 |
|
Hydrocarbon FingerprintingHydrocarbon fingerprinting, sometimes referred to as hydrocarbon
characterization, is a forensic geochemistry technique that is useful
for evaluating releases of petroleum and petroleum products. The
primary source of information for this technique
is high resolution capillary gas chromatography with a flame ionization
detector which gives detailed information not available from routine
EPA methods. This analysis can be done using free product samples,
extracts from soils, and extracts from water samples. Benefits/UsesFour main questions can usually be answered using this technique:
Type of ProductInterpreting the type of product or products present in a sample
using gas chromatography data is primarily a process of visual examination
and comparison with chromatograms of known products. There have
been some advances in using statistical and numerical techniques
to evaluate the data, but these should only be considered tools
to assist in the interpretation and not a means to the final answer.
During the evaluation the interpreter must always be vigilant for
the effects of weathering and mixing of two or more products. Weathering,
discussed in more detail latter, can alter the chemistry of a product,
which in turn effects the appearance of the chromatogram and so
must be taken in to account when interpreting the gas chromatograms. WeatheringThere are three primary types of weathering that may effect the chemistry and thus the character of the gas chromatogram of released products. Evaporation reduces the amount of the most volatile compounds in a product. This is especially noticeable in gasolines or products with compounds in the gasoline range. Biodegradation will selectively remove certain types of hydrocarbons from a product. The bacteria prefer normal paraffins, but when the normal paraffins are gone the bacteria will start to remove other types of hydrocarbons. Water washing (dissolution) preferentially removes the more water soluble components of released products. The oxygenates in gasoline are some of the most soluble compounds followed by the BTEX compounds. Understanding how these three weathering processes effect the chemistry of released products and the chromatograms is essential to accurately interpret the type or types of products released. Age EstimationSeveral techniques have been postulated for estimating the age
or time of release of petroleum products. Of these, the one developed
by Christiansen and Larsen (1993) and further substantiated
by Schmidt, Beckmann and Torkelson (1999) appears to
be the most useful for middle distillates such as diesel fuel.
Christensen and Larsen analyzed a set of samples from sites where
the date of the diesel fuel spill was known. They discovered that
the ratio of nC17 to Pristane had a high correlation with the
age of the spill. From their work a formula (-8.4 x nC17/Pristane
+ 19.8 +/-2) can be used as a way to estimate the age of a spill.
Schmidt, Beckmann and Torkelson expanded the number of known age
spills in the data base and further validated the original work
of Christiansen and Larsen. Comparison to Nearby SamplesComparison of samples collected and analyzed from the same site
can be useful for evaluating the lateral extent of individual
spills (plumes), the commingling of two or more plumes, and likely
direction of the source of a spill. Routine Data PresentationThe results of the gas chromatography analyses are usually e-mailed to clients in the form of an MS-Word document. There will be two pages for each sample. One page has the chromatogram filling the entire page. The second page has four widows and two columns of data. One window shows the entire chromatogram and three smaller windows show different segments of the chromatogram. The two columns of data are the peak heights and peak areas for the peaks identified on the chromatogram. There will also be a scan of the chain of custody form that came with the samples. Semiquantitative ResultsAny of the identified peaks can be reported as percentage of the
entire chromatogram. This is not true quantitative data since there
is no calibration of the instruments. The reported semiquantitative
data is probably plus or minus 10% of a true quantitative analysis,
ie if a true quantitative value is 10% the semiquantitative value
might be reported as low as 9% and as high as 11%. Customized Data PresentationCustomized presentation of the analytical results are available for an additional fee. |
|
|
Copyright © 2007 Torkelson Geochemistry,
Inc. All Rights Reserved |
|
