Screening for GROs and DROs using GC/FID (EPA 8015C)

There are many different types of methods for hydrocarbon analysis.   There are EPA methods for air pollution testing and for spills which contaminate waters and soils.  There are ASTM methods for testing petroleum products.  There are also methods for testing hydrocarbons in food packaging and other products.  Out of all of these methods, the one I probably get asked about the most is EPA 8015C.   More specifically, I get asked for suggestions on how to screen samples for GROs (gasoline range organics) and DROs (diesel range organics) quickly, accurately, and using GC/FID instead of GC/MS (especially for highly contaminated samples).

For those that need to follow the method exactly as written, I suggest contacting the EPA with any screening questions.  For those of you who have some flexibility with your screening procedure, I’ll share with you how I would approach this subject.

If you are not familiar with screening water and soil samples for GROs and DROs, it can be confusing.  I remember the first time I had to screen samples, I was surprised by the large number of peaks, most of which eluted on a hydrocarbon hump.  I thought to myself “How am I going to figure-out which petroleum product this is?”, so like many folks, I panicked and gave it the sniff-test (I am not recommending this to anyone).   My sample smelled like diesel, so then I knew what the main contamination was, but then I thought “How can I be certain that other petroleum products aren’t in my sample?”

Diesel Fuel on Rtx-5

 

Before I provide you my suggestions on how I would screen samples, let’s look at a few relevant sections (pasted below) of EPA 8015C as pertaining to GRO and DRO screening.

Section 1.2:  This method may be applicable to the analysis of other analytes, including triethylamine and petroleum hydrocarbons. The petroleum hydrocarbons include gasoline range organics (GRO) and diesel range organics (DRO).

This section also states that you can analyze GROs using three sample introduction techniques; purge & trap, headspace, and direct injection (for waters).  For DROs, two techniques are listed; headspace and solvent extraction (I personally would not use headspace for DRO analysis).

Section 1.2.2:  GRO corresponds to the range of alkanes from C6 to C10 and a boiling point range of approximately 60°C - 170°C. DRO corresponds to the range of alkanes from C10 to C28 and a boiling point range of approximately 170°C - 430°C.

Section 11.7 discusses sample screening and includes the cautionary statement below about using a GC/FID.

CAUTION: The FID is much less sensitive to halogenated compounds than the MS detector. As a result, a simple peak height comparison for such compounds in the GC/MS standard may underestimate the actual concentration of halogenated compounds. When using this method as a screening tool, such an underestimate could lead to GC/MS results over the calibration range or result in contamination of the GC/MS system. Therefore, the analyst should exercise caution when screening samples that also contain halogenated compounds.

 

Let’s say a customer sends you a few soil samples collected from a nearby spill site and your boss asks you to analyze it.   How should you proceed?  First, your customer should tell you something about the expected contamination and the analytical method you are supposed to follow.  If the expected contamination is gasoline and/or diesel, and no method is provided, you may want to consider using EPA 8015C.  If you think the samples may be contaminated with high concentrations of gasoline and/or diesel, you may want to screen your samples first to avoid instrument contamination and downtime.

So what happens if you are not sure which petroleum contamination is in your sample, or its concentration?  I’m sure others may proceed differently, but I would extract some of the sample because I would be worried that if it contained hydrocarbons larger (heavier) than those found in gasoline (>C10), it may contaminate a purge & trap or headspace unit.  However, before I extracted the sample, I would spike it with a carbon marker standard (C6/n-hexane through C28/n-Octacosane) so that I knew which hydrocarbon range was associated with it.   I would follow EPA 8270 for guidance on sample extraction.

Once I had an extract, I would analyze it (1µL liquid injection using a syringe) on an Rtx-1 (or similar) column using GC/FID.  Column dimensions could vary depending upon the type of contamination expected, but a 30m x 0.25mmID x 0.5µmdf should work well for both GROs and DROs.  If you prefer to use a larger bore (internal diameter) Rtx-1 column, then 30m x 0.32mmID x 1.0µmdf or 30m x 0.53mmID x 1.5µmdf should also work well.

 

If you decided that you also wanted to obtain an accurate concentration of gasoline and/or diesel present in your sample, you could use gasoline and/or diesel calibration standards to generate a calibration curve prior to the screening analysis.  However, in order to get an accurate result, you would need to know how much the sample (more specifically, the gasoline or diesel) has been weathered* (see example chromatogram below).  Once you know, you can choose a reference standard which most closely resembles the sample.   Through screening, you should now have a good idea of the contamination and its concentration.

Note:  I suggest analyzing each gasoline and/or diesel reference standard contained in each kit (links above) individually, and saving a printout of each chromatogram in a folder to use as a “fingerprint”.  These chromatograms should help you quickly determine which unweathered or weathered reference standard(s) to choose when generating a calibration curve.  

Because Section 1.2.3 of EPA 8015C mentions that this method may be applicable to other petroleum hydrocarbons, I also recommend generating chromatograms for kerosene and motor oil.   Kerosene has a hydrocarbon range from approximately C10 to C18 and motor oil from approximately C22 to C40.  The capillary columns listed above would be fine for kerosene, but motor oil should be analyzed on an Rtx-1 column like 10121.

* For a chemical reference standard labeled as 50% weathered, a fresh sample was allowed to evaporate to half of its original weight.  The lighter boiling point compounds will evaporate more quickly than the higher boiling point compounds.  Below is an example showing unweathered gasoline verses 99% weathered gasoline.