Formatted Title:

Recently Discovered Trait in Coal Tar and Creosote Fluorescence Leads to Greatly Reduced Reliance on Validation Sampling

Background/Objectives:

NAPLs such as coal tar, creosote, bunker, and dripolene act as the primary source of contamination at former manufactured gas plants, wood treaters, and petrochemical sites. The Tar-specific Green Optical Screening Tool (TarGOST^®) was specifically designed to log these molecularly “heavy” multicomponent DNAPLs by taking advantage of their fluorescent property and it has been deployed successfully at hundreds of sites since 2004. Unfortunately, fluorescence emitted by biogenic soils and organic detritus remains a significant interferent at DNAPL sites because these “false positives” can fluoresce just as brightly, if not brighter, than the target DNAPL. It is imperative that investigators assume their fluorescence logs contain false responses and work to filter them out - lest they be misinterpreted as DNAPL - resulting in a grossly inaccurate and inflated DNAPL conceptual site model (CSM).

Approach/Activities:

There are usually just enough differences in the fluorescence of the two that they can be separated by color and/or average lifetime. Over the last 20 years a systematic method of targeted sampling/validation has evolved to discover and document these differences and use them to remove the highly problematic non-target fluorescence. Unfortunately, despite the best efforts of experienced practitioners, a consequential number of sites involve DNAPLs and false positives whose fluorescence is nearly identical or so highly variable (due to multiple sources or weathering) that the false positives cannot be satisfactorily rejected. This results in a loss of confidence that the fluorescence logs faithfully represent the DNAPL alone, and justifiably so. This forces investigators to validate an excessive number of locations or even revert to a traditional sampling characterization which greatly increases costs, slows progress, taxes the investigators, increases exposure to personnel, and results in low resolution CSMs.

Recently, a question posed by a consultant determined to decisively identify DNAPL in a TarGOST log at its heavy DNAPL site prompted a serendipitous discovery; we observed that the DNAPL’s fluorescence had a subtly temporal property that none of the fluorescing soils did – a property that the standard advanced data filtering method wasn’t leveraging effectively. This unique fluorescence property was later found to apply to the entire site - and later to all heavy DNAPL sites - regardless of DNAPL type. Twenty years after the introduction of TarGOST this newly recognized trait appears to “crack the code” in the sense that it broadly applies to the entire heavy DNAPL and false positive families, meaning it’s effective even at sites involving significant variation in heavy DNAPL chemistries and or multiple false positive sources.

Results/Lessons Learned:

A new fluorescence filtering algorithm has been implemented that allows for discernment of DNAPL from false positives, in real time, in the field. It appears to greatly reduce the need for validation at many sites, lowers costs, allows for a more adaptive characterization and generally improves confidence in the DNAPL CSM. We will review how the new filter process was developed, how it is applied, and how it has been validated. We will include a case study where the creosote NAPL fluorescence at a site with previously intractable false positive responses was isolated decisively. A disturbingly high fraction (40%) of the significant fluorescence sitewide was found to be due to brightly fluorescing biogenics, not creosote.

David Bordelon

Affiliation

Refine Results

(Group 2, Poster Board #46) Recently Discovered Trait in Coal Tar and Creosote Fluorescence Leads to Greatly Reduced Reliance on Validation Sampling