Eliminating or preventing potential exposures to hydrocarbons in petroleum operations requires efficient and accurate means of measuring airborne concentrations. Typically, petroleum hydrocarbon mixtures are in the carbon range of C2 to C28, with wide variability depending on the source materials encountered at the different stages of the production process, from oil and gas exploration to refining operations, with differences in type of crude oils, natural gas condensates, natural gas liquids, or natural gasoline. The composition of these hydrocarbon mixtures is highly variable depending on factors such as geographical region, geological formation, seasonality, and storage conditions. Because of this variability, assessing occupational exposures to hydrocarbons in petroleum operations is a challenge. The traditional approach for assessing personal exposures in the oil and gas industry was air sampling for benzene, toluene, ethyl benzene, and xylene (the chemicals collectively known as BTEX), along with hydrocarbon  mixtures such as gasoline, kerosene, and diesel (also referred to as light, middle, and heavy distillates). The gasoline, kerosene, and diesel fractions also varied in composition due to specific blending processes and end use. Measurement of the distillate fractions was commonly performed using the concept of “Total Hydrocarbons” (THCs). The resulting THC value has long been used in the IH profession to provide a general assessment of hydrocarbon exposure. An alternative exposure assessment strategy for assessing potential personal exposures to hydrocarbons using the same collection and analysis methods—charcoal sorbent and gas chromatography flame ionization detection (GC-FID)—drops the use of THCs and adds analysis of light, middle, and heavy distillate fractions. TOTAL HYDROCARBONS REVISITED Measuring BTEX vapor concentrations is straightforward, with established methods from both NIOSH and OSHA. However, over the years, the application of the THC result has become misunderstood, in part because “THC” is a misnomer: the THC value does not represent a true “total” mass of hydrocarbons present.  Four critical variables affect the accuracy of applying THC for assessment of hydrocarbon exposures: 

Therefore, a more appropriate definition of the current THC concept is “measurement of potential exposure to any organic compound that can be collected and retained on charcoal media during sampling and storage; extracted (desorbed) using carbon disulfide, the preferred desorption solvent; and measured by GC-FID.” Considering these variables quickly reveals the limitations in the accuracy of the THC result. First, gaseous and highly volatile compounds, like C2 to C4 isomers, are not effectively collected or retained on the charcoal media and so would not be determined. Second, while most aliphatic and aromatic hydrocarbons in the range of C5 to C28 will have a desorption efficiency of greater than 95 percent in carbon disulfide, the actual desorption efficiency of other classes of hydrocarbons (alcohols, ketones, chlorinated hydrocarbons, and other substituted hydrocarbons) that may have been collected on charcoal cannot be specifically known, and a few chemicals known to be present in the petroleum industry have poor recoveries (less than 75 percent desorption efficiency). Third, the laboratory must assume that all of the hydrocarbons collected and desorbed will also pass through the chromatographic column and ultimately be measured by GC-FID. Finally, the choice of calibration material can influence the reported value. The first limitation is mitigated through use of continuous monitors; however, using continuous monitors in combination with sorbent sampling will result in two values that are not additive and would be difficult to evaluate. Without knowledge of the specific chemicals being collected, there is no way to understand the effect of desorption efficiency on the final result or if all chemicals collected passed through the chromatographic column and were measured by FID. Accurate measurement of hydrocarbon compounds by GC-FID is tied to the following analytical variables:

These variables can be controlled within a single laboratory operation, but differences between laboratories affect the final result because there is no methodology protocol published by NIOSH, OSHA, or EPA related to THC measurement. NIOSH has published analytical methods for classes of hydrocarbons (NIOSH methods 1500 and 1501 for general hydrocarbons). OSHA has published analytical methods on specific petroleum hydrocarbons. EPA has environmental methods for gasoline-range organics (GRO) and diesel-range organics (DRO) in water and soil, but neither method addresses the analysis of THC or provides an approach to compounds that would be of interest to the industrial hygienist. Therefore, each laboratory develops its own protocol for the calibration, instrument conditions, and quantitation of THCs.  Analytical laboratories often report THC as “THCs as n-hexane.” This term can be confusing to customers. It actually means that each peak, in the defined range for THCs, was quantitated against a calibration curve made from n-hexane (C6). In the absence of a published method, even the quantitation “as n-hexane” is not prescribed. One can quantitate “as toluene” or other references.  There is no defined exposure limit for THCs. Over the years, the THC results have been compared to occupational exposure limits such as OSHA Permissible Exposure Limits or ACGIH Threshold Limit Values for the reference bulk used in quantitative analysis. However, comparing “THC as n-hexane” to the occupational exposure limit for n-hexane is not appropriate, since THC is not a measure of “n-hexane exposure.” In most cases, this type of comparison could indicate a potential false overexposure to n-hexane since the value reported includes other hydrocarbons present. It is important to note that our historical use of THC was not incorrect or invalid. The concept of THC has been useful in providing a general assessment of hydrocarbon exposure over time for a single blend of hydrocarbons or job classification. Evaluating THC results for a specific hydrocarbon blend and process is efficient and effective, allowing the industrial hygienist to identify changes in hydrocarbon exposure over time, but only for a known and consistent hydrocarbon blend. AN ALTERNATIVE APPROACH Consistent and reliable measurement of occupational exposures to hydrocarbons requires adoption of a different analysis scheme, eliminating the sampling and analysis of THCs, especially in scenarios where the specific hydrocarbon blend is not well characterized. This need was demonstrated in an unusually dramatic situation involving a field IH who was assessing personal exposures to “hydrocarbons” using the traditional THCs approach. The laboratory results, quantitated as n-hexane, were questioned by the field IH because the THC results were incorrectly compared to the diesel exposure limit and exceeded that limit. A review of the sample chromatograms revealed that the distillate was exclusively in the diesel range, making the use of n-hexane as a calibration standard incorrect due to differences in the response factors of n-hexane and diesel distillate. Re-quantitating the results “as diesel” lowered the THC values to a point where there was no exceedance of the occupational standard. While quantitation variability of this magnitude is not typical in THC samples, this example exposes the limitations of the THC concept when the character of the sampling environment is not fully understood.  Because of the variability of hydrocarbon exposure scenarios encountered, this article presents an alternative approach that includes the traditional chemicals found in the petroleum industry—hexane, benzene, toluene, ethyl benzene, and xylene (HBTEX)—and three distillate fractions: light, middle, and heavy. The light, middle, and heavy distillates are defined using both literature review and direct assessment of examples of light distillate (gasoline), middle distillate (kerosene), and heavy distillate (diesel). Although historically BTEX were commonly requested for hydrocarbon exposures, a review of safety data sheets for a cross section of hydrocarbon blends encountered in the industry showed that n-hexane is a common constituent in these blends. The relatively low TLV (50 ppm) for n-hexane warrants its addition to the general BTEX assessment scheme. The resulting exposures would be compared to the TLVs for specific target chemicals (HBTEX): n-hexane, benzene, toluene, ethyl benzene, xylene, and the three distillate fractions defined as hydrocarbon ranges light distillate (gasoline range), middle distillate (kerosene), and heavy distillate (diesel). The distillate ranges are defined by chromatographic retention times for the predominant carbon range for each distillate published in the documentation of TLVs and confirmed through analysis of actual distillate blends. The overall assessment strategy is shown in Table 1.

Use of Surrogates for Quantitative Analysis Currently, assessing potential exposure to specific distillates requires the industrial hygienist to submit a “reference bulk” with each sample set, in an attempt to match the exposure. Providing specific reference samples, in the form of milliliters of the distillate material, is difficult in both collection of the material and shipping. In addition, airborne exposure to a distillate will be skewed in favor of the more volatile components in the distillate when compared to the reference material. This difference between airborne concentration and reference material composition is more pronounced in the light distillate (gasoline) since gasoline comprises a greater amount of more volatile hydrocarbons that tend to become airborne at a greater percentage than the heavier end of gasoline, due to distillate effect.

Eliminating reference bulk and reducing the potential variability of results between laboratories that assess THCs can be addressed through the use of surrogates for quantitation of the different distillate blends. This approach provides a common methodology for assessing distillate blend exposures across the petroleum industry. In developing this methodology, the response factors of C5 through C26 alkanes were compared to the response factors of light, middle, and heavy distillates. Using octane (C8) as a surrogate for light distillate provides a comparable value to specific calibration with a light distillate (gasoline). Using dodecane (C12) as a surrogate for middle distillates and hexadecane (C16) for heavy distillates provides similar agreement when compared to using the actual distillate as a reference. (See Figures 1 and 2 for the chromatographic patterns for the three types of distillate.) Table 2 shows the percent bias of the selected surrogates and the traditional n-hexane when used to quantitate light, middle, and heavy distillate fractions.  Providing a common methodology for assessing these common distillate mixtures across the industry is the overall advantage of using surrogates for quantitative analysis. Eliminating the need to provide a reference sample for the various distillate blends is an added benefit.