The current study presents some initial results attempting to quantify the ambient levels of common fugitive emissions on typical Pawnee National Grassland (PNG) Oil and Natural Gas (O&NG) production sites. Volatile Organic Compounds (VOCs) were quantified in real time and used to determine the spacial and temporal windows of exposure for proximate flora and fauna. Our hypotheses were that (i) VOC concentrations would significantly differ among Pump Groups (ii) VOC concentrations would be dependent on wind direction (iii) VOC concentrations would decrease from 8am to 2pm (iv) VOCs are from O&NG and not from other sources, and (v) VOC levels frequently exceed reference standards. Eleven O&NG sites on the PNG in Northern Colorado were randomly selected and grouped according to production (Pump group) along with 13 control sites from three geographical locations. At each site, samples were collected 25 m from the wellhead or stack in NE, SE and W directions. In each direction, two samples were collected with a Gasmet DX4040 gas analyzer every hour from 8am-2pm (6 hours total), July-October, 2016 (N=900).
Results/Conclusions
PERMANOVA results indicated that Pump group was a significant predictor variable (F 26.9, p <0.001 ***), while Direction was not (F 1.3, p .261) although directional trends were observed. VOC concentrations generally increased during the 6 hr day with the exception of N2O and were predominately the result of O&NG production and not vehicle exhaust. Thirteen of 24 VOCs had significantly different concentrations between groups, frequently above reference standards and thus at biologically relevant levels for shortgrass steppe flora and fauna. The most biologically relevant VOCs, found at concentrations exceeding time weighted average permissible exposure limits (TWA PELs), were benzene and acrolein. Akaike information criterion (AIC) was used to measure the relative quality of statistical models predicting a TWA PEL exceedance (binary variable). Generalized Estimating Equations (GEEs) were used to measure the relative quality of statistical models predicting VOC concentration (continuous variable) on sites.