About This Project
Car Care Service LNAPL, What UVOST Saw, and What BOS 200® Accomplished – Culminating in NFA
When I think about illustrations or case studies for the newsletter, I prefer those that are uncomplicated. There’s not enough space or reader time for lengthy explanations. Pictures are usually best. I have made an exception here because this site is interesting and teaches us a lot.
Car Care Service (Car Care) was a complicated site with a history of LNAPL in monitoring wells ranging from inches to feet and groundwater samples having BTEX and naphthalene concentrations above their solubilities. The geology was clay, clay slits, and occasional sand stringers. Despite the observable presence of LNAPL, a large, discrete area of LNAPL, sufficient to explain the LNAPL present in the monitoring wells, had yet to be located. It was one of those sites that presented more like an LNAPL minefield – drill a well, and maybe you will get LNAPL. This kind of site is often the result of multiple sources of petroleum, historic product extraction efforts, and an aged release. All elements that describe this site perfectly.
To locate the sources of LNAPL, the services of Dakota Technologies Company, LLC were enlisted. While their membrane interface probe (MIP) was used, the most valuable data came from their Ultraviolet Optical Screening Tool (UVOST®). Randy St Germain of Dakota Technologies, Inc. conducted some fascinating QA/QC studies as part of the effort. I’ll only convey some of what Randy did, but he used a bench-top UVOST tool to characterize both “intact” or fresh gasoline and gasoline mixed with varying amounts of BOS 200® to emulate site conditions. He then compared these bench-top waveforms to the in situ waveforms recorded during UVOST logging both before and after BOS 200 injection at Car Care.
To provide some background, the UVOST system emits brief pulses of 308 nm laser light into the soil. PAHs will absorb this energy and emit it as Stokes-shifted light (fluorescence) ranging from 350 to 500 nm depending on the size and substitution of the specific PAH doing the fluorescing. By looking at the shift in color and duration across these four wavelengths, we can gain insights into the chemical characteristics of the petroleum NAPL doing the fluorescing. Figure 1, for instance, is a typical waveform pattern for fresh gasoline.
Figure 1
is a typical waveform pattern for fresh gasoline. The right side of the waveform is color-coded red because it is more toward the 500 nm range, which is towards red colors. The left side of the waveform is blue because it is away from the red and closer to the blue colors. Larger PAHs fluoresce toward the red, while smaller PAHs fluoresce closer to the blue.*
The total fluorescence observed at all the UVOST locations, which relates directly to the total NAPL present, is represented in a bar chart in Figure 2. Taken before the BOS injection, this UVOST survey demonstrated that typical “intact” gasoline fluorescence (PAHs of all sizes) dominated the site. This is noted on the total fluorescence bars as “96% intact”, meaning 96% of the fluorescence had a waveform matching that Randy recorded for fresh gasoline, as shown in Figure 1. Some samples demonstrated a “blue shift” marked with a blue cap on the bar. The blue shift is likely due to natural carbon or other geological materials that absorb the UV fluorescence. The blue shift is like that caused by BOS 200 when it encounters gasoline due to the activated carbon in BOS 200 selectively adsorbing the larger PAHs.
Figure 2. The bars mark the fluorescence as percent return per foot (%RE-ft) provided by the UVOST as collected at individual survey points. The blue caps indicate loss of fluorescence or return due to absorbance by minor constituents of the geology. 96% of the recorded signal is consistent with fresh gasoline, while minor constituents of the geology absorb 4% of the signal.
Figure 3 shows the results for UVOST logging of these same locations twenty months after BOS 200 injection. 11 of 12 sampling points show a decrease in total fluorescence as a percent return compared to the pre-BOS data. It is important to note that BOS 200 itself has no fluorescence and, in fact, is a fluorescence killer because black carbon absorbs all light. So, the decreased fluorescence was primarily attributable to UVOST “seeing BOS” both as a loss of fluorescing PAHs in the LNAPL and the BOS being physically present in the soil. Therefore, after contact with BOS the total fluorescence of the LNAPL dropped. Still, some smaller two-ring or highly substituted single aromatics that are not as effectively adsorbed into the activated carbon continue to fluoresce mainly in the blue range as compared to the “intact” gasoline. The decrease in total fluorescence before the BOS injection is shown as a percent difference above each location’s bar in Figures 2 and 3. Sitewide, there was a 62% loss in fluorescence. These two figures strongly indicate that many of the fluorescing PAHs were adsorbed to the activated carbon, and/or the soil was rendered non-fluorescent by BOS’s dark presence. Note that the activated carbon also adsorbs non-fluorescent hydrocarbons, which are also in the formation; the UVOST tool does not address these hydrocarbons. Nevertheless, the total fluorescence bar graphs demonstrate that activated carbon is present in the subsurface and has contacted the LNAPL.
Figure 3. The bars mark the fluorescence as percent return per foot (%RE-ft) provided by the UVOST as collected at individual survey points twenty months after BOS 200 injection. The blue caps indicate loss of fluorescence or return due to absorbance by minor constituents of the geology and the activated carbon from BOS 200. 84% of the recorded signal is consistent with fresh gasoline, while 16% of the signal is absorbed by minor constituents of the geology and activated carbon. The total loss of RE-ft is 62%, which is the difference read from the x-axis from Figure 2 and Figure 3.
Let’s look at another figure to further illustrate the percent loss of fluorescence (Signal %RE). Figure 4 is a mirror image comparison of the fluorescence before and after BOS 200 injection. The point we’re looking at is the far left 01 survey point in Figures 2 & 3. The loss of 89% RE post-injection is visually striking!
Figure 4 is a mirror image comparison of the reflectance before (01-LIF on the left) and after (01-PB on the right) BOS 200 injection. There is an 89% loss of RE post-injection.
Another remarkable aspect of the UVOST data is that the smaller weakly fluorescing PAHs that persist in the LNAPL after modest contact with BOS are consistent with expectations based on activated carbon adsorption dynamics. Larger PAHs adsorb more tightly to activated carbon than two-ring PAHs or highly substituted single aromatics. Therefore, when LNAPL saturates activated carbon, one might expect this observed blue shift. See Figure 5.
Figure 5. The images on the left are from UVOST logs acquired before BOS 200 injection, while the images on the right are after BOS 200 injection. The % RE is the area under the curve. It may not be obvious, but in the right waveform, there is more relative area under the blue (UV) fluorescence channel than in the left waveform. Note how comparatively small the red waveform on the right has become as compared to the left. The difference is easier to detect in the two lower images. The black dots mark a continuous recording of the average color for the return fluorescence. The further the dots move to the left, the larger the blue shift. Compare the left image to the one on the right. On the right, the blue arrow indicates where the black dots show a marked shift toward the blue (UV) and away from the red. The left image does not have a similar pattern. This horizontal shift of the black dots, wavelength, is a tell-tale indication of activated carbon adsorbance. This type of shift is not typically seen in soil and aquifer materials. And this pattern is consistent with bench-top testing of BOS/LNAPL and is a tell-tale indication of selective activated carbon adsorption of larger (redder fluorescing) PAHs.
Site data is typically presented as groundwater concentrations, and BTEX in groundwater is a key measure of compliance. So, I’ll focus on BTEX. Keep in mind, however, that the relationship between LNAPL and BTEX concentrations in groundwater is controlled by solubility. Graph 1 presents total xylenes post-injection in groundwater. I didn’t want to show all the contaminant data in one messy spaghetti chart, so I selected xylene. Xylene is a good indicator of LNAPL when looking at groundwater BTEX data. It adheres to soil well, but xylene also has sufficient solubility to reflect the presence of LNAPL when significant LNAPL is present close to the sample collection point. After BOS 200 injection, a steadily decreasing trend was established for xylene.
Graph 2 shows benzene being wrestled to more manageable concentrations even in the presence of LNAPL. The groundwater data were collected from 4 individual monitoring wells located in the heart of the LNAPL area as indicated by LNAPL in the monitoring wells, soil samples, and UVOST. The other BTEX constituents have similar graphs (Data not shown).
Graph 1 presents total xylenes post-injection in groundwater. The groundwater data were collected from 4 individual monitoring wells located in the heart of the LNAPL area as indicated by LNAPL in the monitoring wells, soil samples, and UVOST. The blue vertical band marks the time of the BOS 200 injection. Post BOS 200 injection, a steadily decreasing trend was established for xylene.
Graph 2 is presented due to the prominence of benzene as a regulatory driver. The groundwater data were collected from 4 individual monitoring wells located in the heart of the LNAPL area as indicated by LNAPL in the monitoring wells, soil samples, and UVOST. The vertical blue band marks the time of BOS 200 injection. Benzene is not the best indicator of LNAPL because its high solubility can support significant benzene concentration in groundwater without LNAPL being present. It is difficult, however, to lower benzene concentrations when significant LNAPL is present close to the sample collection point.
While groundwater often determines whether a site falls within compliance, subsurface soil and aquifer materials are more difficult to remediate. These geologic media retain a greater mass of contamination than groundwater. Table 1 presents BTEX data samples collected before BOS 200 in situ injection and 33 months after injection. The data demonstrate that the soil and aquifer materials had lower average BTEX concentrations after BOS 200 in situ injection. The standard deviations on the same data were also lower post-injection, demonstrating that the BTEX and TVPH concentrations vary less between samples post-injection.
Table 1. BTEX plus TVPH data is presented for soil samples collected before and after BOS 200 in situ injection. Both the means and standard deviations of those means are lower after BOS 200 injection. Note that a lower standard deviation indicates that there is less variability between the individual samples after injection. The data was collected 33 months after injection.
What does Car Care teach us? UVOST can “see” BOS 200 in the subsurface as a decrease in fluorescence due to both a loss of fluorescing PAHs in the LNAPL and the BOS being physically present in the soil. Sitewide, there was a 62% loss in fluorescence at 20 months. UVOST also revealed a blue shift in the fluorescence return. This blue shift is consistent with the adsorption character of activated carbon, a significant component of BOS 200. Finally, the blue shift is unique in nature and is a tell-tale signature of activated carbon. The consistency of the appearance of the blue shift indicated that BOS 200 was distributed in the subsurface. When we examine the groundwater data as presented in Graphs 1 & 2, we see that at 20 months, the xylene and benzene concentrations developed a steadily decreasing pattern that persisted. Soil and aquifer materials sampling at 33 months post-injection demonstrates lower contaminant concentrations and less variability in contaminant concentration between samples. In 2023, this site received No Further Action (NFA) from the state of Tennessee. BOS 200 can be successfully employed on petroleum sites having LNAPL.
The success on the Car Care Service site arose not just from the product (BOS 200) or the invaluable insight provided by Dakota Technologies Inc. but also from the skill and attention to detail given to the site by the AST crew who performed the work. As they say, many skilled hands can make short work of a big job. Thank You
* Technically, the laser excitation is 308 nm in the ultraviolet range, and the fluorescence is composed of ultraviolet, violet, indigo, and blue light. So, these color designations are assigned by convention.