labeling solvent-enhanced recovery agents or emulsified vegetable oils (EVO) for subsequent detection via LIF.

The photo shown below of dye-labeled vegetable oils illustrates this concept. The emulated TarGOST log below the photo illustrates TarGOST's semi-quantitative performance for one of those dyes used to label EVO. The log's callouts are (top-to-bottom): 0% EVO (pore water concentration), 0.055%, 0.55%, and 5.55%, and 55% EVO in pore water. To illustrate how different dye waveforms can be, a variety of waveforms generated by a few other dyes are shown below the TarGOST log.

Pilot Studies
Dakota has recently worked with a number of clients to pilot-test three variations of dye-tracking of either remedies (injected to affect change), tracers (injected to understand flow patterns) or both. Here are some short descriptions of two interesting applications of LIF toward observing the distribution of fluids at injection pilot projects.

Water Tracing
The client had previously injected some tracer dye into an injection well, but the dye failed to "show up" in the receiver wells as predicted according to the potentiometric groundwater surface. Dakota worked with the client to evaluate several potential dyes and to develop a limit-of-detection (LOD) estimate by doing some benchtop work at Dakota. It was decided to use the TarGOST for this application since it had the best performance for the red-colored (Rhodamine) water tracer (RWT) dye the client decided to use. The client also wondered if simultaneous injection of molasses would be problematic at all, so Dakota ran experiments by adding molasses to the RWT solution. It was determined that molasses fluoresced with TarGOST as well, so it didn't hurt and actually helped the fluorescence intensity increase to some degree.

An injection of 2% molasses and 200 ppm dye was made a month prior to TarGOST being brought on site. The TarGOST log below is one example of the six logs acquired during a day of TarGOST logging. This log was done 8 feet west and 25 feet north of the injection well and it shows clearly that the injected molasses and dye were moving in preferred horizons. The green Signal and the orange Fluor (fluorescence-only) responses indicate the higher porosity horizons where the RWT/molasses solution was residing.

Dakota also applied a method that involves non-negative least squares (NNLS) fitting of the logs' waveforms (example above) to "clean up" and categorize the fluorescence into RWT and/or two forms of molasses fluorescence (perhaps due to oxygen concentration differences).

The client learned a great deal about the behavior of injections at their site and, based on pore water and soil samples from the conducting zones identified with TarGOST, the client later determined an effective field detection limit of 10-20 ppb for RWT (pore water concentration) in dark fine soils and estimated a 1-2 ppb limit for clean sandy soils.

EVO Injection
Dakota worked with a client in California to develop a dye-labeling approach to make their emulsified vegetable oil (EVO) pilot test trackable with LIF so as to allow direct assessment of the radius of influence following the direct push injection of EVO. In lab tests the TarGOST system proved to provide the highest signal-to-noise ratio for the oleophilic (oil-loving) dyes Dakota had available (relative to UVOST). Based on a series of calibration tests that were conducted prior to the pilot test using the vendor's EVO and soil from the site, a limit of detection of >= 0.05% (dye-labeled EVO in pore water) was determined for sandy soils. Dakota and the client decided to further improve TarGOST's ability to detect the injected EVO solution by dye-labeling the dilution water with a water-soluble dye (50 ppm) as well – basically doubling up with two types of dye.

This log was taken approximately three feet from the injection boring between 0 to 48 hours after injection (injection took two days to complete and TarGOST logging started immediately after). The intended depth of injection ranged between 31 and 55 ft. The LIF showed convincingly that, rather than the often-hoped-for homogeneous radius of influence, the EVO solution penetrated only discrete and limited soil horizons (assumed to be preferred pathways due to higher porosity). The distinct horizons at 32 ft and 42 ft, along with an unexpected horizon at 7 ft (unintended injection that perhaps followed annulus of the rod or a rod joint leaked) all yielded waveforms that matched the unique waveforms of EVO and/or water tracer of the injected solution, assuring us that the LIF response was due to the presence of injected material. The client is now using this somewhat surprising information to carefully consider the path going forward for full-scale remediation.