Start Submission

Reading: Distributions of air pollutants associated with oil and natural gas development measured in ...


A- A+
Alt. Display

Research Article

Distributions of air pollutants associated with oil and natural gas development measured in the Upper Green River Basin of Wyoming


R.A. Field ,

Department of Atmospheric Science, University of Wyoming, Laramie, Wyoming, United States, US
X close

J.J. Soltis,

Department of Atmospheric Science, University of Wyoming, Laramie, Wyoming, United States, US
X close

P. Pérez-Ballesta,

European Commission, Joint Research Center, Institute for Environment and Sustainability, Air and Climate Unit, Ispra, Italy, IT
X close

E. Grandesso,

European Commission, Joint Research Center, Institute for Environment and Sustainability, Air and Climate Unit, Ispra, Italy, IT
X close

D.C. Montague

Department of Atmospheric Science, University of Wyoming, Laramie, Wyoming, United States, US
X close


Diffusive sampler monitoring techniques were employed during wintertime studies from 2009 to 2012 to assess the spatial distribution of air pollutants associated with the Pinedale Anticline and Jonah Field oil and natural gas (O&NG) developments in the Upper Green River Basin, Wyoming. Diffusive sampling identified both the extent of wintertime ozone (O3) episodes and the distributions of oxides of nitrogen (NOx), and a suite of 13 C5+ volatile organic compounds (VOC), including BTEX (benzene, toluene, ethylbenzene and xylene isomers), allowing the influence of different O&NG emission sources to be determined. Concentration isopleth mapping of both diffusive sampler and continuous O3 measurements show the importance of localized production and advective transport. As for O3, BTEX and NOx mixing ratios within O&NG development areas were elevated compared to background levels, with localized hotspots also evident. One BTEX hotspot was related to an area with intensive production activities, while a second was located in an area influenced by emissions from a water treatment and recycling facility. Contrastingly, NOx hotspots were at major road intersections with relatively high traffic flows, indicating influence from vehicular emissions. Comparisons of observed selected VOC species ratios at a roadside site in the town of Pinedale with those measured in O&NG development areas show that traffic emissions contribute minimally to VOCs in these latter areas. The spatial distributions of pollutant concentrations identified by diffusive sampling techniques have potential utility for validation of emission inventories that are combined with air quality modeling.
Knowledge Domain: Atmospheric Science
How to Cite: Field, R.A., Soltis, J.J., Pérez-Ballesta, P., Grandesso, E. and Montague, D.C., 2015. Distributions of air pollutants associated with oil and natural gas development measured in the Upper Green River Basin of Wyoming. Elem Sci Anth, 3, p.000074. DOI:
 Published on 21 Oct 2015
 Accepted on 23 Sep 2015            Submitted on 10 Apr 2015
Domain Editor-in-Chief: Detlev Helmig; Institute of Alpine and Arctic Research, University of Colorado Boulder, Boulder, Colorado, United States
Guest Editor: Gabrielle Petron; CIRES, University of Colorado Boulder and NOAA ESRL, Boulder, United States

1. Introduction

Increased onshore oil and natural gas (O&NG) production, driven in part by the development of hydraulic fracturing and directional drilling techniques, continues to reshape the United States (US) energy economy. Natural gas, in particular, has become increasingly important, a trend expected to be sustained in the future, such that by 2040, its share of total primary energy production is predicted to rise to a high of 30%. Furthermore, it is expected to become the most important fuel source for electricity generation by 2040 (US EIA, 2014).

Wyoming has a long history of O&NG development with oil production dating back to the late 19th Century. Currently, the Jonah field and Pinedale Anticline, both in the sparsely populated Upper Green River Basin (UGRB) of Sublette County, Wyoming, are ranked 8th and 6th respectively among US producing gas fields for 2013 (US EIA, 2015). In 2012, ∼4,000 wells from these two fields produced 807.5 million Mcf of natural gas, 6.3 million barrels of oil and 23.9 million barrels of water (Wyoming Oil and Gas Conservation Commission, 2015). The relatively wet nature of the gas allows for production of oil (condensate) that leads to both fields also being ranked within the top hundred of US producing oil fields, with the Jonah field 94th and the Pinedale Anticline 85th, respectively (US EIA, 2015).

Accompanying O&NG development has been the inevitable addition of new atmospheric pollutant emission sources. The large number and diversity of emission sources have accentuated the difficulty of developing quantitatively accurate emission inventories for O&NG field operations. One approach to combating these difficulties is the development of a new inventory for small point (also termed ‘area’), upstream O&NG sources. Discrepancies have emerged between measurement based emissions estimates over, for example, an entire local gas field and calculated inventory source strengths. The latter appear to be under-predicted (Brandt et al., 2014). Efforts to reconcile these differences are already underway (Pétron et al., 2014).

Pollutants from O&NG development and production operations contribute to the photochemical formation of surface ozone (O3). High levels of wintertime O3 were first observed in 2005, close to the developing Jonah and Pinedale Anticline O&NG fields. Since that time a large number of winter O3 episodes have been reported in the UGRB (during 2005, 2008, 2011) and the Uintah Basin of Utah (during 2011, 2013 and 2014). Schnell et al. (2009) summarized the early observations and described the requisite conditions necessary for episode occurrence. These include relatively stagnant conditions and snow cover to both increase the actinic flux and foster the formation of a low-level capping inversion above a ‘cold pool’ that facilitates the accumulation of primary O3 precursor pollutants, specifically volatile organic compounds (VOC) and oxides of nitrogen (NOx) (Carter and Seinfeld, 2011; Oltmans et al., 2014). Under favorable conditions, O3 production rates in the UGRB can exceed 10 ppbv/hr, leading to maximum hourly mixing ratios in excess of 160 ppbv (Rappenglück et al., 2014). The National Ambient Air Quality Standard for O3 is based upon the average of the fourth highest 8-hour average for 3 years of monitoring. An area is defined as being in non-attainment when this 3-year average exceeds 75 ppbv. Fourth highest 8-hour averages have exceeded 75 ppbv in 2005, 2008 and 2011 in the UGRB (Field et al., 2015).

Ozone episodes have typically been associated with urban areas in summer with precursor species from traffic and industrial emissions. The Los Angeles (LA) basin provides an archetypical scenario for these events. In recent years, air quality in the LA basin and many other metropolitan areas has improved, largely because of the implementation of stringent emission controls for traffic produced VOCs and NOx (Pollack et al., 2013; Warneke et al., 2012). The dominance of traffic related pollutant sources has eased the characterization and design of effective pollution control strategies. Emission profiles from O&NG activities are varied and also depend on the basin considered due to differences between characteristics of the extracted hydrocarbon product and the operational approaches. Alkanes are typically the most important VOC class arising from fugitive emission sources, whereas aromatics are emitted in particular during wet gas dehydration operations and water treatment processing (Field et al., 2015). Major VOC sources also include those associated with well completions, venting, processing, and distribution activities, in addition to a wide range of combustion engines (Field et al., 2014). Combustion sources are also the primary source of NOx.

Characterization of the relationship between precursor mixing ratios and the potential for wintertime O3 production under optimal formation conditions has progressed significantly. The importance of the spatial distribution of emitted VOCs and their concentrations has been recognized. Both modeling studies (Carter and Seinfeld, 2011; Edwards et al., 2014) and observations (Field et al., 2015) suggest that O3 production is radical limited and therefore sensitive to VOCs. Thus VOC reductions should lower peak O3 whereas NOx reductions could lead to increases in O3, until a critical lower threshold of NOx is reached. Under VOC sensitive/NOx saturated conditions, modest reductions in NOx can lead to an increase in the rate of O3 production by decreasing the rate of diversion of HOx radicals to inactive HNO3, thereby causing the concentration of NO-oxidizing peroxy radical production from VOCs to rise. Larger decreases in NOx can result in the system becoming NOx rather than VOC sensitive, leading to a reduction in the rate of NO2 photodecomposition, and a consequent lowering of the O3 formation rate (Sillman, 1999). Available measurement data (Field et al., 2015) suggest that O3 precursor concentrations vary across the UGRB, implying that uniformity of O3 formation is highly unlikely. For example, daily average ambient toluene mixing ratios during winter at monitoring locations in the UGRB, described elsewhere (Field et al., 2015), typically vary over two orders of magnitude within a range of 0.0 to ∼50 ppbv. At additional sites considered as proximate to sources, including a fence line location next to a water treatment and recycling facility, toluene values reached 1 ppmv. O3 mixing ratios measured in the UGRB have recently been shown to be strongly influenced by both toluene and m+p-xylene (Field et al., 2015). Modeling studies attempting to compute spatially resolved time dependent wintertime O3 concentrations are clearly limited by the resolution of precursor concentration data (Ahmadov et al., 2015), and are potentially compromised by the difficulties of successfully simulating the topographically influenced weak wind fields in the very shallow surface cold pool. The need for spatially distributed measurements of VOC and NOx with reference to current regulatory modeling approaches, was also noted by Carter and Seinfeld (2011).

Early State of Wyoming Department of Environmental Quality (WDEQ) modeling studies aimed at understanding UGRB wintertime O3 episodes were based on measurements from two locations, namely Jonah and Boulder. They were subsequently supplemented by four additional air quality network stations outside the O&NG development areas, together with a suite of more extensive measurements at the Boulder site (WDEQ, 2015a). Unfortunately, these network upgrades did not provide enough data to characterize the spatial distribution of pollutants across the basin (Carter and Seinfeld, 2011). The distribution of VOC and NOx in wintertime varies due to the non-uniform spatial distributions of their emission sources, and is further influenced by meteorological conditions that affect local advection and pollutant redistribution in the surface cold pool below the capping temperature inversion.

This study aims to characterize the spatial extent of elevated O3 during days anticipated to experience episodic ozone production. A similar characterization during wintertime conditions in the UGRB was performed for O3 precursors, VOC and NOx, at locations in O&NG development areas with primary pollutant emission sources. Local population centers were included as both contrast, due to emissions, and due to concerns related to pollution impacts. Achieving these aims greatly facilitates the identification and strength of major pollutant emission sources, understanding cold pool meteorology in the basin, and identifying communities potentially at risk for experiencing degraded air quality. The study builds on surveys that used canister sampling for VOC measurements (Field et al., 2015), by employing a passive diffusive sampler network to measure the spatial distributions of O3, NOx, and VOC (including benzene, toluene, ethylbenzene and xylene isomers (BTEX)), over a wider area of the UGRB, and with a greater number of sampling sites. The resultant observational dataset could be used to assess the quality and accuracy of the WDEQ spatially resolved emissions inventory, which has yet to be made publicly available.

2. Materials and methods

Measurement surveys of the spatial distributions of species of interest were conducted using passive diffusive samplers. This technique was first developed in 1973 (Palmes and Gunnison, 1973), and has been subsequently refined and is now well established (e.g., Varns et al., 2001; Carmichael et al., 2003; Muckerjee et al., 2004; Bruno et al., 2005; Gerboles et al., 2006; Pérez Ballesta et al., 2008; Cocheo et al., 2008; Buczynska et al., 2009; Martin et al, 2014). The method, which relies on the diffusion of a pollutant species to a collection medium, typically an adsorbent material, allows a wide range of trace gas concentrations to be measured in both indoor and outdoor environments (Zabiegala et al., 2010). Brown (2000) has described the underlying theory of diffusive sampling in detail, while Carmichael et al. (2003), among others, have discussed the prerequisites for its successful operation, which include an appropriate consideration of sampler design, sampling rate, sampler orientation during sample collection, exposure conditions, and analytical procedures.

Diffusive sampling yields averaged species concentrations for the chosen sampling time period, which may range from a few hours, appropriate for O3 episode monitoring, to a week or more, an often useful time period for VOC measurements in this context. A particular advantage of the method, especially in remote areas, is the flexible ease of sampler placement, because no electricity is required for sampler operation. In Europe diffusive sampling has qualified as an indicative method for European Air Quality Directives, by achieving defined accuracy requirements (Buzica et al., 2005). For example, the Radiello® diffusive sampler meets the quality requirements according to the European standard for measuring benzene in ambient air (Pennequin-Cardinal et al., 2005). Diffusive sampling is regularly employed in Europe and frequent inter-comparison studies are performed to ensure the quality of diffusive samplers and their application as a technique for assessing ambient air quality (Martin et al, 2014).

2.1 Sampling approach

In this study, sampling surveys were conducted in the UGRB of Sublette County, Wyoming, using species specific diffusive passive samplers deployed at over 60 sites. Table 1 summarizes the sampler types deployed, the analysis laboratories and analytical methods employed, sampling duration periods, and survey dates. Table 2 shows a summary of key quality assurance parameters for the different diffusive sampling and analysis methodologies. Ozone surveys were performed to determine both the spatial extent of episodes, and whether observed high O3 concentrations resulted from long-distance regional transport or local production driven by precursor emissions related to UGRB O&NG development. BTEX and NOx surveys were performed to show the distribution of these O3 precursors. A single survey of VOC was carried out at 13 sites to determine the variability of these pollutants at selected established monitoring sites within the basin.


Table 1.

Diffusive sampler methods

Species Sampler Analysisa Analysis group Duration Survey dates (m/yy)
O3 Ogawa IC RTI Int. 8-hours 2/09; 6/09
NOx Ogawa IC RTI Int. 3-7 days 2/11; 11/11; 2/12
BTEXb Radiello® TD/GC/MS Air Toxics Inc. 3-7 days 2/11; 11/11; 2/12
VOCc PODs TD/GC/FID ERLAP 3-days 2/12

Table 2.

Overview of methodology quality assurance

Species Analysis group Accuracy Detection limit
O3 RTI Int. 15% 3 ppbv
NOx RTI Int. 20% 0.3 ppbv
BTEXa Air Toxics Inc. 30% 0.1 ppbv
VOCb ERLAP 20% 0.1 ppbv

Potential monitoring locations were initially determined by dividing the study area into a 10 x 10 grid and evaluating each of the 100 grid boxes in the field for potentially suitable sites. The primary criteria used for site subsequent selection were accessibility, representativeness, and safety. Site locations, Pinedale Anticline and Jonah field wells, and select geographical markers are shown in Figure 1. The UGRB, at an elevation of approximately 2150 m msl, is enclosed by the Wyoming mountain range to the west, the Gros Ventre range to the north, and the Wind River mountains to the northeast. To the south it broadens and opens onto lower elevation plains and the Interstate-80 corridor. Monitoring sites were located towards the center of the basin in and around O&NG developments. As funding was from the Pinedale Anticline Project Office the designated oversight agency (WDEQ) did not support the placement of many samplers in and around the Jonah field. The greater density of passive samplers in the Pinedale Anticline compared to the Jonah field is evident in Figure 2, as is the wider distribution of continuous O3 monitors, which extend further to the south into the Jonah field and beyond.

doi: 10.12952/journal.elementa.000074.f001.
Figure 1.  

Study area in the Upper Green River Basin of Wyoming with diffusive sampling sites and well pad locations.

Natural gas wells active in 2011 in the Pinedale Anticline (grey) and the Jonah Field (black) developments are shown together with locations of diffusive sampling sites (blue and red as labeled).

doi: 10.12952/journal.elementa.000074.f002.
Figure 2.  

Ozone mixing ratio (ppbv) isopleth contour plots for February 22, 2009.

Upper panel, Figure 2a: Contour plot constructed from 8 hour average O3 passive diffusion sampler data. Locations of sampling sites are indicated by pink triangles. Lower panel, Figure 2b: Contour plot is constructed from 8 hour averaged data from continuous monitors located at the sites marked by pink triangles. The dashed outline rectangle shows the area depicted in the upper panel. Wells are superimposed to aid as geographic markers.

A key experimental requirement was that samplers should be deployed and retrieved at similar times, ideally simultaneously, especially for short duration samples. To achieve this objective, short deployment routes were designed, each with 3 to 5 sampling sites, thereby ensuring that the monitoring periods for each sampler were within ±15 minutes of the target start and duration times. Up to fourteen teams of trained volunteers, representing employees of the State of Wyoming, the Federal Government, Sublette County, the City of Pinedale, the University of Wyoming, and the natural gas industry, together with local citizens, carried out coordinated sampler deployment on days when meteorological forecasts predicted conditions expected to allow O3 production and/or pollutant accumulation. Consistency of handling was ensured through in-person training with respect to standardized operating procedures. Each team was supplied with a clear route map and easy to follow sampler-handling instructions. This approach ensured that only 1 sample out of a total of ∼400 was lost due to handling issues.

Samplers were mounted on isolated custom-installed or existing posts at ∼1.5 m above ground level. After exposure, they were prepared for analysis immediately upon receipt from the placement teams. Samplers were refrigerated overnight and were express shipped in an insulated container to the analysis laboratory (an unnecessary procedure for sorbent tubes). The week-long extended sampling period of the first O3 precursor survey yielded samplers with relatively high mass loadings of adsorbed material, such that while they were within the sampler saturation limit for BTEX, they required split flow during analysis. A reduced sampling duration was therefore adopted during the second and third surveys, to eliminate the need for split flow analysis. Prototype Pocket Diffusive (POD) samplers developed by the European Reference Laboratory of Air Pollution (ERLAP) of the European Commission (Pérez Ballesta et al., 2014) were used for VOC.

2.2 Analytical approach

Samples were analyzed following protocols recommended by the requisite sampler manufacturer. The analytical approach differed based upon the compound considered and the sample collection approach.

Ogawa samplers were used for determination of ambient mixing ratios for O3 and NOx. This sampler, developed by Koutrakis et al. (1993), has a tubular design with two chambers at each end that each hold a collection pad set within stainless steel shields behind an end cap. Each end cap has a series of holes engineered to ensure an appropriate airflow rate, and thereby diffusion, to a cellulose fiber analyte collection pad. The collection pads for O3 measurements are coated with a nitrite based solution that is oxidized to nitrate during sampling. After sampling the pad is extracted and analyzed with ion chromatography Dionex 2000i series. The nitrate ion concentration is used to determine O3. This determination relies upon calculations that consider the diffusion path (length and area of diffusion) and diffusion coefficient (determined by the collection temperature and pressure). The measurement of NOx with the Ogawa sampler follows the same principles as those for O3 measurement, albeit with different collection pads, analytical procedures and diffusion co-efficient. For determination of NOx in ambient air, Ogawa sampling pads are coated with triethanolamine. After exposure the collection pads were stored in vials prior to shipping for analysis. They were subsequently extracted and the NOx content determined by the spectrophotometric Saltzman reaction using a TCI-NOx 1000 automatic flow injection analyzer (Tokyo Kasei Kogyo). RTI International performed the analyses of Ogawa samplers and supplied us with spreadsheets of the mass loading results for each sampled adsorbent. Corresponding raw O3 and NOx mixing ratios were then calculated using the appropriate concentration conversion coefficients incorporating ambient temperature, relative humidity, and the water vapor pressure coefficient, as described by Ogawa (2006). Ambient O3 and NOX mixing ratios were derived with an uptake rate suitable for the sampling conditions with blank correction applied from five travel blanks for each survey.

Radiello® samplers (Fondazione Salvatore Maugeri (FSM)) were used to determine the ambient mixing ratios of BTEX. This sampler, developed by Cocheo et al. (1996), consists of a cylinder set within a micro-porous polyethylene membrane, held by a polycarbonate support plate. The sample is collected by adsorption upon 350 ± 10 mg of graphitized carbon (Carbograph 4) sorbent material. The design of the stainless steel net cylinder containing the sorbent promotes radial diffusion to the sorbent surface. After sampling the exposed sorbent cylinders were individually stored in glass vials prior to shipping. Collected analyte is recovered by thermal desorption, with analyses performed by capillary gas chromatography with MS detection. Sampling times for typical ambient environments may be up to two weeks. Given the possibility of high ambient levels in the UGRB, sampling times were reduced to 3-days to minimize the possibility of back diffusion, while allowing sufficient time for collection of a suitable quantity of the target compounds. Air Toxics Inc. performed the analyses of Radiello® samplers and supplied spreadsheets with a mass loading for each sampler. Corresponding mixing ratios were subsequently calculated using appropriate concentration conversion coefficients incorporating ambient temperature, relative humidity, and the water vapor pressure coefficient as described by FSM (2006). Ambient mixing ratios were derived with uptake rates suitable for the sampling conditions with blank correction applied from five travel blanks for each survey.

POD samplers were used for the determination of ambient mixing ratios of VOC. This was the first field deployment of POD samplers, so the results presented here are regarded as a pilot project. This sampler, developed by Pérez Ballesta (2014), is a zero equilibrium interface concentration device (Mayer et al., 2003). The POD sampler consists of a metallic diffusion body with a porosity of 20 µm and an adsorbent cartridge located inside the diffusion body. The adsorbent cartridge contains 34±1 mg of Carbopack-X 60–80 mesh (Supelco). The diffusion body is held on a support body for sampling and is hermetically closed with a PTFE cover for transport and storage. After sampling sealed samplers are shipped in sets of three within glass tubes. Collected analyte is recovered by thermal desorption, with analysis performed by capillary gas chromatography with MS detection. Sampling rates were experimentally evaluated in laboratory tests. For VOCs the effect of concentration (0.1 to 50 µg/m3 depending on the compound), temperature (-30 to 40°C), humidity (10 to 70%) and air speed (0.1 to 4 m/s) was evaluated. The sampling rate was not significantly influenced by concentration (variation <10%) humidity (<15%) and air speed (< 6%). Sampling rate was lower for compounds with the higher diffusion coefficients. Laboratory tests allowed the estimation of sampling rates for VOCs that were not evaluated, as a function of their diffusion coefficient. Ambient VOC mixing ratios were derived for thirteen compounds, with uptake rates suitable for the sampling conditions with blank correction applied from five travel blanks and five laboratory blanks.

Reported values for each sampler were calculated by applying standard uptake rates and blank corrections (based on travel blanks) to the mass loadings. Average UGRB meteorological conditions reported at WDEQ monitoring sites were used to calculate species mixing ratios. Data quality comparisons were conducted by co-located measurements, with identical sampling times, using independent methodologies. Table 3 shows linear regression results for O3 and NOx from Ogawa diffusive samplers compared to averaged hourly values from WDEQ stations, and also for BTEX species from Radiello® diffusive samplers with those from UW canister measurements (Field et al., 2015). The results confirm that Ogawa and Radiello® passive samplers perform well, indicating that they can accurately assay the targeted pollutants. For VOC, excellent agreement was also found for thirteen C5 to C8 species mixing ratios measured by POD diffusive samplers that were measured with co-located canister sampling at thirteen sites performed during the study of Field et al. (2015) (CPOD sampler = 1.06 Ccanister, R2 = 0.99).


Table 3.

Comparison of co-located measurements of pollutant concentrations (C) using diffusive sampling (ds) with established methodologies (em)a

Species Sampler Linear regression parametersb Survey dates (m/yy)
O3 Ogawa Cds= 0.93 Cem, b = -1.1; R2= 0.98; n = 8 2/09; 6/09
NOx Ogawa Cds= 1.02 Cem, b = 0.6; R2= 0.85; n = 18 2/11; 11/11; 2/12
BTEX Radiello® Cds = 0.93 Cem, b = 0.0; R2= 0.99; n = 120 2/11; 11/11; 2/12
VOC PODs Cds= 1.06 Cem, b = 0.0; R2= 0.99; n = 169 2/12

3. Results and discussion

Summary results for each survey for O3, NOx and benzene are given in Table 4. The winter survey yielded higher O3 mixing ratios than that in summer. NOx mixing ratios are similar in the three winter surveys with measured values ranging from 0.1 ppbv to 16 ppbv. For BTEX surveys, benzene mixing ratios range from 0.1 to 10 ppbv. By contrast benzene peaks at over 50 ppbv during the VOC survey that was performed concurrently with VOC canister sampling (Field et al., 2015). Diffusive VOC sampling sites were selected from those used for BTEX surveys with the exception of one site which was positioned at the fence line of the water treatment facility discussed in sections 3.2 and 3.3. While Table 4 show elevated pollutant levels, spatial distributions of pollutant species can be visualized by constructing concentration contour plots of their ambient concentrations (or mixing ratios). Mixing ratio zones on such plots are defined with greater confidence in areas where the density of monitoring sites is high. Contour plots shown here were created with the Surfer® software plotting routine available from Golden Software LLC, Golden, CO. The plotting routine uses point kriging and a default linear variogram (Isaaks and Srivastava, 1989). Such plots are critically dependent upon the relationship between pollutant emissions and meteorological conditions, in particular wind speed and direction. Wind roses for all surveys are given in the supplemental materials Figure S1.


Table 4.

Selected summary statistics for the diffusive sampling surveys for ozone, oxides of nitrogen and benzene

Species Sampler No. Min (ppbv) Mean (ppbv) Max (ppbv) Start date (m/dd/yyyy)
O3 Ogawa 50 32 48 60 2/22/2009a
O3 Ogawa 50 28 35 43 6/3/2009a
NOx Ogawa 50 2 7 15 2/22/2011b
NOx Ogawa 42 0 4 16 11/15/2011c
NOx Ogawa 48 1 7 16 2/7/2012c
benzene Radiello® 49 0.2 1.3 4.9 2/22/2011b
benzene Radiello® 49 0.2 1.1 9.6 11/15/2011c
benzene Radiello® 47 0.1 1.1 5.9 2/7/2012c
benzene PODs 13 1.4 9.3 55.2 2/14/2012

3.1 Ozone

Eight-hour average O3 mixing ratio data obtained for UGRB diffusive sampling surveys are plotted for February 22nd 2009 (9:00 to 17:00) in the contour map shown in Figure 2a. This plot indicates that on this day O3 levels present throughout the Pinedale Anticline and Jonah field were elevated above the background levels observed to the west. Higher O3 values likely extended into nearby areas to the northwest. During the day winds were variable with speeds less than 10 m/s.

An analogous O3 contour plot extending further south, can also be constructed for the same time period (February 22, 2009) using data from continuous monitors. A special study conducted by WDEQ during this period increased the total number of O3 monitors in the Pinedale Anticline and Jonah field to 21 (Environ, 2010). The data they provide allow the O3 contour plot shown in Figure 2b to be constructed from hourly averaged data for the same time period as Figure 2a. The distribution patterns of O3 mixing ratio in Figures 2a and 2b are similar. Furthermore, in areas where passive samplers and continuous monitors are located in close proximity, or where the spatial proximity of measurements from both techniques is high, observed O3 mixing ratios are in excellent mutual agreement. The largest differences occur in areas where passive samplers and continuous monitors are not co-located and coverage is sparse. Contour plot patterns for these areas, especially in Fig. 2b, should therefore be interpreted with caution. The greater geographic coverage provided by the continuously operating monitors allows the approximate southern boundary of elevated O3 levels to be roughly delineated, as shown in Figure 2b. The plot confirms that on this day, high O3 mixing ratios are essentially confined to the gas development fields and their immediate surroundings, rather than being distributed throughout the entire region.

The four panels of Figure 3 show the development of O3 distributions constructed from the continuous monitor data at 06:00, 12:00, 17:00, and 22:00 MST on March 2, 2009. This day experienced some of the highest O3 values during the winter of 2009. Animations of hourly O3 mixing ratio distributions illustrate the development pattern more clearly. Higher O3 mixing ratios, centered in particular on the Pinedale Anticline, are observed in the afternoon. Other days (not shown) exhibit similar temporal behavior, but also show slightly different O3 spatial variability, that presumably results from small differences in the prevailing wind fields. The weak episode on February 22, 2009 (Figure 2) was centered upon the Jonah field, when winds throughout the day were generally light and variable. On March 2, 2009, winds were also variable with speeds less than 10 m/s during the morning, but became southeasterly in the late afternoon. It is possible that these southeasterlies are an incipient barrier wind that occasionally forms along the western flank of the Wind River mountains and subsequently extends over the UGRB (Emery et al., 2015). This wind results in the advection of O3 and its precursors from source regions in the Pinedale Anticline and Jonah field towards the northwestern boundary of the basin.

doi: 10.12952/journal.elementa.000074.f003.
Figure 3.  

Ozone mixing ratio (ppbv) isopleth contour maps derived from continuous analyzer data for March 2, 2009 at 06:00, 12:00, 17:00 and 22:00 MST.

Contour maps are constructed by interpolating ozone measurements from 21 continuously operating monitoring sites. Site locations are indicated by pink triangles. Wells are superimposed to aid as geographic markers.

The O3 distributions on February 22, 2009 and March 2, 2009, when elevated O3 is found largely in the Pinedale Anticline and Jonah field, are typical of those found on days with similar wind patterns, suggesting that localized production processes (Schnell et al., 2009) are dominant. Measurements of high wintertime O3 in surrounding areas only occurs under particular wind regimes when advection transports O3 and/or its precursors to those areas, such as during barrier wind events as noted above (Emery et al. 2015). In 2011 the observation of elevated O3 levels at WDEQ monitoring stations in the town of Pinedale, and in undeveloped rural areas on the northwestern perimeter of the basin, was evident when southerly wind flow occurred from areas that had earlier experienced episodic O3 production. While meteorology remains an important factor, a lack of wintertime O3 episodes from 2012 to 2015 is most likely due to reduced emissions from O&NG sources, from a combination of voluntary efforts by operators (e.g. leak detection and repair programs), regulatory actions (WDEQ, 2015b) and reduced explorations activities, as indicated by rig counts in the Green River Basin (Sublette, Lincoln, Sweetwater and Lincoln counties) declining from a peak of 32 in 2011 to ∼14 from 2012 to 2014 (Baker Hughes, 2015).

3.2 BTEX

Passive diffusive sampling surveys of BTEX were carried out on three occasions. The first survey was performed over a one week period (2/2011), whereas surveys two (11/2011) and three (2/2012) were each of 3 days duration. Figures 4a4e illustrate BTEX distributions in a series of iso-concentration contour maps for the third survey. The depicted third survey was dominated by northwesterly winds up to 10 m/s. Within the O&NG development areas, mixing ratios are clearly above background levels with two significant ‘hotspots’ evident. Contour plots for the other two sampling surveys (not shown) exhibit very similar mixing ratio patterns with the same two hotspots. Benzene mixing ratios are greatest (> 4 ppbv) at the more northerly hotspot (designated ‘Hotspot 1’), whereas ethylbenzene and xylenes are largest at the southerly hotspot (designated ‘Hotspot 2’), indicating different predominant emission sources at the two locations. Peak toluene mixing ratios are similar at the two hotspots, while the extent of the northern hotspot is greater.

doi: 10.12952/journal.elementa.000074.f004.
Figure 4.  

BTEX and NOx (ppbv) isopleth contour maps derived from passive diffusive sampling data for February 7–10, 2012.

Contour maps are constructed from 72 hour averaged measurements at ∼50 diffusive sampling sites. Site locations are indicated by pink rectangles. Figures 4a–4e show maps for 5 individual BTEX species as labeled. Figure 4f shows the contour map for NOx. Note the different color-coded mixing ratio scales for each pollutant. Wells are superimposed to aid as geographic markers.

Hotspot 1 (∼42.75°N, 109.85°W) is located in the Mesa area of the Pinedale Anticline natural gas field, an area with widespread production and numerous exploration activities. Intensive drilling was occurring in this area during the survey, with some 11 – 16 actively operating drill rigs, powered by engines known to emit VOC. In addition, ancillary activities associated with well drilling and completion, such as flow back, are known to have high BTEX emissions (Field et al., 2015). The area has also been observed to contain grandfathered uncontrolled dehydration units. In sum, measurements suggest that elevated BTEX levels at this hotspot may be attributed to intensive local O&NG operations. The southerly Hotspot 2, shown in Figure 4 at ∼42.65°N, 109.75°W, indicates the influence of BTEX emissions from a wastewater treatment and recycling facility. While not defined as a Title V facility, it is known to be an important point source of VOC (Field et al., 2015), with a far higher emission density than elsewhere in the UGRB.

These non-uniform distributions of elevated BTEX mixing ratios illustrate the utility of detailed, high-resolution spatial measurements when evaluating emissions in O&NG development areas. For the UGRB, the origin of the southerly hotspot, the wastewater treatment facility, is now well established. Additional fence line measurements, to be reported elsewhere, performed in 2014 and 2015, utilizing passive sampling with GC-MS analysis and mobile monitoring using PTR-TOF-MS analysis, have confirmed the hotspots noted here. However, the relative contributions of the specific sources associated with gas well drilling, completion, and production operations such as those close to Hotspot 1 on the Mesa have yet to be unambiguously determined. It is clearly important that the influence, strength, and temporal behavior of emission sources are quantitatively characterized, not only to develop better emissions inventories, but to enable observed ambient levels of hazardous pollutants and their concentration gradients to be both explained and predicted. To this end we are conducting on-going investigations to better understand the importance of the role of wet gas processing at Hotspot 1, and at other locations, to ambient BTEX levels. By contrast to typical well pads, as evident in the Jonah Field, that have on site storage of produced water and condensate, those at the Pinedale Anticline are directly connected to liquids gathering facilities that transport water and condensate off site in pipelines, eliminating the need for storage tanks at well pads.

3.3. Oxides of nitrogen

Diffusive sampling surveys of NOx were performed in conjunction with BTEX measurements, for the same time periods. Nitrogen oxides are generated from internal combustion engines, and the presence of NOx in ambient air often signifies a traffic source, especially when highway vehicle densities are high (Beevers et al., 2012). Measurements in the UGRB are in line with this conclusion (Soltis and Field, 2012a), and, as samplers were placed at roadside locations, traffic influences are to be expected. Traffic surveys have shown the dominance of traffic related to O&NG activities (Field and Soltis, 2009a) with ∼70% pickup trucks and ∼30% multi-axle vehicles. Figure 4f illustrates NOx distributions as an iso- concentration contour map for the third survey, carried out in February 2012. All three surveys exhibited similar pollutant distributions and NOx mixing ratios. While two hotspots are shown, they are close to but not coincident with those for BTEX. The more northerly hotspot (NOx Hotspot 1 at ∼42.70N, 109.80W) is associated with the Pinedale Complex in an area known as Gobblers Knob near the junction of Paradise and North Anticline Roads. Gas compression facilities are also present in this area. The second area (NOx Hotspot 2 at ∼42.60N, 109.70W) includes the junctions of US Highways 191 and 351, and Middle Crest Road. The passive sampler data show that NOx mixing ratios are generally less than 10 ppbv above the background level (<1 ppbv), while those at the two hotspots are up to 20 ppbv above background.

One purpose of the construction of BTEX and NOx contour plots was to assist on-going photochemical grid modeling studies designed to predict O3 formation in the UGRB. Unfortunately, the WDEQ emissions inventory database used in these studies (Rodriguez et al. 2014) could not be evaluated or refined using the BTEX and NOx observations as it has yet to be released for external review. Rodriguez et al. (2014) report that their O3 modeling efforts are unable to replicate episodic O3 production, in part because the emissions database includes VOC data that appear to be an order of magnitude too low. Comparisons of ambient data measured at the University of Wyoming Boulder South Road monitoring site and the WDEQ winter emission inventory show excellent agreement in terms of relative masses of VOC (as total NMHC), n-hexane, benzene, toluene and xylenes (R2 = 0.99). However if NOx values are correct then the VOC emissions rates in the 2011 WDEQ winter (February and March) inventory could be a factor of approximately 5 too low. Edwards et al. (2014) and Ahmadov et al. (2015) demonstrate that measurement based model inputs of appropriate values of VOC and NOx, together with appropriate derivation of radicals for photochemical mechanisms, are critical to ensure effective modeling of wintertime O3 in the Uintah Basin, Utah. Model development for the UGRB remains critical, despite a lack of O3 episodes in the past few years, so that further development activities in the basin do not result in 8-hour O3 averages that exceed the current 75 ppbv standard value, or the new standard value of 70 ppbv set in October 2015 (US EPA, 2014).

3.4 Volatile organic compounds

Diffusive sampling of VOC was also performed, in conjunction with a longer-term VOC monitoring project that included canister sampling (Field et al., 2015) at 13 UGRB monitoring sites. Diffusive sampling results from six of the sites are shown in Figure 5, from a pilot survey for prototype samplers conducted in and around the Pinedale Anticline commencing on February 14, 2012 and extending for 72 hours. This survey was dominated by northwesterly winds up to 10 m/s. Four of the sites in Figure 5 are the identified NOx and BTEX hotspots. Also shown are data from the WDEQ Boulder monitoring station, which typically receives air containing well-mixed O&NG emissions. A site on the heavily trafficked main street in the town of Pinedale is also included. Concentrations for four of thirteen measured VOC species are shown, selected as marker compounds because their relative abundances reflect the strengths of different emission sources. For example, the ratio by mass of n-pentane to toluene is typically about 4:1 in fugitive natural gas emissions, but only about 1:4 in condensate (Wyoming Oil and Gas Conservation Commission, 2014). While fugitive unprocessed wet natural gas emissions should match those of wet natural gas, emissions from condensate may vary depending on the adopted product handling approach and other environmental conditions. The i-pentane to n-pentane ratio (iC5/nC5) can also be used to contrast areas with significant O&NG emissions from those with greater influence from traffic (Gilman et al., 2013). The iC5/nC5 ratio of ∼1.2:1 (median) from the VOC survey, excluding town of Pinedale sites, indicates the dominance of O&NG emissions and is similar to that reported in natural gas (1.4:1). However while the iC5/nC5 ratio is almost constant (r2 = 0.98), that of toluene to n-pentane shows greater variation. Emissions from traffic sources within the Pinedale Anticline are reported to contribute minimally to VOC levels compared to O&NG sources (Field et al., 2015). Higher ratios of toluene to n-pentane suggest a greater contribution of emissions other than those from fugitive natural gas. Aromatic compounds appear to be relatively elevated in the Pinedale Anticline compared to measurements in the Uintah basin (Helmig et al., 2014; Warneke et al., 2014), when considering ratios expressed relative to mixing ratios of alkanes. This difference is likely due to a combination of the selection of hotspot sites (closer to emission sources in the UGRB), the consistently high non-methane hydrocarbon content of natural gas in the Pinedale Anticline, the hydrocarbon product handling approach, and the density of development in the UGRB. With the exception of BTEX Hotspot 2 and the town of Pinedale site, species distributions suggest that a mix of O&NG emission sources determine the observed compound concentrations at the remaining four sites shown in Figure 5. At BTEX Hotspot 2, relatively high toluene to benzene (∼3:1), and n-octane to n-pentane (∼1:1) ratios represent nearby emissions rich in heavier VOC, as expected due to the close proximity of this site to a water treatment and recycling facility. By contrast, at the town of Pinedale site, low ratios for both toluene to benzene (∼1:1) and n-octane to n-pentane (∼0.01:1) indicate the predominance of traffic emissions, albeit at low mixing ratios due to the low traffic flows compared to more populous urban areas (Schauer et al., 2002).

doi: 10.12952/journal.elementa.000074.f005.
Figure 5.  

Averaged mixing ratios (ppbv) of selected VOC at six sites during the diffusive sampling survey of February 14–17 2012.

Data from the WDEQ Boulder monitoring site and a roadside site in downtown Pinedale are presented with remaining four sites that correspond to those with high BTEX and NOx mixing ratios, identified by the plots in Figure 4 (see text). Figure 5: Mixing ratios (ppbv) of four selected VOC species, as labeled.

4. Conclusions

Passive diffusive sampling measurements were carried out to determine spatial distributions of O3, NOx, VOC, and BTEX in the UGRB of Wyoming, where emissions are dominated by O&NG sources. Visualization of pollutant distributions, facilitated by constructing iso-concentration contour plots, enable an appreciation of geographical differences driven by contributing emission sources.

Contour maps for O3 reveal locations where episodic levels of photochemically produced O3 are found in winter. Measurements from continuously operating O3 monitors, including long-term WDEQ sites, confirm the diffusive sampler O3 distribution patterns, while providing a clear resolution of the localized nature of wintertime episodes. BTEX contour maps show two hotspots with mixing ratios elevated above background. The first of these, in the Mesa area of the Pinedale Anticline, is associated with production activities, whereas the second results from emissions from a wastewater treatment and recycling facility. While NOx mixing ratios were also elevated above background levels, two hotspots different from those for BTEX were identified, both associated with traffic emissions. Diffusive sampling of VOC supports the findings of the more extensive BTEX sampling. BTEX hotspots have minimal traffic influence, a finding supported by the observed patterns and ratios of VOC species, confirming the dominance of O&NG sources.

Passive diffusive sampling is an inexpensive and simple-to-use measurement technique that provides valuable quantitative assessments of pollutant spatial distributions that are readily visualized by contour mapping. In addition to providing spatial distribution information such mapping techniques have the potential to support the development and evaluation of emission inventories and air quality modeling. Model output invariably forms a key factor influencing regulatory air quality management. Photochemical grid modeling approaches are reliant upon the accuracy of spatially aggregated emissions. Spatially distributed and temporally integrated ambient air quality measurements can provide supplemental information for input parameters (e.g. O3 precursors) to assess emission inventories and air quality model simulations (e.g. O3). The value of future diffusive air sampling studies can be enhanced by increasing both the number of reported chemical species, which for VOCs increases the utility of the data for source apportionment analysis, and the sampler distribution density, which improves the confidence of derived species spatial distributions. Besides serving as an effective tool for local and regional studies, longer term temporal monitoring is also likely to become more common as the technique becomes better accepted as a methodology with known, and relatively low, detection limits and measurement uncertainty.

Data accessibility statement

The data used for this article are available from the corresponding author upon request. The data for the O3i and PASQUA projects are also available from the Wyoming Department of Environmental Quality.


© 2015 Field et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Supplemental material

Figure S1.
Wind roses for WDEQ Boulder site, February 22, 2009, March 2, 2009, February 7–10, 2012, and February 14–17, 2012.
File Type: DOC
File Size: 4.78 MB


Contributed to conception and design: RAF, JJS

Contributed to acquisition of data: RAF, JJS

Contributed to analysis and interpretation of data: RAF, JJS, DCM, EG, PPB

Drafted and/or revised the article: RAF, JJS, DCM

Approved the submitted version for publication: RAF, JJS, DCM, EG, PPB

Competing interests

The authors declare that they do not have competing interests.

Funding information

R.A. Field, D.C. Montague and J.J. Soltis gratefully acknowledge the financial support of the Wyoming Department of Environmental Quality for the Upper Green River Ozone (O3i) project, the Pinedale Anticline Project Office (PAPO) for the Pinedale Anticline Spatial Air Quality Assessment (PASQUA) project and the UW School of Energy Resources, which made the presented analyses possible.


First and foremost we would like to thank all of the people that deployed samplers for the ambient surveys. This project would not have been possible without this voluntary commitment and support. Our sincere gratitude is sent to the following people and the organizations that they represented: Sublette County Conservation District (Meghann Durbrow, Delsa Allen and Loren Racich); Sublette County Health District (Keith Raney); Wyoming BLM (Tim Zebulske, Shelly and Jed Gregory, Mark Thonoff, Kelly O’Kell, Kyle Schumacher, Brian Roberts, Cindi Etcheverry and Janet Bellis); USFS (Ted Porwoll and Ed Sherman); The Log Cabin Motel (Forest and Sofia Wakefield); Citizens United for Responsible Energy Development (Bev and Sam Sharp, Carmel Kail, Perry Walker, Tina Rock, Elaine Crumpley and Mary Lynn Worl); WDEQ (Jennifer Frazier and Brandi O’Brien); Town of Pinedale (Steve Smith, Mary Hogarty, Kate Grimes, Patty and Dave Racich); The Wyoming Outdoor Council (Dave Hohl); Shell (Jim Sewell, David Osborn, Charles Travis, Bill Cook, Levi Ackland, Jasmine Allison, Aimee Davidson and Nancy Feck); BP (Don Brooks and Todd Morgan); and QEP (Kevin Williams). We also thank Kelly Bott, Ken Rairigh, Lori Bocchino of WDEQ for their oversight of the project. Finally the support of ShihMing Huang and Michael McCarthy with Figure 1 is appreciated. Finally we really appreciate time spend by the reviewers of the manuscript. The revisions suggested by the two anonymous referees and the editorial team of Detlev Helmig and Gaby Pétron improved the quality of the final manuscript.


  1. Ahmadov R , McKeen S , Trainer M , Banta R , Brewer A , et al. 2015. Understanding high wintertime ozone pollution events in an oil and natural gas producing region of the western US. Atmos Chem Phys 15: 411–429. doi: 10.5194/acp-15-411-2015.

  2. Baker Hughes. 2015. North American Rig Count.

  3. Beevers SD, Westmoreland E, De Jong M, Williams ML, Carslaw DC. 2012. Trends in NOx and NO2 emissions from road traffic in Great Britain. Atmos Environ 54: 107–116. doi: 10.1016/j.atmosenv.2012.02.028.

  4. Brandt AR , Heath GA , Kort EA , O’Sullivan F , Pétron G , et al. 2014. Methane leaks from North American natural gas systems. Science 343: 733–735. doi: 10.1126/science.1247045.

  5. Brown RH. 2000. Monitoring the ambient environment with diffusive samplers: Theory and practical considerations. J Environ Monit 2: 1–9. doi: 10.1039/A906404D.

  6. Bruno P, Caputi M, Caselli M, de Gennaro G, de Rienzo M. 2005. Reliability of a BTEX radial diffusive sampler for thermal desorption: Field measurements. Atmos Environ 39: 1347–1355. doi: 10.1016/j.atmosenv.2004.11.012.

  7. Buczynska AJ , Krata A , Stranger M , Godoi AFL , Kontozova-Deutsch V , et al. 2009. Atmospheric BTEX concentrations in an area with intensive street traffic. Atmos Environ 43: 311–318. doi: 10.1016/j.atmosenv.208.09.071.

  8. Buzica D, Gerboles M, Amantini L, Pérez Ballesta P, De Saeger E. 2005. Modelling of the uptake rate of the nitrogen dioxide Palmes diffusive sampler based on the effect of environmental factors. J Environ Monit 7: 169–174. doi: 10.1039/B411474D.

  9. Carmichael GR , Ferm M , Thongboonchoo N , Woo J-H , Chan LY , et al. 2003. Measurements of sulfur dioxide, ozone, and ammonia concentrations in Asia, Africa, and South America using passive samplers. Atmos Environ 37: 1293–1308. doi: 10.1016/S1352-2310(02)01009-9.

  10. Carter WP, Seinfeld JH. 2011. Winter ozone formation and VOC incremental reactivities in the Upper Green River Basin of Wyoming. Atmos Environ 50: 255–266. doi: 10.1016/j.atmosenv.2011.12.025.

  11. Cocheo C , Sacco P , Pérez-Ballesta P , Donato E , Garcia S , et al. 2008. Evaluation of the best compromise between the urban air quality monitoring resolution by diffusive sampling and resource requirements. J Environ Monit 10: 941–950. doi: 10.1039/B806910G.

  12. Cocheo V, Boaretto C, Sacco P. 1996. High uptake rate radial diffusive sampler suitable for both solvent and thermal desorption. Am Ind Hyg Assoc J 57: 897–904.

  13. Edwards PM , Brown SS , Roberts JM , Ahmadov R , Banta RM , et al. 2014. Unconventional photochemistry underlying winter ozone in an oil and gas producing region. Nature 514: 351–354. doi: 10.1038/nature13767.

  14. Emery BR, Field RA, Parish TR, Montague DC. 2015. Barrier wind formation in the Upper Green River Basin of Sublette County, Wyoming and Its Relationship to Wintertime Elevated Ozone Distributions. J Appl Met Climat , in press. doi: 10.1175/JAMC-D-15-0103.1.

  15. Environ International Corporation. 2010. Final Report: 2009 Upper Green River winter ozone study.

  16. Field RA, Soltis J. 2009a. Upper Green River Ozone Investigation (O3i) Luman and Paradise Road Traffic Count Survey. Report . University of Wyoming.

  17. Field RA, Soltis J. 2009b. Upper Green River Ozone Investigation (O3i) Ozone Spatial Distribution Survey. Report . University of Wyoming.

  18. Field RA, Soltis J, McCarthy MC, Murphy SM, Montague DC. 2015. Influence of oil and gas field operations on spatial and temporal distributions of atmospheric non-methane hydrocarbons and their effect on ozone formation in winter. Atmos Chem Phys 15: 3527–3542. doi: 10.5194/acp-15-3527-2015.

  19. Field RA, Soltis J, Murphy S. 2014. Air Quality Concerns of Shale Gas Production. Environ Sci . doi: 10.1039/C4EM00081A.

  20. FSM (Fondazione Salvatore Maugeri). 2006. Radiello Manual English Version 01.’s%20manual%2001-06.pdf.

  21. Gerboles M, Buzica D, Amantini L, Lagler F. 2006. Laboratory and field comparison of measurements obtained using the available diffusive samplers for ozone and nitrogen dioxide in ambient air. J Environ Monit 8: 112–119. doi: 10.1039/B511271K.

  22. Gilman JB, Lerner BM, Kuster WC, de Gouw JA. 2013. Source signature of volatile organic compounds from oil and natural gas operations in Northeastern Colorado. Environ Sci Technol 47: 1297–1305.

  23. Helmig D , Thompson CR , Evans J , Boylan P , Hueber J , et al. 2014. Highly elevated atmospheric levels of volatile organic compounds in the Uintah Basin, Utah. Environ Sci Technol 48(9): 4707–4715. doi: 10.1021/es405046r.

  24. Isaaks EH, Srivastava RM. 1989. An Introduction to Applied Geostatistics . New York: Oxford University Press.

  25. Koutrakis P , Wolfson JM , Bunyaviroch A , Froelich SE , Hirano K et al. 1993. Measurement of ambient ozone using a nitrite-saturated filter. Anal Chem 65: 210–214.

  26. Martin NA , Helmore JJ , White S , Barker Snook IL , Parish A , et al. 2014. Measurement of nitrogen dioxide diffusive sampling rates for Palmes diffusion tubes using a controlled atmosphere test facility (CATFAC). Atmos Environ 94: 529–537. doi: 10.1016/j.atmosenv.2014.05.064.

  27. Mayer P, Tolls J, Hermens L, Mackay D. 2003. Equilibrium Sampling Devices. Environ Sci Technol 37: 184a–191a.

  28. Muckerjee S , Smith LA , Norris GA , Morandi MT , Gonzales M , et al. 2004. Field method comparison between passive air samplers and continuous monitors for VOCs and NO2 in El Paso, Texas. J Air Waste Manage Assoc 54(3): 307–319. doi: 10.1080/10473289.2004.10470903.

  29. Ogawa. 2006. NO, NO2, NOx, and SO2 sampling protocol using the Ogawa sampler.

  30. Oltmans S , Schnell R , Johnson B , Pétron G , Mefford T , et al. 2014. Anatomy of wintertime ozone associated with oil and natural gas extraction activity in Wyoming and Utah. Elem Sci Anth 2: 000024. doi: 10.12952/journal.elementa.000024.

  31. Palmes ED, Gunnison AF. 1973. Personal monitoring device for gaseous contaminants. Am Ind Hyg Assoc J 34: 78–81. doi: 10.1080/0002889738506810.

  32. Pennequin-Cardinal A , Plaisance H , Locoge N , Ramalho O , Kirchner S , et al. 2005. Performances of the Radiellos diffusive sampler for BTEX measurements: Influence of environmental conditions and determination of modelled sampling rates. Atmos Environ 39: 2535–2544.

  33. Pérez Ballesta P. 2014. Diffusive Sampling Device. P. Number Wo 2014/076153 A1 22.05.2014. World Intellectual Property Organization. International Application Published under the Patent Cooperation Treaty (PCT).PCT/EP2013/073765 .

  34. Pérez Ballesta P , Field RA , Fernández-Patier R , Galán Madruga D , Connolly R , et al. 2008. An approach for the evaluation of exposure patterns of urban populations to air pollution. Atmos Environ 42(21): 5350–5364. doi: 10.1016/j.atmosenv.2008.02.047.

  35. Pérez Ballesta P, Grandesso E, Field RA. 2014. The modeling and validation of a novel diffusive sampler for determining concentrations of volatile organic compounds in air. J Anal Chem A : in preparation.

  36. Pétron G , Karion A , Sweeney C , Miller BR , Montzka SA , et al. 2014. A new look at methane and non-methane hydrocarbon emissions from oil and natural gas operations in the Colorado Denver-Julesburg Basin. J Geophys Res 119(11): 6836–6852. doi: 10.1002/2013JD021272.

  37. Pollack IB , Ryerson TB , Trainer M , Neuman JA , Roberts JM , et al. 2013. Trends in ozone, its precursors, and related secondary oxidation products in Los Angeles, California: A synthesis of measurements from 1960 to 2010. J Geophys Res 118(11): 5893–5911. doi: 10.1002/jgrd.50472.

  38. Rappenglück B , Ackermann L , Alvarez S , Golovko J , Buhr M , et al. 2014. Strong wintertime ozone events in the Upper Green River Basin, Wyoming. Atmos Chem Phys 14: 4909–4934. doi: 10.5194/acp-14-4909-2014.

  39. Rodriguez M , Reed J , Chien C-J , Taylor C , Reid S , et al. 2014. Ozone modeling results and analyses for winter in Sublette County, Sweetwater County and Lincoln County Wyoming. Interim report from AECOM to the Wyoming Department of Environmental Quality .

  40. Schauer JJ, Fraser MP, Cass GR, Simoneit BRT. 2002. Source reconciliation of atmospheric gas-phase and particle-phase pollutants during a severe photochemical smog episode. Environ Sci Technol 36(17): 3806–3814. doi: 10.1021/es011458j.

  41. Schnell RC , Oltmans SJ , Neely RR , Endres MS , Molenar JV , et al. 2009. Rapid photochemical production of ozone at high concentrations in a rural site during winter. Nature Geoscience 2: 120–122. doi: 10.1038/ngeo415.

  42. Sillman S. 1999. The relation between ozone, NOx and hydrocarbons in urban and polluted rural environments. Atmos Environ 33: 1821–1845.

  43. Soltis J, Field RA. 2011. Pinedale anticline spatial air quality assessment (PASQUA). 2010–2011 Spatial Distribution Surveys. Report . University of Wyoming.

  44. Soltis J, Field RA. 2012a. Pinedale anticline spatial air quality assessment (PASQUA). Mobile laboratory monitoring of ozone precursors at Boulder South Road site 10/31/2011 to 03/30/2012. Report . University of Wyoming.

  45. Soltis J, Field RA. 2012b. Pinedale anticline spatial air quality assessment (PASQUA). 2011–2012 Spatial Distribution Surveys. Report . University of Wyoming.

  46. United States Energy Information Administration (US EIA). 2014. Annual energy outlook 2014. DOE/EIA-0383(2014).

  47. United States Energy Information Administration (US EIA). 2015. Top 100 U.S. oil and gas fields.

  48. United States Environmental Protection Agency (US EPA). 2014. National Ambient Air Quality Standards for Ozone: Proposed rule. Federal Register Vol. 79, No. 242/Wednesday, December 17, 2014/Proposed Rules .

  49. Varns JL , Mulik JD , Sather ME , Glen G , Smith L , et al. 2001. Passive ozone network of Dallas: A modeling opportunity with community involvement 1. Environ Sci Technol 35: 845–855. doi: 10.1021/es001311c.

  50. Warneke C , de Gouw JA , Holloway JS , Peischl J , Ryerson TB , et al. 2012. Multiyear trends in volatile organic compounds in Los Angeles, California: Five decades of decreasing emissions. J Geophys Res 117: D00V17. doi: 10.1029/2012JD017899.

  51. Warneke C , Geiger F , Edwards PM , Dube W , Pétron G , et al. 2014. Volatile organic compound emissions from the oil and natural gas industry in the Uintah Basin, Utah: Oil and gas well pad emissions compared to ambient air composition. Atmos Chem Phys 14: 10977–10988. doi: 10.5194/acp-14-10977-2014.

  52. Wyoming Department of Environmental Quality (WDEQ). 2015a. Winter Ozone study.

  53. Wyoming Department of Environmental Quality (WDEQ). 2015b. Upper Green River Basin Ozone Strategy. April 2015.

  54. Wyoming Oil and Gas Conservation Commission. 2014. Data base for Gas and Liquid Analysis by Field, Pinedale records.

  55. Wyoming Oil and Gas Conservation Commission. 2015. Data base for Production by Field, Pinedale and Jonah records.

  56. Zabiegala B, Kot-Wasik A, Urbanowicz M, Namieśnik J. 2010. Passive sampling as a tool for obtaining reliable analytical information in environmental quality monitoring. Analytical Bioanalytical Chem 396(1): 273–296. doi: 10.1007/s00216-009-3244-4.

comments powered by Disqus