Chemico-Biological Interactions xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Chemico-Biological Interactions journal homepage: www.elsevier.com/locate/chembioint
Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012 Jessica L. Myers, Tracie Phillips, Roberta L. Grant ⇑ Texas Commission on Environmental Quality, Toxicology Division, Austin, TX, USA
a r t i c l e
i n f o
Article history: Available online xxxx Keywords: Lower olefins Monitoring Butadiene Ethylene Isoprene Propylene
a b s t r a c t Texas has the largest ambient air monitoring network in the country with approximately 83 monitoring sites that measure ambient air concentrations of volatile organic compounds (VOCs). The lower olefins, including 1,3-butadiene, ethylene, isoprene, and propylene, are a group of VOCs that can be measured in both 24 h/every sixth-day canister samples and continuous 1-h Automated Gas Chromatography (AutoGC) samples. Based on 2012 Toxics Release Inventory data, the total reported industrial air emissions in Texas for these olefins, as compared to total national reported air emissions, were 79% for 1,3butadiene, 62% for ethylene, 76% for isoprene, and 54% for propylene, illustrating that Texas industries are some of the major emitters for these olefins. The purpose of this study was to look at the patterns of annual average air monitoring data from 2002 to 2012 using Texas Commission on Environmental Quality (TCEQ) data for these four lower olefins. It should be emphasized that monitors may not be located close to or downwind of the highest emitters of these lower olefins. In addition, air monitors only provide a snapshot in time of air concentrations for their respective locations, and may not be able to discriminate emissions between specific sources. In 2012, the highest annual average air concentration for 1,3-butadiene was 1.28 ppb by volume (ppbv), which was measured at the Port Neches monitoring site in Region 10-Beaumont. For ethylene, the highest 2012 annual average air concentration was 5.77 ppbv, which was measured at the Dona Park monitoring site in TCEQ Region 14-Corpus Christi. Although reported industrial emissions of isoprene are predominantly from the Houston and Beaumont regions, trees are natural emitters of isoprene, and the highest ambient air concentrations tend to be from regions with large areas of coniferous and hardwood forests. This was observed with TCEQ Region 5-Tyler, which had the two highest isoprene annual average air concentrations for 2012: 0.56 ppbv at the Karnack monitoring site and 0.47 ppbv at the Longview monitoring site. For propylene, the highest 2012 annual average air concentration was recorded at the HRM 7 monitoring site in TCEQ Region 12-Houston, which was 7.9 ppbv. A significant portion of the total 2012 industrial propylene emissions were also reported in TCEQ Region 12Houston. Although some individual monitors showed increased annual averages from 2002 to 2012, there was a general decreasing trend present across the state for all four lower olefins examined. The annual average air concentrations of the four lower olefins were well below their respective Air Monitoring Comparison Values (AMCVs) and are not expected to cause long-term or chronic adverse health effects. Published by Elsevier Ireland Ltd.
Abbreviations: AL, Air Laboratory; AMCV, Air Monitoring Comparison Value; APWL, Air Pollutant Watch List; AutoGC, Automated Gas Chromatography; CFR, Code of Federal Regulation; EI, Emissions Inventory; EPA, Environmental Protection Agency; HECT, HRVOC Emission Cap and Trade; HRM, Houston Regional Monitoring; HRVOC, highly reactive VOC; LIDAR, Differential Absorption Light Detection and Ranging; LOD, limit of detection; MDL, method detection limit; ppbv, parts per billion by volume; RL, reporting limit; RN, Registered Identification Number; TAMIS, Texas Air Monitoring Information System; TCEQ, Texas Commission on Environmental Quality; TPY, tons per year; TRI, Toxics Release Inventory; VOC, volatile organic compound. ⇑ Corresponding author at: Texas Commission in Environmental Quality, P.O. Box 13087 (MC-168), Austin, TX 78711-3087, USA. E-mail address: [email protected] (R.L. Grant).
1. Introduction 1.1. Industrial use of the lower olefins in Texas Texas industries include some of the major emitters of 1,3-butadiene, ethylene, isoprene, and propylene in the nation. Based on the Environmental Protection Agency’s (EPA’s) 2012 Toxics Release Inventory (TRI) data, the total reported industrial air emissions in Texas for these olefins, as compared to total national reported air emissions, were 79% for 1,3-butadiene, 62% for ethylene, 76% for isoprene, and 54% for propylene.
http://dx.doi.org/10.1016/j.cbi.2015.02.008 0009-2797/Published by Elsevier Ireland Ltd.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
2
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
1,3-Butadiene is used as an intermediate to produce a variety of industrial chemicals (e.g., polymers, elastomers), but its major uses are in the manufacturing of styrene-butadiene rubber, or synthetic rubber, and thermoplastic resins. The primary way 1,3-butadiene is released into the environment is from emissions from gasolineand diesel-powered vehicles and equipment [18]. Ethylene is produced through both anthropogenic and natural activities [19]. It is a high-production-volume chemical product of the petrochemical industry, and is primarily used as an intermediate in the production of other chemicals. As a natural compound, ethylene is a plant hormone that enhances the ripening process of fruits, vegetables, and flowers. Some microbes and higher plants naturally produce ethylene. Ambient ethylene is also produced during incomplete combustion of biomass and fossil fuels, forest fires, and active volcanic events. A relatively large proportion of ethylene in urban air is due to vehicular traffic emissions [1]. Isoprene, the 2-methyl analogue of 1,3-butadiene, is also produced through both anthropogenic and natural activities. It is used in the manufacturing of synthetic rubber, styrene–isoprene-styrene block co-polymers and butyl rubber, and in the production of hydrocarbon resins and synthesis of terpenes [3,12,16]. It can also be released in oil fires, tobacco smoke, and automobile exhaust [10,16]. Isoprene is produced naturally by living organisms (flora and fauna), and the amount of isoprene produced and emitted by biological systems far exceeds that which is produced synthetically [10,17]. There are over 200 different plant species that emit isoprene, with trees being a large emitter [11]. Propylene is produced and consumed in refinery operations for the production of components of gasoline, and is also a chemical intermediate in the manufacture of various chemicals (e.g., acetone, isopropylbenzene, isopropanol, propylene oxide, etc.). It is used in polymerized form as polypropylene for plastics and carpet fibers and as an aerosol propellant [9]. As with isoprene and ethylene, propylene is emitted by certain tree species, and produced by other types of vegetation. Propylene is often found in air samples in industrialized areas and is also a combustion product present in cigarette smoke, aircraft exhaust, and motor vehicle exhaust.
1.2. Texas ambient air toxics monitoring network Texas has one of the most extensive ambient air toxics monitoring networks in the country. With approximately 100 air toxics monitoring sites across the state, the Texas Commission on Environmental Quality (TCEQ) analyzes for approximately 120 of the air toxics compounds, as identified by the 1990 EPA Clean Air Act. These toxics include volatile organic compounds (VOCs), semivolatile organics, carbonyls, and metals. The TCEQ has extensive ambient air monitoring data for 1,3-butadiene, ethylene, isoprene, and propylene across the state for numerous years. With a state that covers a large area, in order to better provide services to the citizens and regulated community, the TCEQ has broken down the state into four regional areas, which are further broken down into 16 individual regions. The monitoring sites that the TCEQ oversees are located within these regions; these monitors include stationary monitors that belong to various entities (state, local governments, councils of governments, and industrysponsored). There are two primary types of stationary ambient air monitors that measure VOCs in the state: canister samplers that collect twenty-four-hour samples every sixth-day, and Automated Gas Chromatography (AutoGC) samplers that collect 40 min samples every hour (referred to as a one-hour sample). It should be emphasized that monitors only provide a snapshot in time of ambient air concentrations at their respective locations.
1.3. Spatial and temporal trends of the identified lower olefins The purpose of this study was to evaluate the spatial and temporal trends of the identified lower olefins (1,3-butadiene, ethylene, isoprene, and propylene) across the state of Texas for 2002 through 2012. Chloroprene was not included in this analysis since there are no available monitoring data. The five monitoring sites measuring the highest concentrations in 2012 for each olefin were identified (hereafter referred to as the top five monitoring sites). Trend data of annual average ambient air concentrations for each olefin from 2002 to 2012 for the top five monitoring sites are presented. A further analysis of the ambient air monitors with the highest olefin annual average concentrations was then conducted. The five industrial sources that had the highest 2012 atmospheric emissions in Texas for each olefin were also identified through the Texas Emissions Inventory (EI) (hereafter referred to as the top five EI sites). Monitoring sites located within two miles of the top five EI sites property boundaries were also identified. It should be noted that this study does not discriminate emissions between specific industries, as ambient air monitoring data cannot provide that type of information. 2. Material and methods 2.1. Air monitoring data Air monitoring data (24-h canister and 1-h AutoGC samples) were acquired from either the TCEQ air monitoring databases (e.g., the Texas Air Monitoring Information System (TAMIS)) or from an industry-sponsored monitoring network. Due to proximity to the identified industries in this evaluation, data for three monitoring sites were requested from the Houston Regional Monitoring (HRM) Network and included in this evaluation: two canister sites – HRM 7 and HMR 8; and one AutoGC site – HRM 16. The canister samples are collected continuously for 24 h in silica-coated stainless steel canisters every sixth-day, and then sent off to be analyzed by gas chromatography/mass spectroscopy by the TCEQ Austin Air Laboratory (AL) using modified EPA Method TO-15. The AutoGC samples are collected continuously for 40 min every hour and analyzed automatically on-site using modified EPA Method TO-14. For 2012, there were 81 monitors in the TCEQ databases and 3 monitors from an industry-sponsored monitoring network that had data available on the identified olefins. Data from 2002 through 2012 (where available) were also collected for each of the four identified chemicals from the top five monitoring sites to visualize trends over the last eleven years. 2.2. Analysis of the air monitoring data For each of the monitoring sites examined, the data was considered complete when 75% of the total possible samples were collected and validated for the calendar year. For the 24-h canister data collected every sixth-day, 75% completeness is represented by at least 45 out of a possible 60 samples each year. For the 1-h AutoGC continuous data, 75% completeness is represented by at least 6570 out of a possible 8760 samples each year. Unless otherwise indicated, data were generally used from monitoring sites if it met 75% completeness. One exception includes the 2002–2012 trend graphs of the top five monitoring sites for each olefin, where incomplete previous years were still used for visualization of the trends (identified by #). Air Monitoring Comparison Values (AMCVs) set by the TCEQ are chemical-specific air concentrations derived to protect human health and welfare, including odor, animals, and vegetation; each data set was compared to its lowest respective AMCV.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
The TCEQ AL determines data quality objectives in terms of the limit of detection (LOD). For all VOCs, the LODs are established using the methods outlined in 40 Code of Federal Regulations (CFR) Part 136, Appendix B. The AL reports uncensored concentrations as low as the method reporting limit (RL), which is 0.01 ppbv for all four olefins, for canister concentrations that are less than the LODs. The RL is the value below which the instrument is not capable of measuring and reporting a value (i.e., not detected). A value of one-half the RL (0.005 ppbv) was substituted in lieu of results that were less than the RL for statistical calculations and for the calculation of annual averages; this is a simple substitution method that is a common practice to deal with values that fall below the RL. Since it is not known what the value is below the RL, but it is likely that the value is not zero and equally as likely that the value may not be the RL, a substitution of one-half the RL is conservatively applied. There is no RL for AutoGC data as the AutoGC reports concentrations all the way down to zero. Therefore, for the AutoGC analysis, a value of one-half the LOD was substituted for zeros in the calculation of annual averages and statistics. Information was not readily available on the LODs and RLs for the industry-sponsored data. 2.3. Statistical trend analysis A trend analysis of annual average air concentrations for each monitoring site from 2002 to 2012 was conducted using the Spearman’s rank correlation coefficient. The significance of this correlation was determined by comparing the coefficient to critical values provided by [6]. If the absolute value of the calculated coefficient was larger than the critical value for a one-tailed test (level of significance a of 0.050), then the correlation is significant (p < 0.050). 2.4. Emissions data Texas-specific industrial emissions were obtained from the EI. This data is estimated by the individual industrial sites and reported directly to the TCEQ in tons per year (TPY), so the data can be organized by highest reported estimated emissions by industrial company/site. Data from all of the companies/sites were collected for the most recent completed year, 2012, and from 2002, if available (i.e., the company was established/active), for comparison. The difference in reported estimated emissions from 2002 to 2012 was calculated and given as a percent change for the total state of Texas. For each of the olefins examined, the 2012 emissions data were separated by site and combined by region to determine the highest reported estimated emission areas. It was then compared to the previously described TCEQ monitoring data to see if correlations existed between the two data sets. Finally, Texas emissions were compared to the national emissions using the EPA’s TRI data, which was obtained from EPA’s website. Similar to the EI data, TRI data is estimated by the individual industrial sites and reported directly to the EPA, but the units are in pounds per year. 2.5. Geographical analysis Spatial analysis of monitor and industry geographical locations was done using the ArcGIS 10.1 geographical information system software developed by Esri. A layer of ambient air monitors was created using decimal degree latitude and longitude coordinates obtained from the TCEQ TAMIS database. The TCEQ Air Permitting Division maintains a current layer of property boundaries for entities seeking an air authorization from the Agency. This layer was used to identify industries via the Registered Identification Number (RN) obtained from the EI data. A one-meter statewide orthoimagery for Texas, maintained by the National
3
Agriculture Imagery Program, was used as the base layer. Identification of total TPY of emission sources near monitoring sites was done by creating a two-mile proximity buffer around each site. The industry boundary data was then extracted by using the ArcGIS 10.1 clipping tool to identify only the industries within the monitor proximity buffers. The industrial sources within that two-mile proximity buffer were then compared to the EI data described in Section 2.4. The total TPY of emissions around each monitor were then summed. The same method was used to identify the ambient monitoring sites located near the identified industries. A two-mile proximity buffer was created around the boundaries of each identified industry. The monitoring site data were then extracted by clipping to the industry proximity buffers. Monitoring sites that were located within the two-mile proximity buffer were then identified.
3. Results 3.1. 1,3-Butadiene In 2012, annual average air concentrations of 1,3-butadiene, as measured at 81 air toxics monitors across Texas, ranged from not detected to 1.28 ppbv; the highest annual average was at the Port Neches monitoring site, which is located in TCEQ Region 10Beaumont. The long-term AMCV for 1,3-butadiene of 9.1 ppbv, set to protect against chronic, life-time exposure [18], is 7 times greater than the highest 2012 annual average. Monitored concentrations and statistical summary data from the top five monitoring sites for 1,3-butadiene are listed in Table 1. 1,3-Butadiene was detected above the LOD in 16–72% of canister samples from this analysis (AutoGC data does not report the percent above the LOD). This is a measure of how often the measured concentration is reliably measured in the sample. The sample completeness, ranging from 84% to 100% for 1,3-butadiene, is a measure of the likelihood that the calculated average reflects ambient conditions throughout the year. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1A. In 2002, the monitoring site with the highest annual average of 1,3-butadiene was Milby Park (TCEQ Region 12-Houston). Monitored levels at this site have significantly decreased (p < 0.01) over the past 11 years; the annual average in 2002 was of 2.12 ppbv, which has reduced to an annual average of 0.62 ppbv in 2012. Annual averages have also decreased at the Channelview monitoring site (p < 0.01), although not as dramatically as seen in Milby Park. In 2012, the highest annual average was at the Port Neches monitoring site, which was increased compared to previous years (0.76 ppbv in 2002 and 1.28 ppbv in 2012), although no significant trend was observed. An increase was also seen in the second highest annual average at the Groves monitoring site (TCEQ Region 10Beaumont), where annual averages increased from 0.11 ppbv in 2002 to 0.80 ppbv in 2012 and a significant trend was observed (p < 0.05). It is important to note that all of these annual averages are still well below the long-term AMCV of 9.1 ppbv, and are not expected to cause adverse health effects. Fig. 2A shows the location and wind directions for the two highest monitoring sites for 1,3-butadiene: Port Neches and Groves. These two monitoring sites are located within approximately 3.7 miles of each other. In the figure, the yellow shaded areas represent the property boundaries for registered facilities in the surrounding area and the orange circle represents a two-mile radius around the identified monitoring site. For the Port Neches monitoring site, a total of 19.7 TPY of 1,3-butadiene are reported to be emitted within the two mile radius. This is 4.8% of the total 2012 1,3-butadiene emissions in Texas. For the Groves monitoring site, a total of 53.5 TPY are reported to be emitted within the
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
4
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
Table 1 Monitoring data from the top 5 monitors for each of the evaluated olefins. Chemical
Site
Region
Mean ± SD (ppb)
Median (ppb)
High 24 h (ppb)
% >LODa
Sample comp.b (%)
1,3-Butadiene
Port Neches Groves Milby Park⁄ Manchester/Central Channelview⁄
10 10 12 12 12
Beaumont Beaumont Houston Houston Houston
1.28 ± 1.62 0.80 ± 1.51 0.62 ± 1.90 0.43 ± 1.70 0.31 ± 0.90
0.62 0.12 – 0.09 –
9.08 7.16 9.37 12.06 3.97
72 41 – 16 –
100 97 87 84 77
Ethylene
Dona Park Baytown Laredo Bridge Lynchburg Ferry Jacinto Port
14 12 16 12 12
Corpus Christi Houston Laredo Houston Houston
5.77 ± 8.81 3.96 ± 2.18 3.70 ± 1.31 3.51 ± 2.01 3.44 ± 2.52
2.82 3.77 3.62 3.61 2.72
51.79 9.24 6.19 8.10 12.74
89 100 100 100 100
90 79 79 90 89
Isoprene
Karnack Longview Danciger⁄ HRM #3 Haden Rd Flower Mound⁄
5 Tyler 5 Tyler 12 Houston 12 Houston 4 Dallas
0.56 ± 0.76 0.47 ± 0.59 0.43 ± 0.80 0.38 ± 0.45 0.35 ± 0.90
0.19 0.13 – 0.17 –
3.42 2.43 3.74 1.76 2.68
47 45 – 38 –
98 95 83 98 83
Propylene
HRM 7c Baytown HRM 8c HRM #3 Haden Rd Groves
12 12 12 12 10
7.90 5.78 ± 7.17 3.33 3.17 ± 6.53 3.01 ± 3.04
5.34 4.51 2.05 1.24 2.10
113 47.56 14.3 32.85 17.32
100 94 100 83 97
98 79 88 98 97
Houston Houston Houston Houston Beaumont
Data based on hourly AutoGC data⁄ or every sixth-day 24-h canister data. Some information not provided for AutoGC and industry-sponsored network data. a LOD – limit of detection (this measurement is currently showing up as method detection limit (MDL) in TAMIS web-based reports). b Sample completeness, 75% of possible samples is considered complete. c Industry-sponsored network monitor.
Fig. 1. Monitoring trends over the last 11 years for the top five monitoring sites for each of the evaluated olefins. Monitoring data for (A) 1,3-butadiene, (B) ethylene, (C) isoprene, and (D) propyleneare based on hourly AutoGC data⁄ or every sixth-day 24-h canister data. # – data not complete (less than 75% of possible samples).
two-mile radius, 13.1% of the total 1,3-butadiene emissions in Texas. Although Port Neches reported a 2012 annual average for 1,3-butadiene that was just over 1.5 times greater than Groves (1.28 and 0.8 ppbv, respectively), the reported emissions around Groves are twice as high as those around Port Neches (53.5 and 19.7 TPY, respectively). This illustrates how monitored concentrations do not necessarily correlate with reported emissions, in part because a monitor takes into account all sources of a chemical, and industrial emissions may not be the primary source. The total emissions of 1,3-butadiene for Texas in 2012, as reported to the EI, was 406.9 TPY. This is a decrease of 40% from 2002, where state-wide emissions were reported to be 675.3 TPY. The majority
of emissions are reported from Region 12-Houston (242.0 TPY) and Region 10-Beaumont (131.2 TPY). 3.2. Ethylene The 2012 annual average air concentrations of ethylene measured across the state ranged from not detected to 5.77 ppbv; the highest was at the Dona Park monitoring site, which is located in TCEQ Region 14-Corpus Christi. The lowest TCEQ long-term AMCV for ethylene is 30 ppbv, set to protect against vegetation effects [18], which is over 5 times greater than the highest 2012 annual average. The health-based long-term AMCV for ethylene
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
5
Fig. 2. Monitor locations and wind directions for some of the top monitors for the evaluated olefins. Monitors represent the highest site(s) for 1,3-butadiene (Port Neches and Groves), ethylene (Dona Park and Baytown), isoprene (Karnack) and propylene (HRM 7 and Baytown). The green dot is the monitor location and the orange circle represents a two mile radius around the monitor. The yellow shaded areas represent the property boundaries for registered facilities in the area. The wind rose is an average of the wind conditions recorded at that monitoring site over 2012 and is represented by the radial color blocks emanating from the monitoring site: the wind speed is represented by the different colors (green is calm, yellow is 1 mph, aqua is 4 mph, red is 8 mph, fuchsia is 13 mph, blue is 19 mph, and light green is 24 mph), the arms represent the direction that the wind was coming from (i.e., if an arm is point straight up, the wind was coming from the north), the radius lines represent the percentage of time over the year that wind was coming from that direction (the first band is 6%, the second band is 12%, the third band is 18%, and the fourth band is 24%). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
is set at 5300 ppbv, which is 900 times greater than the highest 2012 annual average. Monitored concentrations and statistical summary data from the top five monitoring sites for ethylene are listed in Table 1. For the top five monitoring sites, ethylene was detected above the LOD in 89–100% of the samples, and the sample completeness ranged from 79% to 90%. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1B. In 2012, the highest annual average was at the Dona Park monitor. There was a slight increase compared to previous years (4.56 ppbv in 2002 and 5.77 ppbv in 2012) although it was not significant. A decrease was observed at Baytown (TCEQ Region12Houston) (5.66 ppbv in 2002 and 3.96 ppbv in 2012), but the overall trend was not significant. Significant decreasing trends were observed at Laredo Bridge (TCEQ Region 16-Laredo) (7.08 ppbv in 2002 and 3.70 ppbv in 2012), and Lynchburg Ferry (TCEQ Region 12-Houston) (6.98 ppbv in 2002 and 3.51 ppbv in 2012), with p < 0.01 and p < 0.05, respectively. Once again, all of these monitored values are still well below the vegetation-based long-term AMCV for ethylene of 30 ppbv, and are not expected to cause adverse vegetation or health effects. Interestingly, the third highest monitor for ethylene is located at Laredo Bridge, an area that does not have a large industrial presence or reported industrial emissions for any of the evaluated olefins, which indicates this site may be more representative of vehicular emissions. The location and wind directions for the two highest monitoring sites for ethylene, Dona Park and Baytown, are shown in Fig. 2B and C. Around the Dona Park monitor, the highest monitor for
ethylene, 9.75 TPY are reported to be emitted within a two mile radius. This is only 0.2% of the total 2012 ethylene emissions in Texas. For Baytown, the second highest monitor, 114.3 TPY are reported to be emitted, 2.4% of the total ethylene emissions in Texas. Although the percentage of emissions is still low for the Baytown area, they are 12 times higher than emissions reported in the area surrounding Dona Park, the highest monitor for ethylene in 2012. As mentioned previously, this is an example of how monitored concentrations do not necessarily correlate with reported emissions, in part because a monitor takes into account all sources of a chemical. The total emissions of ethylene for Texas in 2012, as reported to the EI, were 4766.2 TPY. This is a decrease of just over 30% from 2002, where state-wide emissions were reported to be 6999.5 TPY. The majority of emissions are reported from Region 12-Houston (1677.4 TPY), Region 5-Tyler (1318.4 TPY), Region 10-Beaumont (894.7 TPY), and Region 14-Corpus Christi (855.1 TPY). 3.3. Isoprene The 2012 annual average air concentrations of isoprene measured across the state ranged from not detected to 0.56 ppbv; the highest annual average was at the Karnack monitoring site, which is located in TCEQ Region 5-Tyler. The long-term AMCV for isoprene is 2 ppbv, which is 3.5 times greater than the highest 2012 annual average. Isoprene is a unique olefin in that natural sources account for most of its emissions into the atmosphere.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
6
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
This can be demonstrated by looking at the location of the monitoring sites with the highest annual averages, which are located in TCEQ Region 5-Tyler, an area with very few industrial sources but large areas of coniferous and hardwood forests. Monitored concentrations and statistical summary data from the top five monitoring sites for isoprene are listed in Table 1. For the top five monitoring sites, isoprene was detected above the LOD in 38–47% of the canister samples, the lowest percent detected of all the evaluated olefins, and the sample completeness ranged from 83% to 98%. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1C. Karnack (TCEQ Region 5-Tyler) does not show a significant trend over the last eleven years, while Longview (TCEQ Region 5-Tyler) shows a slight but significant increasing trend (p < 0.05). Danciger (TCEQ Region 12-Houston) shows a decreasing trend (p < 0.05), although the values from 2002 (0.46 ppbv) to 2012 (0.43 ppbv) do not differ significantly. The total emissions of isoprene for Texas in 2012, as reported to the EI, were 77.9 TPY. This is a decrease of over 60% from 2002, where state-wide emissions were reported to be 211.6 TPY. The majority of isoprene emissions are reported from TCEQ Region 10-Beaumont (40.4 TPY) and TCEQ Region 12-Houston (37.2 TPY). Although the two highest monitors are located in TCEQ Region 5-Tyler, the total reported emissions from that area are 0.0001 TPY, which indicates that there is likely a natural source of isoprene in the region. A map showing a two-mile radius around the Karnack monitor shows there are no EI sources in the area (Fig. 2D), with EI data reporting 0 TPY. The Longview monitor shows a similar lack of sources and reported emissions within a two-mile radius (data not shown).
3.5. Further analysis of possible nonindustrial olefin sources The highest 2012 annual average for isoprene was measured at the Karnack monitoring site (Fig. 1). The Karnack site is located in an area surrounded by large areas of coniferous and hardwood forests (Fig. 2D). There are no identified olefin EI industries located within two miles of the Karnack site, therefore there are no reported emissions for any of the evaluated olefins. Because of the lack of industrial olefin sources, the measured olefin concentrations in the ambient air around this monitor may represent natural vegetation emissions of olefins. A VOC canister sampler was activated at the Karnack monitoring site in 2004, which means that olefin monitoring data are only available at this site from 2004 to 2012 (9 years). The 9-year olefin averages (and the percent detected over the LOD) are as follows: butadiene 0.007 ppbv (0–1.6%); ethylene 0.63 ppbv (37–75%); isoprene 0.61 ppbv (35–47%); and propylene 0.19 ppbv (0–10%). 1,3-Butadiene and propylene were rarely detected above the LOD. The third highest 2012 annual average concentration for ethylene was measured at the Laredo Bridge monitoring site (Fig. 1). Laredo Bridge is located near substantial mobile source emissions, however there are no olefin emissions reported within two miles of this site (i.e., EI olefin data are zero). Since there are no EI olefin data near this site, measured olefin concentrations at this site may be representative of mobile source emissions. The 10-year olefin averages (and the percent detected over the LOD) are as follows: ethylene 3.9 ppbv (96–100%); 1,3-butadiene 0.18 ppbv (6.5–55%); isoprene 0.16 ppbv (4.3–24%); and propylene 1.6 ppbv (87–100%).
3.6. Top five reported emission sources for each evaluated olefin 3.4. Propylene The 2012 annual average air concentrations of propylene measured across the state ranged from not detected to 7.9 ppbv; the highest annual average was at the HRM 7 monitoring site, which is located in TCEQ Region 12-Houston. The TCEQ classifies propylene as a simple asphyxiant; toxicity is due to it displacing air, lowering the partial pressure of oxygen, and causing hypoxia at sufficiently high concentrations. Therefore, AMCVs are not derived for propylene. Monitor concentrations and measurement data from the top five monitoring sites for propylene are listed in Table 1. For the top five monitoring sites, propylene was detected above the LOD in 83–100% of the canister samples, and the sample completeness ranged from 79% to 98%. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1D. Baytown (TCEQ Region-12 Houston), HRM #3 (TCEQ Region-12 Houston), and Groves (TCEQ Region-12 Houston) all show a decreasing trend for propylene over the last eleven years (p < 0.01). HRM 8, however, has been steadily increasing since 2002 (p < 0.05). The location and wind directions for the two highest monitors for propylene, HRM 7 and Baytown, are shown in Fig. 2C. Around the HRM 7 and Baytown monitoring sites, 189 TPY are reported to be emitted, 8.7% of the total propylene emissions in Texas. These two monitors are approximately one mile apart from each other; it is not unexpected that they have the same reported propylene emissions within their two-mile radiuses. HRM 7, however, is located closer to other industrial facilities, which may explain why the propylene concentrations at this monitoring site are slightly higher than at the Baytown monitoring site. The total emissions of propylene in 2012, as reported to the EI, were 2167.6 TPY. This is a decrease of over 30% from 2002, where state-wide emissions were reported to be 3251.4 TPY. The majority of emissions are reported from Region 12-Houston (1294.6 TPY).
Table 2 shows the top five EI sites based on the EI data for each of the evaluated olefins as a percentage of the total olefin emissions across the state. While most of these facilities emit multiple olefins, they are not necessarily ranked in the top five for more than one olefin. However, a few facilities (denoted in bold) are ranked in the top five for more than one olefin (e.g., Equistar Chemicals LP, Channelview is a top emitter for both 1,3-butadiene and isoprene). This table identifies monitoring sites that are located within two miles of the property boundaries of the identified top emitters, which suggests that these facilities could contribute to measured concentrations at that monitor. For most of the top five EI sites, monitors are not located within a two mile radius of the facility property boundary, or if they are, they may be located upwind of the facility. However, the following monitoring sites reported higher annual average concentrations for olefins (Table 1) and are located within two miles of the top five EI sites: Groves (ranked #2 for 1,3-butadiene and #5 for propylene); Channelview (ranked #5 for 1,3-butadiene); HRM 7 (ranked #1 for propylene); HRM 8 (ranked #3 for propylene); and Baytown (ranked # 2 for both propylene and ethylene).
4. Discussion 4.1. Olefin monitoring data Although Texas industries are some of the major emitters of lower olefins in the nation, measured ambient air concentrations across the state for these olefins are in the low ppbv range. Measured concentrations were well below health-based AMCVs for 1,3-butadiene [18,8] and isoprene, and were well below vegetation- and health-based AMCVs for ethylene [19,5]. Propylene does not have an AMCV because it is a simple asphyxiant.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
7
Table 2 Emissions data from the top five emitters for each of the evaluated olefins.
BD – 1,3-butadiene, ET – ethylene, IS – isoprene, PR – propylene. Bolded facilities are in the top five emitters for more than one olefin. * Monitors ranked in the top five for the specific olefin indicated: Groves (#2 for butadiene, #5 for propylene); Channelview (#5 for 1,3-butadiene); HRM 7 (#1 for propylene); HRM 8 (#3 for propylene) and Baytown (# 2 for both propylene and ethylene).
Annual average air concentration trends from 2002 to 2012 for nine monitoring sites were not significantly different, whereas for three monitoring sites increasing trends were noted (i.e., 1,3butadiene at Groves, isoprene at Longview, and propylene at HRM 8). The toxicological significance of these increasing concentrations was minimal since all measured concentrations were well below their relevant AMCVs. Eight monitoring sites had significant decreasing trends of a specific olefin: 1,3-butadiene at Milby Park and Channelview; ethylene at Laredo Bridge and Lynchburg Ferry; isoprene at Danciger; and propylene at Baytown, HRM #3 Haden Rd, and Groves. Most of the sites showing decreasing trends from 2002 to 2012 are located in the Houston–Galveston–Brazoria ozone nonattainment areas. 1,3-Butadiene, propylene, and ethylene have been identified as highly reactive VOCs (HRVOCs) in Harris County, while ethylene and propylene are identified as HRVOCs in Brazoria County. HRVOCs contribute to ground-level ozone formation in these areas [21]. Isoprene has been identified as equally or even more reactive in terms of ozone production, but has not historically been emitted in large quantities by anthropogenic sources. Efforts in the Houston–Galveston–Brazoria ozone nonattainment area to reduce ozone have led to a decrease in HRVOCs. Reasons for this may include the TCEQ HRVOC rules, the TCEQ HRVOC Emission Cap and Trade (HECT) program within Harris County, and increased general awareness of photochemical reactivity in the production of ozone. Improvements in remote sensing of emissions, including Differential Absorption Light Detection and Ranging (LIDAR), HAWK infrared (IR) video cameras and GasFind IR cameras, and their subsequent use have anecdotally led to reductions in fugitive emissions, including olefin fugitive emissions [21]. Grant et al. [7] reported on canister spatial and temporal trends for 1,3-butadiene at Milby Park from 1999 (when the site began operations) to 2003. In 2003, the Milby Park monitoring site reported the highest annual average concentrations in the state
(3.2 ppbv). In 2012, annual average concentrations at Milby Park were 0.62 ppbv (this average is from AutoGC data as the canister sampler was deactivated in 2005, at which time an AutoGC was activated). The TCEQ established the Air Pollutant Watch List (APWL) for areas of the state where air toxics are persistently monitored at elevated concentrations as discussed in [4,20]. Listing and delisting procedures for APWL sites are outlined in the APWL Protocol. Efforts to reduce emissions in APWL areas have contributed to decreases in 1,3-butadiene annual average concentrations in the Milby Park area; this area was removed from the APWL in 2009. It should be noted that personal exposure to olefins or other air toxics cannot be assessed based on the use of ambient air concentrations. A monitor measures a snapshot in time, so these ambient air measurements are not likely to be representative of an individual’s exposure to air toxics. There are several considerations that should be taken into account when attempting to determine personal exposure, including the significant amount of time spent indoors, at home, at work, at school, in the car, and other locations [2,13,14,15].
4.2. Olefin emissions data Estimated EI data for 2012 for Texas were lower when compared to 2002 EI data for all olefins. However, consistent correlations between reported olefin EI estimates and measured ambient olefin concentrations were not observed. Industrial emissions are not the only source of emissions in the environment. All lower olefins are byproducts of combustion; mobile sources and other miscellaneous combustion sources may contribute to ambient concentrations, as well as natural sources (i.e., plants). The Laredo Bridge monitoring site may be representative of mobile source emissions, whereas the Karnack monitoring site in East
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
8
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
Texas may be representative of olefin background vegetation/tree emissions. As seen in Table 2, monitoring sites are not always located close to the top five EI sites for each olefin, and there are several possible reasons for this. Monitoring sites are located based on a specific monitoring objective (e.g., background, population exposure, source oriented, etc.). The TCEQ typically locates more air toxic monitors in, or near, communities that are adjacent to, or nearby, heavily industrialized areas. This allows for the assessment of the impact emissions may have on those residential areas. Several factors are considered when a monitoring site is located, including the magnitude of pollution emissions within a 10-km radius of the potential site, the predominant wind direction, population density, the degree of public concern, traffic patterns in the vicinity, sources of electricity, and security. When selecting a monitoring site, TCEQ also ensures that placement of the monitors follows the requirements specified in 40 CFR Part 58, Appendices D and E. Conflict of interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found in the online version.
Acknowledgements We wish to thank Jill Dickey of the TCEQ for providing the Texas Emissions Inventory data and Heather Reddick of the TCEQ for collecting the EPA TRI data. We also wish to thank the HRM Technical Advisory Committee for providing the olefin statistical summary data for the HRM-sponsored monitoring sites HRM 7, 8, and 16. References [1] F.B. Abeles, H.E. Heggestad, Ethylene: an urban air pollutant, J. Air Pollut. Control Assoc. 23 (1973) 517–521. [2] J.L. Adgate, T.R. Church, A.D. Ryan, G. Ramachandran, A.L. Fredrickson, T.H. Stock, M.T. Morandi, K. Saxton, Outdoor, indoor, and personal exposure to VOCs in children, Environ. Health Perspect. 112 (2004) 1386–1392.
[3] BG Chemie [Berfsgenossenschaft der chemischen Industrie], Toxicological Evaluation: Isoprene No. 105, BG Chemie, Heidelberg, Germany, 2000. [4] T. Capobianco, S.M. Hildebrand, M. Honeycutt, J.S. Lee, D. McCant, R.L. Grant, Impact of three interactive texas state regulatory programs to decrease ambient air toxic levels, J. Air Waste Manage. Assoc. 63 (5) (2013) 507–520. [5] N.K. Erraguntla, R.L. Grant, Health-based and vegetative based effect screening values for ethylene, Chem. Biol. Interact. 2014, submitted for publication, in this issue. [6] T.D. Gauthier, Detecting trends using Spearman’s Rank correlation coefficient, Environ. Forensics 2 (2001) 359–362. [7] R.L. Grant, V. Leopold, D. McCant, M. Honeycutt, Spatial and temporal trend evaluation of ambient air concentrations of 1,3-butadiene and chloroprene in Texas, Chem. Biol. Interact. 166 (2007) 44–51. [8] R.L. Grant, J. Haney, A.L. Curry, M. Honeycutt, Development of a unit risk factor for 1,3-butadiene based on an updated carcinogenic toxicity assessment, Risk Anal. 29 (2009) 1726–1742. [9] Hazardous Substances Data Bank (HSDB), United States National Library of Medicine, (accessed September 23, 2014). [10] H.E. Hurst, Toxicology of 1,3-butadiene, chloroprene, and isoprene, Rev. Environ. Contam. Toxicol. 189 (2007) 131–179. [11] F. Loreto, Emission of isoprenoids by plants: their role in atmospheric chemistry, response to the environment, and biochemical pathways, J Environ Pathol Toxicol Oncol 16 (2–3) (1997) 119–124. [12] R.L. Melnick, R.C. Sills, et al., Inhalation toxicity and carcinogenicity of isoprene in rats and mice: comparisons with 1,3-butadiene, Toxicology 113 (1–3) (1996) 247–252. [13] D.C. Payne-Sturges, T.A. Burke, P. Breysse, M. Diener-West, T.J. Buckley, Personal exposure meets risk assessment: a comparison of measured and modeled exposures and risks in an urban community, Environ. Health Perspect. 112 (2004) 589–598. [14] K. Sexton, J.L. Adgate, G. Ramachandran, G.C. Pratt, S.J. Mongin, T.H. Stock, M.T. Morandi, Comparison of personal, indoor, and outdoor exposures to hazardous air pollutants in three urban communities, Environ. Sci. Technol. 38 (2004) 423–430. [15] K. Sexton, J.L. Adgate, G. Ramachandran, G.C. Pratt, S.J. Mongin, T.H. Stock, M.T. Morandi, Evaluating differences between measured personal exposures to volatile organic compounds and concentrations in outdoor and indoor air, Environ. Sci. Technol. 38 (2004) 2593–2602. [16] T.D. Sharkey, Isoprene synthesis by plants and animals, Endeavour 20 (2) (1996) 74–78. [17] J. Song, W. Vizuete, et al., Comparison of observed and modeled isoprene concentrations in southeast Texas during the Texas Air Quality Study, 2005. [18] Texas Commission on Environmental Quality (TCEQ), Development support document 1,3-butadiene CAS registry number: 106-99-0, Austin, TX, , 2008a (accessed October 9, 2014). [19] Texas Commission on Environmental Quality (TCEQ), Development support document ethylene CAS registry number: 74-85-1, Austin, TX, , 2008b (accessed October 9, 2014). [20] Texas Commission on Environmental Quality (TCEQ), Protocol for notification and work group functions for evaluating potential and active Air Pollutant Watch List (APWL) areas, Austin, TX, , 2012 (accessed October 9, 2014). [21] R. Thomas, J. Smith, M. Jones, J. MacKay, J. Jarvie, Emissions modeling of specific highly reactive volatile organic compounds (HRVOC) in the Houston– Galveston–Brazoria ozone nonattainment area, in: 17th International Emissions Inventory Conference, Portland, OR, June 2–5, 2008, , 2008 (accessed September 29, 2014).
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
Contents lists available at ScienceDirect
Chemico-Biological Interactions journal homepage: www.elsevier.com/locate/chembioint
Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012 Jessica L. Myers, Tracie Phillips, Roberta L. Grant ⇑ Texas Commission on Environmental Quality, Toxicology Division, Austin, TX, USA
a r t i c l e
i n f o
Article history: Available online xxxx Keywords: Lower olefins Monitoring Butadiene Ethylene Isoprene Propylene
a b s t r a c t Texas has the largest ambient air monitoring network in the country with approximately 83 monitoring sites that measure ambient air concentrations of volatile organic compounds (VOCs). The lower olefins, including 1,3-butadiene, ethylene, isoprene, and propylene, are a group of VOCs that can be measured in both 24 h/every sixth-day canister samples and continuous 1-h Automated Gas Chromatography (AutoGC) samples. Based on 2012 Toxics Release Inventory data, the total reported industrial air emissions in Texas for these olefins, as compared to total national reported air emissions, were 79% for 1,3butadiene, 62% for ethylene, 76% for isoprene, and 54% for propylene, illustrating that Texas industries are some of the major emitters for these olefins. The purpose of this study was to look at the patterns of annual average air monitoring data from 2002 to 2012 using Texas Commission on Environmental Quality (TCEQ) data for these four lower olefins. It should be emphasized that monitors may not be located close to or downwind of the highest emitters of these lower olefins. In addition, air monitors only provide a snapshot in time of air concentrations for their respective locations, and may not be able to discriminate emissions between specific sources. In 2012, the highest annual average air concentration for 1,3-butadiene was 1.28 ppb by volume (ppbv), which was measured at the Port Neches monitoring site in Region 10-Beaumont. For ethylene, the highest 2012 annual average air concentration was 5.77 ppbv, which was measured at the Dona Park monitoring site in TCEQ Region 14-Corpus Christi. Although reported industrial emissions of isoprene are predominantly from the Houston and Beaumont regions, trees are natural emitters of isoprene, and the highest ambient air concentrations tend to be from regions with large areas of coniferous and hardwood forests. This was observed with TCEQ Region 5-Tyler, which had the two highest isoprene annual average air concentrations for 2012: 0.56 ppbv at the Karnack monitoring site and 0.47 ppbv at the Longview monitoring site. For propylene, the highest 2012 annual average air concentration was recorded at the HRM 7 monitoring site in TCEQ Region 12-Houston, which was 7.9 ppbv. A significant portion of the total 2012 industrial propylene emissions were also reported in TCEQ Region 12Houston. Although some individual monitors showed increased annual averages from 2002 to 2012, there was a general decreasing trend present across the state for all four lower olefins examined. The annual average air concentrations of the four lower olefins were well below their respective Air Monitoring Comparison Values (AMCVs) and are not expected to cause long-term or chronic adverse health effects. Published by Elsevier Ireland Ltd.
Abbreviations: AL, Air Laboratory; AMCV, Air Monitoring Comparison Value; APWL, Air Pollutant Watch List; AutoGC, Automated Gas Chromatography; CFR, Code of Federal Regulation; EI, Emissions Inventory; EPA, Environmental Protection Agency; HECT, HRVOC Emission Cap and Trade; HRM, Houston Regional Monitoring; HRVOC, highly reactive VOC; LIDAR, Differential Absorption Light Detection and Ranging; LOD, limit of detection; MDL, method detection limit; ppbv, parts per billion by volume; RL, reporting limit; RN, Registered Identification Number; TAMIS, Texas Air Monitoring Information System; TCEQ, Texas Commission on Environmental Quality; TPY, tons per year; TRI, Toxics Release Inventory; VOC, volatile organic compound. ⇑ Corresponding author at: Texas Commission in Environmental Quality, P.O. Box 13087 (MC-168), Austin, TX 78711-3087, USA. E-mail address: [email protected] (R.L. Grant).
1. Introduction 1.1. Industrial use of the lower olefins in Texas Texas industries include some of the major emitters of 1,3-butadiene, ethylene, isoprene, and propylene in the nation. Based on the Environmental Protection Agency’s (EPA’s) 2012 Toxics Release Inventory (TRI) data, the total reported industrial air emissions in Texas for these olefins, as compared to total national reported air emissions, were 79% for 1,3-butadiene, 62% for ethylene, 76% for isoprene, and 54% for propylene.
http://dx.doi.org/10.1016/j.cbi.2015.02.008 0009-2797/Published by Elsevier Ireland Ltd.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
2
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
1,3-Butadiene is used as an intermediate to produce a variety of industrial chemicals (e.g., polymers, elastomers), but its major uses are in the manufacturing of styrene-butadiene rubber, or synthetic rubber, and thermoplastic resins. The primary way 1,3-butadiene is released into the environment is from emissions from gasolineand diesel-powered vehicles and equipment [18]. Ethylene is produced through both anthropogenic and natural activities [19]. It is a high-production-volume chemical product of the petrochemical industry, and is primarily used as an intermediate in the production of other chemicals. As a natural compound, ethylene is a plant hormone that enhances the ripening process of fruits, vegetables, and flowers. Some microbes and higher plants naturally produce ethylene. Ambient ethylene is also produced during incomplete combustion of biomass and fossil fuels, forest fires, and active volcanic events. A relatively large proportion of ethylene in urban air is due to vehicular traffic emissions [1]. Isoprene, the 2-methyl analogue of 1,3-butadiene, is also produced through both anthropogenic and natural activities. It is used in the manufacturing of synthetic rubber, styrene–isoprene-styrene block co-polymers and butyl rubber, and in the production of hydrocarbon resins and synthesis of terpenes [3,12,16]. It can also be released in oil fires, tobacco smoke, and automobile exhaust [10,16]. Isoprene is produced naturally by living organisms (flora and fauna), and the amount of isoprene produced and emitted by biological systems far exceeds that which is produced synthetically [10,17]. There are over 200 different plant species that emit isoprene, with trees being a large emitter [11]. Propylene is produced and consumed in refinery operations for the production of components of gasoline, and is also a chemical intermediate in the manufacture of various chemicals (e.g., acetone, isopropylbenzene, isopropanol, propylene oxide, etc.). It is used in polymerized form as polypropylene for plastics and carpet fibers and as an aerosol propellant [9]. As with isoprene and ethylene, propylene is emitted by certain tree species, and produced by other types of vegetation. Propylene is often found in air samples in industrialized areas and is also a combustion product present in cigarette smoke, aircraft exhaust, and motor vehicle exhaust.
1.2. Texas ambient air toxics monitoring network Texas has one of the most extensive ambient air toxics monitoring networks in the country. With approximately 100 air toxics monitoring sites across the state, the Texas Commission on Environmental Quality (TCEQ) analyzes for approximately 120 of the air toxics compounds, as identified by the 1990 EPA Clean Air Act. These toxics include volatile organic compounds (VOCs), semivolatile organics, carbonyls, and metals. The TCEQ has extensive ambient air monitoring data for 1,3-butadiene, ethylene, isoprene, and propylene across the state for numerous years. With a state that covers a large area, in order to better provide services to the citizens and regulated community, the TCEQ has broken down the state into four regional areas, which are further broken down into 16 individual regions. The monitoring sites that the TCEQ oversees are located within these regions; these monitors include stationary monitors that belong to various entities (state, local governments, councils of governments, and industrysponsored). There are two primary types of stationary ambient air monitors that measure VOCs in the state: canister samplers that collect twenty-four-hour samples every sixth-day, and Automated Gas Chromatography (AutoGC) samplers that collect 40 min samples every hour (referred to as a one-hour sample). It should be emphasized that monitors only provide a snapshot in time of ambient air concentrations at their respective locations.
1.3. Spatial and temporal trends of the identified lower olefins The purpose of this study was to evaluate the spatial and temporal trends of the identified lower olefins (1,3-butadiene, ethylene, isoprene, and propylene) across the state of Texas for 2002 through 2012. Chloroprene was not included in this analysis since there are no available monitoring data. The five monitoring sites measuring the highest concentrations in 2012 for each olefin were identified (hereafter referred to as the top five monitoring sites). Trend data of annual average ambient air concentrations for each olefin from 2002 to 2012 for the top five monitoring sites are presented. A further analysis of the ambient air monitors with the highest olefin annual average concentrations was then conducted. The five industrial sources that had the highest 2012 atmospheric emissions in Texas for each olefin were also identified through the Texas Emissions Inventory (EI) (hereafter referred to as the top five EI sites). Monitoring sites located within two miles of the top five EI sites property boundaries were also identified. It should be noted that this study does not discriminate emissions between specific industries, as ambient air monitoring data cannot provide that type of information. 2. Material and methods 2.1. Air monitoring data Air monitoring data (24-h canister and 1-h AutoGC samples) were acquired from either the TCEQ air monitoring databases (e.g., the Texas Air Monitoring Information System (TAMIS)) or from an industry-sponsored monitoring network. Due to proximity to the identified industries in this evaluation, data for three monitoring sites were requested from the Houston Regional Monitoring (HRM) Network and included in this evaluation: two canister sites – HRM 7 and HMR 8; and one AutoGC site – HRM 16. The canister samples are collected continuously for 24 h in silica-coated stainless steel canisters every sixth-day, and then sent off to be analyzed by gas chromatography/mass spectroscopy by the TCEQ Austin Air Laboratory (AL) using modified EPA Method TO-15. The AutoGC samples are collected continuously for 40 min every hour and analyzed automatically on-site using modified EPA Method TO-14. For 2012, there were 81 monitors in the TCEQ databases and 3 monitors from an industry-sponsored monitoring network that had data available on the identified olefins. Data from 2002 through 2012 (where available) were also collected for each of the four identified chemicals from the top five monitoring sites to visualize trends over the last eleven years. 2.2. Analysis of the air monitoring data For each of the monitoring sites examined, the data was considered complete when 75% of the total possible samples were collected and validated for the calendar year. For the 24-h canister data collected every sixth-day, 75% completeness is represented by at least 45 out of a possible 60 samples each year. For the 1-h AutoGC continuous data, 75% completeness is represented by at least 6570 out of a possible 8760 samples each year. Unless otherwise indicated, data were generally used from monitoring sites if it met 75% completeness. One exception includes the 2002–2012 trend graphs of the top five monitoring sites for each olefin, where incomplete previous years were still used for visualization of the trends (identified by #). Air Monitoring Comparison Values (AMCVs) set by the TCEQ are chemical-specific air concentrations derived to protect human health and welfare, including odor, animals, and vegetation; each data set was compared to its lowest respective AMCV.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
The TCEQ AL determines data quality objectives in terms of the limit of detection (LOD). For all VOCs, the LODs are established using the methods outlined in 40 Code of Federal Regulations (CFR) Part 136, Appendix B. The AL reports uncensored concentrations as low as the method reporting limit (RL), which is 0.01 ppbv for all four olefins, for canister concentrations that are less than the LODs. The RL is the value below which the instrument is not capable of measuring and reporting a value (i.e., not detected). A value of one-half the RL (0.005 ppbv) was substituted in lieu of results that were less than the RL for statistical calculations and for the calculation of annual averages; this is a simple substitution method that is a common practice to deal with values that fall below the RL. Since it is not known what the value is below the RL, but it is likely that the value is not zero and equally as likely that the value may not be the RL, a substitution of one-half the RL is conservatively applied. There is no RL for AutoGC data as the AutoGC reports concentrations all the way down to zero. Therefore, for the AutoGC analysis, a value of one-half the LOD was substituted for zeros in the calculation of annual averages and statistics. Information was not readily available on the LODs and RLs for the industry-sponsored data. 2.3. Statistical trend analysis A trend analysis of annual average air concentrations for each monitoring site from 2002 to 2012 was conducted using the Spearman’s rank correlation coefficient. The significance of this correlation was determined by comparing the coefficient to critical values provided by [6]. If the absolute value of the calculated coefficient was larger than the critical value for a one-tailed test (level of significance a of 0.050), then the correlation is significant (p < 0.050). 2.4. Emissions data Texas-specific industrial emissions were obtained from the EI. This data is estimated by the individual industrial sites and reported directly to the TCEQ in tons per year (TPY), so the data can be organized by highest reported estimated emissions by industrial company/site. Data from all of the companies/sites were collected for the most recent completed year, 2012, and from 2002, if available (i.e., the company was established/active), for comparison. The difference in reported estimated emissions from 2002 to 2012 was calculated and given as a percent change for the total state of Texas. For each of the olefins examined, the 2012 emissions data were separated by site and combined by region to determine the highest reported estimated emission areas. It was then compared to the previously described TCEQ monitoring data to see if correlations existed between the two data sets. Finally, Texas emissions were compared to the national emissions using the EPA’s TRI data, which was obtained from EPA’s website. Similar to the EI data, TRI data is estimated by the individual industrial sites and reported directly to the EPA, but the units are in pounds per year. 2.5. Geographical analysis Spatial analysis of monitor and industry geographical locations was done using the ArcGIS 10.1 geographical information system software developed by Esri. A layer of ambient air monitors was created using decimal degree latitude and longitude coordinates obtained from the TCEQ TAMIS database. The TCEQ Air Permitting Division maintains a current layer of property boundaries for entities seeking an air authorization from the Agency. This layer was used to identify industries via the Registered Identification Number (RN) obtained from the EI data. A one-meter statewide orthoimagery for Texas, maintained by the National
3
Agriculture Imagery Program, was used as the base layer. Identification of total TPY of emission sources near monitoring sites was done by creating a two-mile proximity buffer around each site. The industry boundary data was then extracted by using the ArcGIS 10.1 clipping tool to identify only the industries within the monitor proximity buffers. The industrial sources within that two-mile proximity buffer were then compared to the EI data described in Section 2.4. The total TPY of emissions around each monitor were then summed. The same method was used to identify the ambient monitoring sites located near the identified industries. A two-mile proximity buffer was created around the boundaries of each identified industry. The monitoring site data were then extracted by clipping to the industry proximity buffers. Monitoring sites that were located within the two-mile proximity buffer were then identified.
3. Results 3.1. 1,3-Butadiene In 2012, annual average air concentrations of 1,3-butadiene, as measured at 81 air toxics monitors across Texas, ranged from not detected to 1.28 ppbv; the highest annual average was at the Port Neches monitoring site, which is located in TCEQ Region 10Beaumont. The long-term AMCV for 1,3-butadiene of 9.1 ppbv, set to protect against chronic, life-time exposure [18], is 7 times greater than the highest 2012 annual average. Monitored concentrations and statistical summary data from the top five monitoring sites for 1,3-butadiene are listed in Table 1. 1,3-Butadiene was detected above the LOD in 16–72% of canister samples from this analysis (AutoGC data does not report the percent above the LOD). This is a measure of how often the measured concentration is reliably measured in the sample. The sample completeness, ranging from 84% to 100% for 1,3-butadiene, is a measure of the likelihood that the calculated average reflects ambient conditions throughout the year. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1A. In 2002, the monitoring site with the highest annual average of 1,3-butadiene was Milby Park (TCEQ Region 12-Houston). Monitored levels at this site have significantly decreased (p < 0.01) over the past 11 years; the annual average in 2002 was of 2.12 ppbv, which has reduced to an annual average of 0.62 ppbv in 2012. Annual averages have also decreased at the Channelview monitoring site (p < 0.01), although not as dramatically as seen in Milby Park. In 2012, the highest annual average was at the Port Neches monitoring site, which was increased compared to previous years (0.76 ppbv in 2002 and 1.28 ppbv in 2012), although no significant trend was observed. An increase was also seen in the second highest annual average at the Groves monitoring site (TCEQ Region 10Beaumont), where annual averages increased from 0.11 ppbv in 2002 to 0.80 ppbv in 2012 and a significant trend was observed (p < 0.05). It is important to note that all of these annual averages are still well below the long-term AMCV of 9.1 ppbv, and are not expected to cause adverse health effects. Fig. 2A shows the location and wind directions for the two highest monitoring sites for 1,3-butadiene: Port Neches and Groves. These two monitoring sites are located within approximately 3.7 miles of each other. In the figure, the yellow shaded areas represent the property boundaries for registered facilities in the surrounding area and the orange circle represents a two-mile radius around the identified monitoring site. For the Port Neches monitoring site, a total of 19.7 TPY of 1,3-butadiene are reported to be emitted within the two mile radius. This is 4.8% of the total 2012 1,3-butadiene emissions in Texas. For the Groves monitoring site, a total of 53.5 TPY are reported to be emitted within the
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
4
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
Table 1 Monitoring data from the top 5 monitors for each of the evaluated olefins. Chemical
Site
Region
Mean ± SD (ppb)
Median (ppb)
High 24 h (ppb)
% >LODa
Sample comp.b (%)
1,3-Butadiene
Port Neches Groves Milby Park⁄ Manchester/Central Channelview⁄
10 10 12 12 12
Beaumont Beaumont Houston Houston Houston
1.28 ± 1.62 0.80 ± 1.51 0.62 ± 1.90 0.43 ± 1.70 0.31 ± 0.90
0.62 0.12 – 0.09 –
9.08 7.16 9.37 12.06 3.97
72 41 – 16 –
100 97 87 84 77
Ethylene
Dona Park Baytown Laredo Bridge Lynchburg Ferry Jacinto Port
14 12 16 12 12
Corpus Christi Houston Laredo Houston Houston
5.77 ± 8.81 3.96 ± 2.18 3.70 ± 1.31 3.51 ± 2.01 3.44 ± 2.52
2.82 3.77 3.62 3.61 2.72
51.79 9.24 6.19 8.10 12.74
89 100 100 100 100
90 79 79 90 89
Isoprene
Karnack Longview Danciger⁄ HRM #3 Haden Rd Flower Mound⁄
5 Tyler 5 Tyler 12 Houston 12 Houston 4 Dallas
0.56 ± 0.76 0.47 ± 0.59 0.43 ± 0.80 0.38 ± 0.45 0.35 ± 0.90
0.19 0.13 – 0.17 –
3.42 2.43 3.74 1.76 2.68
47 45 – 38 –
98 95 83 98 83
Propylene
HRM 7c Baytown HRM 8c HRM #3 Haden Rd Groves
12 12 12 12 10
7.90 5.78 ± 7.17 3.33 3.17 ± 6.53 3.01 ± 3.04
5.34 4.51 2.05 1.24 2.10
113 47.56 14.3 32.85 17.32
100 94 100 83 97
98 79 88 98 97
Houston Houston Houston Houston Beaumont
Data based on hourly AutoGC data⁄ or every sixth-day 24-h canister data. Some information not provided for AutoGC and industry-sponsored network data. a LOD – limit of detection (this measurement is currently showing up as method detection limit (MDL) in TAMIS web-based reports). b Sample completeness, 75% of possible samples is considered complete. c Industry-sponsored network monitor.
Fig. 1. Monitoring trends over the last 11 years for the top five monitoring sites for each of the evaluated olefins. Monitoring data for (A) 1,3-butadiene, (B) ethylene, (C) isoprene, and (D) propyleneare based on hourly AutoGC data⁄ or every sixth-day 24-h canister data. # – data not complete (less than 75% of possible samples).
two-mile radius, 13.1% of the total 1,3-butadiene emissions in Texas. Although Port Neches reported a 2012 annual average for 1,3-butadiene that was just over 1.5 times greater than Groves (1.28 and 0.8 ppbv, respectively), the reported emissions around Groves are twice as high as those around Port Neches (53.5 and 19.7 TPY, respectively). This illustrates how monitored concentrations do not necessarily correlate with reported emissions, in part because a monitor takes into account all sources of a chemical, and industrial emissions may not be the primary source. The total emissions of 1,3-butadiene for Texas in 2012, as reported to the EI, was 406.9 TPY. This is a decrease of 40% from 2002, where state-wide emissions were reported to be 675.3 TPY. The majority
of emissions are reported from Region 12-Houston (242.0 TPY) and Region 10-Beaumont (131.2 TPY). 3.2. Ethylene The 2012 annual average air concentrations of ethylene measured across the state ranged from not detected to 5.77 ppbv; the highest was at the Dona Park monitoring site, which is located in TCEQ Region 14-Corpus Christi. The lowest TCEQ long-term AMCV for ethylene is 30 ppbv, set to protect against vegetation effects [18], which is over 5 times greater than the highest 2012 annual average. The health-based long-term AMCV for ethylene
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
5
Fig. 2. Monitor locations and wind directions for some of the top monitors for the evaluated olefins. Monitors represent the highest site(s) for 1,3-butadiene (Port Neches and Groves), ethylene (Dona Park and Baytown), isoprene (Karnack) and propylene (HRM 7 and Baytown). The green dot is the monitor location and the orange circle represents a two mile radius around the monitor. The yellow shaded areas represent the property boundaries for registered facilities in the area. The wind rose is an average of the wind conditions recorded at that monitoring site over 2012 and is represented by the radial color blocks emanating from the monitoring site: the wind speed is represented by the different colors (green is calm, yellow is 1 mph, aqua is 4 mph, red is 8 mph, fuchsia is 13 mph, blue is 19 mph, and light green is 24 mph), the arms represent the direction that the wind was coming from (i.e., if an arm is point straight up, the wind was coming from the north), the radius lines represent the percentage of time over the year that wind was coming from that direction (the first band is 6%, the second band is 12%, the third band is 18%, and the fourth band is 24%). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
is set at 5300 ppbv, which is 900 times greater than the highest 2012 annual average. Monitored concentrations and statistical summary data from the top five monitoring sites for ethylene are listed in Table 1. For the top five monitoring sites, ethylene was detected above the LOD in 89–100% of the samples, and the sample completeness ranged from 79% to 90%. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1B. In 2012, the highest annual average was at the Dona Park monitor. There was a slight increase compared to previous years (4.56 ppbv in 2002 and 5.77 ppbv in 2012) although it was not significant. A decrease was observed at Baytown (TCEQ Region12Houston) (5.66 ppbv in 2002 and 3.96 ppbv in 2012), but the overall trend was not significant. Significant decreasing trends were observed at Laredo Bridge (TCEQ Region 16-Laredo) (7.08 ppbv in 2002 and 3.70 ppbv in 2012), and Lynchburg Ferry (TCEQ Region 12-Houston) (6.98 ppbv in 2002 and 3.51 ppbv in 2012), with p < 0.01 and p < 0.05, respectively. Once again, all of these monitored values are still well below the vegetation-based long-term AMCV for ethylene of 30 ppbv, and are not expected to cause adverse vegetation or health effects. Interestingly, the third highest monitor for ethylene is located at Laredo Bridge, an area that does not have a large industrial presence or reported industrial emissions for any of the evaluated olefins, which indicates this site may be more representative of vehicular emissions. The location and wind directions for the two highest monitoring sites for ethylene, Dona Park and Baytown, are shown in Fig. 2B and C. Around the Dona Park monitor, the highest monitor for
ethylene, 9.75 TPY are reported to be emitted within a two mile radius. This is only 0.2% of the total 2012 ethylene emissions in Texas. For Baytown, the second highest monitor, 114.3 TPY are reported to be emitted, 2.4% of the total ethylene emissions in Texas. Although the percentage of emissions is still low for the Baytown area, they are 12 times higher than emissions reported in the area surrounding Dona Park, the highest monitor for ethylene in 2012. As mentioned previously, this is an example of how monitored concentrations do not necessarily correlate with reported emissions, in part because a monitor takes into account all sources of a chemical. The total emissions of ethylene for Texas in 2012, as reported to the EI, were 4766.2 TPY. This is a decrease of just over 30% from 2002, where state-wide emissions were reported to be 6999.5 TPY. The majority of emissions are reported from Region 12-Houston (1677.4 TPY), Region 5-Tyler (1318.4 TPY), Region 10-Beaumont (894.7 TPY), and Region 14-Corpus Christi (855.1 TPY). 3.3. Isoprene The 2012 annual average air concentrations of isoprene measured across the state ranged from not detected to 0.56 ppbv; the highest annual average was at the Karnack monitoring site, which is located in TCEQ Region 5-Tyler. The long-term AMCV for isoprene is 2 ppbv, which is 3.5 times greater than the highest 2012 annual average. Isoprene is a unique olefin in that natural sources account for most of its emissions into the atmosphere.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
6
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
This can be demonstrated by looking at the location of the monitoring sites with the highest annual averages, which are located in TCEQ Region 5-Tyler, an area with very few industrial sources but large areas of coniferous and hardwood forests. Monitored concentrations and statistical summary data from the top five monitoring sites for isoprene are listed in Table 1. For the top five monitoring sites, isoprene was detected above the LOD in 38–47% of the canister samples, the lowest percent detected of all the evaluated olefins, and the sample completeness ranged from 83% to 98%. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1C. Karnack (TCEQ Region 5-Tyler) does not show a significant trend over the last eleven years, while Longview (TCEQ Region 5-Tyler) shows a slight but significant increasing trend (p < 0.05). Danciger (TCEQ Region 12-Houston) shows a decreasing trend (p < 0.05), although the values from 2002 (0.46 ppbv) to 2012 (0.43 ppbv) do not differ significantly. The total emissions of isoprene for Texas in 2012, as reported to the EI, were 77.9 TPY. This is a decrease of over 60% from 2002, where state-wide emissions were reported to be 211.6 TPY. The majority of isoprene emissions are reported from TCEQ Region 10-Beaumont (40.4 TPY) and TCEQ Region 12-Houston (37.2 TPY). Although the two highest monitors are located in TCEQ Region 5-Tyler, the total reported emissions from that area are 0.0001 TPY, which indicates that there is likely a natural source of isoprene in the region. A map showing a two-mile radius around the Karnack monitor shows there are no EI sources in the area (Fig. 2D), with EI data reporting 0 TPY. The Longview monitor shows a similar lack of sources and reported emissions within a two-mile radius (data not shown).
3.5. Further analysis of possible nonindustrial olefin sources The highest 2012 annual average for isoprene was measured at the Karnack monitoring site (Fig. 1). The Karnack site is located in an area surrounded by large areas of coniferous and hardwood forests (Fig. 2D). There are no identified olefin EI industries located within two miles of the Karnack site, therefore there are no reported emissions for any of the evaluated olefins. Because of the lack of industrial olefin sources, the measured olefin concentrations in the ambient air around this monitor may represent natural vegetation emissions of olefins. A VOC canister sampler was activated at the Karnack monitoring site in 2004, which means that olefin monitoring data are only available at this site from 2004 to 2012 (9 years). The 9-year olefin averages (and the percent detected over the LOD) are as follows: butadiene 0.007 ppbv (0–1.6%); ethylene 0.63 ppbv (37–75%); isoprene 0.61 ppbv (35–47%); and propylene 0.19 ppbv (0–10%). 1,3-Butadiene and propylene were rarely detected above the LOD. The third highest 2012 annual average concentration for ethylene was measured at the Laredo Bridge monitoring site (Fig. 1). Laredo Bridge is located near substantial mobile source emissions, however there are no olefin emissions reported within two miles of this site (i.e., EI olefin data are zero). Since there are no EI olefin data near this site, measured olefin concentrations at this site may be representative of mobile source emissions. The 10-year olefin averages (and the percent detected over the LOD) are as follows: ethylene 3.9 ppbv (96–100%); 1,3-butadiene 0.18 ppbv (6.5–55%); isoprene 0.16 ppbv (4.3–24%); and propylene 1.6 ppbv (87–100%).
3.6. Top five reported emission sources for each evaluated olefin 3.4. Propylene The 2012 annual average air concentrations of propylene measured across the state ranged from not detected to 7.9 ppbv; the highest annual average was at the HRM 7 monitoring site, which is located in TCEQ Region 12-Houston. The TCEQ classifies propylene as a simple asphyxiant; toxicity is due to it displacing air, lowering the partial pressure of oxygen, and causing hypoxia at sufficiently high concentrations. Therefore, AMCVs are not derived for propylene. Monitor concentrations and measurement data from the top five monitoring sites for propylene are listed in Table 1. For the top five monitoring sites, propylene was detected above the LOD in 83–100% of the canister samples, and the sample completeness ranged from 79% to 98%. The temporal trends for the top five monitoring sites from 2002 to 2012 are presented in Fig. 1D. Baytown (TCEQ Region-12 Houston), HRM #3 (TCEQ Region-12 Houston), and Groves (TCEQ Region-12 Houston) all show a decreasing trend for propylene over the last eleven years (p < 0.01). HRM 8, however, has been steadily increasing since 2002 (p < 0.05). The location and wind directions for the two highest monitors for propylene, HRM 7 and Baytown, are shown in Fig. 2C. Around the HRM 7 and Baytown monitoring sites, 189 TPY are reported to be emitted, 8.7% of the total propylene emissions in Texas. These two monitors are approximately one mile apart from each other; it is not unexpected that they have the same reported propylene emissions within their two-mile radiuses. HRM 7, however, is located closer to other industrial facilities, which may explain why the propylene concentrations at this monitoring site are slightly higher than at the Baytown monitoring site. The total emissions of propylene in 2012, as reported to the EI, were 2167.6 TPY. This is a decrease of over 30% from 2002, where state-wide emissions were reported to be 3251.4 TPY. The majority of emissions are reported from Region 12-Houston (1294.6 TPY).
Table 2 shows the top five EI sites based on the EI data for each of the evaluated olefins as a percentage of the total olefin emissions across the state. While most of these facilities emit multiple olefins, they are not necessarily ranked in the top five for more than one olefin. However, a few facilities (denoted in bold) are ranked in the top five for more than one olefin (e.g., Equistar Chemicals LP, Channelview is a top emitter for both 1,3-butadiene and isoprene). This table identifies monitoring sites that are located within two miles of the property boundaries of the identified top emitters, which suggests that these facilities could contribute to measured concentrations at that monitor. For most of the top five EI sites, monitors are not located within a two mile radius of the facility property boundary, or if they are, they may be located upwind of the facility. However, the following monitoring sites reported higher annual average concentrations for olefins (Table 1) and are located within two miles of the top five EI sites: Groves (ranked #2 for 1,3-butadiene and #5 for propylene); Channelview (ranked #5 for 1,3-butadiene); HRM 7 (ranked #1 for propylene); HRM 8 (ranked #3 for propylene); and Baytown (ranked # 2 for both propylene and ethylene).
4. Discussion 4.1. Olefin monitoring data Although Texas industries are some of the major emitters of lower olefins in the nation, measured ambient air concentrations across the state for these olefins are in the low ppbv range. Measured concentrations were well below health-based AMCVs for 1,3-butadiene [18,8] and isoprene, and were well below vegetation- and health-based AMCVs for ethylene [19,5]. Propylene does not have an AMCV because it is a simple asphyxiant.
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
7
Table 2 Emissions data from the top five emitters for each of the evaluated olefins.
BD – 1,3-butadiene, ET – ethylene, IS – isoprene, PR – propylene. Bolded facilities are in the top five emitters for more than one olefin. * Monitors ranked in the top five for the specific olefin indicated: Groves (#2 for butadiene, #5 for propylene); Channelview (#5 for 1,3-butadiene); HRM 7 (#1 for propylene); HRM 8 (#3 for propylene) and Baytown (# 2 for both propylene and ethylene).
Annual average air concentration trends from 2002 to 2012 for nine monitoring sites were not significantly different, whereas for three monitoring sites increasing trends were noted (i.e., 1,3butadiene at Groves, isoprene at Longview, and propylene at HRM 8). The toxicological significance of these increasing concentrations was minimal since all measured concentrations were well below their relevant AMCVs. Eight monitoring sites had significant decreasing trends of a specific olefin: 1,3-butadiene at Milby Park and Channelview; ethylene at Laredo Bridge and Lynchburg Ferry; isoprene at Danciger; and propylene at Baytown, HRM #3 Haden Rd, and Groves. Most of the sites showing decreasing trends from 2002 to 2012 are located in the Houston–Galveston–Brazoria ozone nonattainment areas. 1,3-Butadiene, propylene, and ethylene have been identified as highly reactive VOCs (HRVOCs) in Harris County, while ethylene and propylene are identified as HRVOCs in Brazoria County. HRVOCs contribute to ground-level ozone formation in these areas [21]. Isoprene has been identified as equally or even more reactive in terms of ozone production, but has not historically been emitted in large quantities by anthropogenic sources. Efforts in the Houston–Galveston–Brazoria ozone nonattainment area to reduce ozone have led to a decrease in HRVOCs. Reasons for this may include the TCEQ HRVOC rules, the TCEQ HRVOC Emission Cap and Trade (HECT) program within Harris County, and increased general awareness of photochemical reactivity in the production of ozone. Improvements in remote sensing of emissions, including Differential Absorption Light Detection and Ranging (LIDAR), HAWK infrared (IR) video cameras and GasFind IR cameras, and their subsequent use have anecdotally led to reductions in fugitive emissions, including olefin fugitive emissions [21]. Grant et al. [7] reported on canister spatial and temporal trends for 1,3-butadiene at Milby Park from 1999 (when the site began operations) to 2003. In 2003, the Milby Park monitoring site reported the highest annual average concentrations in the state
(3.2 ppbv). In 2012, annual average concentrations at Milby Park were 0.62 ppbv (this average is from AutoGC data as the canister sampler was deactivated in 2005, at which time an AutoGC was activated). The TCEQ established the Air Pollutant Watch List (APWL) for areas of the state where air toxics are persistently monitored at elevated concentrations as discussed in [4,20]. Listing and delisting procedures for APWL sites are outlined in the APWL Protocol. Efforts to reduce emissions in APWL areas have contributed to decreases in 1,3-butadiene annual average concentrations in the Milby Park area; this area was removed from the APWL in 2009. It should be noted that personal exposure to olefins or other air toxics cannot be assessed based on the use of ambient air concentrations. A monitor measures a snapshot in time, so these ambient air measurements are not likely to be representative of an individual’s exposure to air toxics. There are several considerations that should be taken into account when attempting to determine personal exposure, including the significant amount of time spent indoors, at home, at work, at school, in the car, and other locations [2,13,14,15].
4.2. Olefin emissions data Estimated EI data for 2012 for Texas were lower when compared to 2002 EI data for all olefins. However, consistent correlations between reported olefin EI estimates and measured ambient olefin concentrations were not observed. Industrial emissions are not the only source of emissions in the environment. All lower olefins are byproducts of combustion; mobile sources and other miscellaneous combustion sources may contribute to ambient concentrations, as well as natural sources (i.e., plants). The Laredo Bridge monitoring site may be representative of mobile source emissions, whereas the Karnack monitoring site in East
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
8
J.L. Myers et al. / Chemico-Biological Interactions xxx (2015) xxx–xxx
Texas may be representative of olefin background vegetation/tree emissions. As seen in Table 2, monitoring sites are not always located close to the top five EI sites for each olefin, and there are several possible reasons for this. Monitoring sites are located based on a specific monitoring objective (e.g., background, population exposure, source oriented, etc.). The TCEQ typically locates more air toxic monitors in, or near, communities that are adjacent to, or nearby, heavily industrialized areas. This allows for the assessment of the impact emissions may have on those residential areas. Several factors are considered when a monitoring site is located, including the magnitude of pollution emissions within a 10-km radius of the potential site, the predominant wind direction, population density, the degree of public concern, traffic patterns in the vicinity, sources of electricity, and security. When selecting a monitoring site, TCEQ also ensures that placement of the monitors follows the requirements specified in 40 CFR Part 58, Appendices D and E. Conflict of interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found in the online version.
Acknowledgements We wish to thank Jill Dickey of the TCEQ for providing the Texas Emissions Inventory data and Heather Reddick of the TCEQ for collecting the EPA TRI data. We also wish to thank the HRM Technical Advisory Committee for providing the olefin statistical summary data for the HRM-sponsored monitoring sites HRM 7, 8, and 16. References [1] F.B. Abeles, H.E. Heggestad, Ethylene: an urban air pollutant, J. Air Pollut. Control Assoc. 23 (1973) 517–521. [2] J.L. Adgate, T.R. Church, A.D. Ryan, G. Ramachandran, A.L. Fredrickson, T.H. Stock, M.T. Morandi, K. Saxton, Outdoor, indoor, and personal exposure to VOCs in children, Environ. Health Perspect. 112 (2004) 1386–1392.
[3] BG Chemie [Berfsgenossenschaft der chemischen Industrie], Toxicological Evaluation: Isoprene No. 105, BG Chemie, Heidelberg, Germany, 2000. [4] T. Capobianco, S.M. Hildebrand, M. Honeycutt, J.S. Lee, D. McCant, R.L. Grant, Impact of three interactive texas state regulatory programs to decrease ambient air toxic levels, J. Air Waste Manage. Assoc. 63 (5) (2013) 507–520. [5] N.K. Erraguntla, R.L. Grant, Health-based and vegetative based effect screening values for ethylene, Chem. Biol. Interact. 2014, submitted for publication, in this issue. [6] T.D. Gauthier, Detecting trends using Spearman’s Rank correlation coefficient, Environ. Forensics 2 (2001) 359–362. [7] R.L. Grant, V. Leopold, D. McCant, M. Honeycutt, Spatial and temporal trend evaluation of ambient air concentrations of 1,3-butadiene and chloroprene in Texas, Chem. Biol. Interact. 166 (2007) 44–51. [8] R.L. Grant, J. Haney, A.L. Curry, M. Honeycutt, Development of a unit risk factor for 1,3-butadiene based on an updated carcinogenic toxicity assessment, Risk Anal. 29 (2009) 1726–1742. [9] Hazardous Substances Data Bank (HSDB), United States National Library of Medicine, (accessed September 23, 2014). [10] H.E. Hurst, Toxicology of 1,3-butadiene, chloroprene, and isoprene, Rev. Environ. Contam. Toxicol. 189 (2007) 131–179. [11] F. Loreto, Emission of isoprenoids by plants: their role in atmospheric chemistry, response to the environment, and biochemical pathways, J Environ Pathol Toxicol Oncol 16 (2–3) (1997) 119–124. [12] R.L. Melnick, R.C. Sills, et al., Inhalation toxicity and carcinogenicity of isoprene in rats and mice: comparisons with 1,3-butadiene, Toxicology 113 (1–3) (1996) 247–252. [13] D.C. Payne-Sturges, T.A. Burke, P. Breysse, M. Diener-West, T.J. Buckley, Personal exposure meets risk assessment: a comparison of measured and modeled exposures and risks in an urban community, Environ. Health Perspect. 112 (2004) 589–598. [14] K. Sexton, J.L. Adgate, G. Ramachandran, G.C. Pratt, S.J. Mongin, T.H. Stock, M.T. Morandi, Comparison of personal, indoor, and outdoor exposures to hazardous air pollutants in three urban communities, Environ. Sci. Technol. 38 (2004) 423–430. [15] K. Sexton, J.L. Adgate, G. Ramachandran, G.C. Pratt, S.J. Mongin, T.H. Stock, M.T. Morandi, Evaluating differences between measured personal exposures to volatile organic compounds and concentrations in outdoor and indoor air, Environ. Sci. Technol. 38 (2004) 2593–2602. [16] T.D. Sharkey, Isoprene synthesis by plants and animals, Endeavour 20 (2) (1996) 74–78. [17] J. Song, W. Vizuete, et al., Comparison of observed and modeled isoprene concentrations in southeast Texas during the Texas Air Quality Study, 2005. [18] Texas Commission on Environmental Quality (TCEQ), Development support document 1,3-butadiene CAS registry number: 106-99-0, Austin, TX, , 2008a (accessed October 9, 2014). [19] Texas Commission on Environmental Quality (TCEQ), Development support document ethylene CAS registry number: 74-85-1, Austin, TX, , 2008b (accessed October 9, 2014). [20] Texas Commission on Environmental Quality (TCEQ), Protocol for notification and work group functions for evaluating potential and active Air Pollutant Watch List (APWL) areas, Austin, TX, , 2012 (accessed October 9, 2014). [21] R. Thomas, J. Smith, M. Jones, J. MacKay, J. Jarvie, Emissions modeling of specific highly reactive volatile organic compounds (HRVOC) in the Houston– Galveston–Brazoria ozone nonattainment area, in: 17th International Emissions Inventory Conference, Portland, OR, June 2–5, 2008, , 2008 (accessed September 29, 2014).
Please cite this article in press as: J.L. Myers et al., Emissions and ambient air monitoring trends of lower olefins across Texas from 2002 to 2012, ChemicoBiological Interactions (2015), http://dx.doi.org/10.1016/j.cbi.2015.02.008
