FY14-15 PGAFY14-1 Task 3.1 CAPITAL AREA COUNCIL OF GOVERNMENTS OZONE MONITORING NETWORK REVIEW AND OPTIMIZATION PLAN Task 3.1, Contract 582-14-40052, PGA FY14-01 Prepared by the Capital Area Council of Governments Air Quality Program April 2014 PREPARED UNDER A GRANT FROM THE TEXAS COMMISSION ON ENVIRONMENTAL QUALITY The preparation of this report was financed through grants from the State of Texas through the Texas Commission on Environmental Quality. The content, findings, opinions, and conclusions are the work of the author(s) and do not necessarily represent findings, opinions, or conclusions of the TCEQ.
Executive Summary This report is designed to enable the Capital Area Council of Governments (CAPCOG) Air Quality Program to optimize its deployment and operation of ozone monitoring assets throughout the region in 2014 and 2015. Based on its review of the goals of CAPCOG s monitoring program and a series of analyses included in this report, CAPCOG s general plan for deployment and operation of ozone monitoring stations in 2014 and 2015 includes: Operate all eight monitoring stations in both 2014 and 2015; Starting the monitoring period for 2015 by March 1; Operate monitoring equipment through the first week of November in both 2014 and 2015; and Add additional quality control or calibration procedures to the monitoring operations for 2014 and 2015. This ozone monitoring network review and optimization plan includes the following: 1. a description of the goals of CAPCOG s air monitoring program and background for this analysis; 2. site-by-site analyses and bottom-up analyses comparing the value of each monitoring station; 3. analyses of the appropriate start and end dates for operating CAPCOG s monitors, data quality needs, and resources available; and 4. quantitative and qualitative ranking of each CAPCOG monitoring station by the analyses contained in this report. The geographic area covered by this analysis is the 10-County CAPCOG region, which consists of the same counties as the Texas Commission on Environmental Quality (TCEQ) Region 11 office: Bastrop, Blanco, Burnet, Caldwell, Fayette, Hays, Lee, Llano, Travis, and Williamson Counties. Within the CAPCOG region, Bastrop, Caldwell, Hays, Travis, and Williamson Counties are particularly important since they would constitute the presumptive boundaries of a nonattainment area if formally designated the region nonattainment for the ozone National Ambient Air Quality Standard (NAAQS), based on the most recent guidance the U.S. Environmental Protection Agency (EPA) has issued on ozone nonattainment designations *. Therefore, most of the analysis in this report relates to these counties, which make up the Austin-Round Rock Metropolitan Statistical Area (Austin-Round Rock MSA). This review relies largely on data and analysis prepared for the monitoring network review completed in February 2013, updated with the monitoring data collected in 2013 and with a more direct understanding of the funding that will be available to conduct ozone monitoring in 2014 and 2015. * Meyers, Robert J.; Principal Deputy Assistant Administrator, U.S. Environmental Protection Agency. Memorandum to Regional Administrators, Regions I-X: Area Designations for the 2008 Revised Ozone National Ambient Air Quality Standards. December 4, 2008. http://www.epa.gov/ozonedesignations/2008standards/documents/area_designations_for_the_2008_revised_o zone_naaqs.pdf CAPCOG. Capital Area Council of Governments Ozone Monitoring Network Assessment. February 2013. http://www.capcog.org/documents/airquality/reports/2013/capcog_air_quality_monitoring_2013_netwo RK_ASSESSMENT.pdf 2
Table of Contents Executive Summary... 2 Section 1: Introduction... 6 1.1: EPA Monitoring Network Assessment Guidance and Regulations... 6 1.2: CAPCOG Ozone Monitoring Goals... 9 1.3: CAPCOG s Air Monitoring Stations... 11 1.4: Equipment Available for 2014-2015 Monitoring... 12 1.5: Other Logistical Constraints... 13 1.6: Overall Level of Monitoring... 13 1.7: Ozone Monitoring History in CAPCOG Region... 15 Section 2: Site-By-Site Network Analysis... 19 2.1: Number of Parameters Measured... 20 2.2: Trend Impacts... 21 2.3: Measured Concentrations... 22 2.3.1: Fourth Highest Eight-Hour O 3 Averages... 22 2.3.2: Cumulative Seasonal Ozone Exposure... 23 2.4: Area and Population Served... 24 2.5: Correlation Analysis... 28 2.6: Probability of Peak and Minimum Ozone Measurements on High Ozone Days... 31 2.6.1: Probability of Measuring the Region-wide Maximum 8-Hour Ozone Concentration on High Ozone Days.... 31 2.6.2: Probability of Measuring the Region-wide Minimum 8-Hour Ozone Average on High Ozone Days... 32 2.7: Jurisdictional Value for Ozone Action Program Action Plan Counties... 34 Section 3: Bottom-Up Analysis... 34 3.1: Emissions Inventory Analysis... 35 3.2: Photochemical Modeling Analysis... 43 3.2.1: Comparisons of 4 th -Highest 8-Hour Ozone Daily Max for June 2006... 48 Section 4: Other Analyses Used for Monitoring Network Optimization... 49 4.1: Assessment of Monitoring Start and End Dates... 49 4.2: Data Handling Procedures... Error! Bookmark not defined.53 4.3: Funding Available for 2014-2015 Monitoring... 55 Section 5: Conclusions and Recommendations... 56 Figure 1: Map of Regional Air Pollution Monitoring Station Locations (white = TCEQ, green = CAPCOG) 12 Figure 2: Region 11 Ozone Monitors 1973-2013... 18 Figure 3: Temperature v. Peak 8-Hour Ozone at CAMS 3, April - October, 2006-2011... 20 Figure 4: Relative Humidity at Camp Mabry v. Peak 8-Hour Ozone at CAMS 3, April - October, 2006-2011... 21 Figure 5: Area and Population Served by Each Ozone Monitor in the CAPCOG Region... 26 Figure 6: Number of Days with Resultant Wind Direction from 6:00 18:00 CST at CAMS 3 on Days >= 60, 65, and 70 ppb, 2001-2009... 37 Figure 7: Wells Permitted and Completed in the Eagle Ford Shale Play... 39 Figure 8: Positions of CAPCOG Ozone Monitors and Major Point Sources of NO X in the Region Overlaid with CAMS 3 Wind Rose... 40 3
Figure 9: Modeled Daily Maximum 8-Hour Ozone Concentrations, June 3, 2006... 44 Figure 10: Modeled Maximum 8-Hour Ozone Concentrations, June 13, 2006... 45 Figure 11: Daily Maximum 8-Hour Ozone, June 29, 2006... 46 Figure 12: Daily Maximum 8-Hour Ozone, June 14, 2006... 47 Figure 13: 4th Highest Daily 8-Hour Ozone Concentrations in Central Texas, June 2006 Episode... 48 Figure 14: Number of High 8-Hour Ozone Days by Month, 2007-2013... 51 Figure 15: Monthly Distribution of 4 Highest 8-Hour Ozone Averages, 2007-2013... 52 Figure 16: Unadjusted and Adjusted 4th Highest 8-Hour Ozone Averages, 2013... 55 Table 1: EPA Air Monitoring Goals Potentially Applicable to CAPCOG's Ozone Monitoring Program... 8 Table 2: CAPCOG Ozone Monitoring Program Goals... 9 Table 3: Locations and Site Lease Details of CAPCOG Monitoring Stations... 11 Table 4: Monitoring Equipment Available for 2014-2015 Monitoring Network Operations... 13 Table 5: Total Ozone Monitors, Population, and Land Area for Ozone Planning Areas in Texas, 2013... 14 Table 6: Total Ozone Monitors, Population, and Land Area in CAPCOG Region and Texas, 2013... 15 Table 7: CAPCOG Region Ozone Monitoring Network Milestones... 16 Table 8: Site-by-Site Analysis Techniques... 19 Table 9: Number of Parameters Measured by Site... 21 Table 10: Number of Ozone Seasons in Operation for CAPCOG Region Ozone Monitor Stations and Trend Impacts Score... 22 Table 11: Fourth Highest Daily 8-Hour Ozone Averages 2011-2013 by Station... 23 Table 12: Cumulative Seasonal Exposure at CAPCOG Sites Using W126 Statistic... 24 Table 13: CAPCOG Land Area and Population Covered by Regional Ozone Monitors... 27 Table 14: Land Area and Population Served Scores for CAPCOG Monitoring Stations.... 28 Table 15: Correlation Coefficients for Monitors on Days Over 60 ppb 2007-2013 (n = 317)... 29 Table 16: Correlation Coefficients for Monitors on Days Over 65 ppb (n = 186)... 29 Table 17: Correlation Coefficients for Monitors on Days Over 70 ppb (n = 90)... 30 Table 18: Correlation Coefficients for Monitors on Days Over 75 ppb (n = 36)... 30 Table 19: Average Correlation Coefficients Among Monitors and Corresponding Scores... 31 Table 20: Probability of a Monitoring Station Measuring the Region's Highest 8-Hour Ozone Average on High Ozone Days... 32 Table 21: Probability of a Monitoring Station Measuring the Region's Lowest Peak 8-Hour Ozone Average on High Ozone Days... 33 Table 22: "Political Value" Score Assigned to Each Monitoring Station... 34 Table 23: Bottom-Up Analysis Techniques... 35 Table 24: Number of Days with Resultant Wind Direction from 6:00-18:00 CST at CAMS 3 and 38 on Days >= 75 ppb 2006-2011... 38 Table 25: 2011 NO X and VOC Emissions by County and Land Area - Selected Counties... 41 Table 26: Assignment of NO X Emissions from Nearby Counties to Monitoring Stations for Emissions Inventory Analysis... 42 Table 27: Emissions Inventory Scores Assigned to Monitoring Stations... 43 Table 28: Modeled 4th Highest 8-Hour Ozone Average for Monitoring Stations in MSA and Associated Score... 49 Table 29: Ozone Season Monitoring Start and End Month Requirements Under Appendix D to 40 CFR Part 58... 50 Table 30: Earliest and Latest Calendar Dates with High-Ozone Days Measured at any Site, 2007-2013.. 51 4
Table 31: Earliest and Latest Calendar Dates Measuring One of the Fourth Highest 8-Hour O 3 Averages for a Year 2007-2013... 52 Table 32: Calculated Intercept and Slope Adjustments for CAMS 690, 2013... 53 Table 33: Selected Data Adjustments for CAMS 690 for 2013 on Days when 8-Hour Ozone Average >= 69 ppb... 54 Table 34: Ozone Monitoring Operation Costs for 2014-2015... 56 Table 35: Weights Assigned to Each Analysis Technique Score... 56 Table 36: Scoring and Rankings for CAPCOG Sites... 57 Table 37: Cost-Benefit Ranking of CAPCOG Monitors... 58 5
Section 1: Introduction This monitoring network review and optimization plan is designed to provide CAPCOG s Air Quality Program with an analysis of the value each of CAPCOG s current ozone monitors is providing in meeting various air quality planning goals, and whether any adjustments to the network, its operations, or its data-handling procedures, are warranted for the 2014-2015 monitoring period. Since CAPCOG s monitoring network is not designed to meet EPA regulatory requirements, the specific goals of CAPCOG s monitoring efforts may be somewhat different from the goals that might exist if the network was designed for regulatory purposes. This monitoring network assessment and plan includes several changes from the previous monitoring network assessment completed in early 2013. Specifically, this analysis assigns specific quantitative scores to each CAPCOG monitoring station, makes specific recommendations for the start and end dates for monitoring operations, and makes specific recommendations for data handling and reporting. Since CAPCOG s monitoring network is designed to supplement TCEQ s ozone monitors, the analyses in this document must account for how CAPCOG s monitors provide data that is distinct from the data that is being collected by TCEQ s sites. For each specific analysis that relate to the value of each monitoring station, CAPCOG scored each station on a scale from 0 to 100, with zero being the lowest value and 100 being the highest value. In order to ensure that the relative value of one CAPCOG monitor relative to another CAPCOG monitor took into account the monitoring network as a whole, it was sometimes necessary to use TCEQ monitoring data to set the 0-100 point scale. For instance, the site with the region s largest population served was TCEQ s CAMS 3; in order to calculate the scores of CAPCOG s monitoring stations, it was necessary to set their scores at levels that considered CAMS 3 to have a score of 100 and all other monitors to have lower scores. The federal regulations for ozone monitoring and EPA s guidance on conducting monitoring network reviews does provide a good starting place for this analysis, however, and CAPCOG used these two sources of information as guidance for its preparation of this analysis and optimization plan. Except as they relate meeting specific regulatory requirements, formally establishing compliance with NAAQS, for a few other elements that are outside the scope of this analysis, the guidance and analyses recommended by EPA can be used in an analysis of a non-regulatory monitoring network such as CAPCOG s. 1.1: EPA Monitoring Network Assessment Guidance and Regulations While CAPCOG does not conduct regulatory ozone monitoring, the basic goals EPA defines for monitoring networks under 40 CFR Part 58, Appendix D Network Design Criteria for Ambient Air Quality Monitoring are still largely applicable to CAPCOG s monitoring network and useful for this assessment. The Network Design Criteria defines three basic monitoring objectives that EPA uses in evaluating the adequacy of air pollutant monitoring networks. These three objectives are to: a) Provide air pollution data to the general public in a timely manner; b) Support compliance with ambient air quality standards and emissions strategy development; and c) Support for air pollution research studies. 6
Federal regulations further define six basic types of monitoring stations: 1. Sites located to determine the highest concentrations expected to occur in the area covered by the network; 2. Sites located to measure typical concentrations in areas of high population density; 3. Sites located to determine the impact of significant sources or source categories on air quality; 4. Sites located to determine general background concentration levels; 5. Sites located to determine the extent of regional pollutant transport among populated areas and in support of secondary standards; 6. Sites located to measure air pollution impacts on visibility, vegetation damage, or other welfarebased impacts. In accordance with federal requirements, TCEQ operates two ozone monitors in the Austin-Round Rock MSA based on population and the most recent ozone design value. * The table below shows the required number of monitors for each area. Table: Minimum Monitors Required for MSAs by Population MSA Population Most Recent 3-Year Design Value Concentrations 85% of any O3 NAAQS Most Recent 3-Year Design Value Concentrations < 85% of any O3 NAAQS >10 million 4 2 4-10 million 3 1 350,000 <4 million 2 1 50,000 - <350,000 1 0 TCEQ s two monitoring stations CAMS 3 and CAMS 38 in Travis County fulfill these requirements and are sited primarily to measure peak ozone and establish compliance with the ozone NAAQS. CAPCOG s monitoring program is not subject to these requirements, and is deployed for reasons other than determining NAAQS compliance. EPA s network assessment guidance provides more detailed list of typical goals for an air quality monitoring network. Not all of the goals mentioned by EPA are applicable to ozone or to CAPCOG s nonregulatory ozone monitoring program. For instance, speciation is relevant for particulate matter monitoring but not directly for ozone monitoring, and compliance with NAAQS is relevant for regulatory ozone monitoring, but not non-regulatory ozone monitoring. The following list represents the goals identified in this guidance document that are applicable or could be applicable to CAPCOG s ozone monitoring program. * Table D-2 of Appendix D to 40 CFR Part 58 SLAMS Minimum O3 Monitoring Requirements. U.S. Environmental Protection Agency. Ambient Air Monitoring Network Assessment Guidance. EPA-454/D-07-001. February 2007. http://www.epa.gov/ttnamti1/files/ambient/pm25/datamang/network-assessmentguidance.pdf 7
Table 1: EPA Air Monitoring Goals Potentially Applicable to CAPCOG's Ozone Monitoring Program EPA-Specified Monitoring Goals Compliance with NAAQS Air Quality Model Evaluation Trend Tracking Monitor the Area of Maximum Precursor Emissions Monitor the Area of Maximum Pollutant Concentration Monitor Background Concentrations Monitor Surrogate Pollutants Transport/Border Characterization Interpolation and Understanding of Pollution Gradients Accountability/Performance Measurement Forecasting Assistance Environmental Justice Public Reporting of the Air Quality Index Applicable to CAPCOG Ozone Monitoring Program In the past, CAPCOG has operated both nitrogen oxides (NO X ) analyzers and sulfur dioxide (SO 2 ) analyzers to measure precursor and surrogate pollutants, so broadly, the goals for measuring maximum precursor emissions and monitoring surrogate pollutants are potentially appropriate for CAPCOG s monitoring program. However, CAPCOG s review of the costs of repairing its two NO X and two SO 2 monitors compared to the potential benefit makes these goals beyond the scope of this study. The extent to which the goal to Monitor the Area of Maximum Pollutant Concentration is applicable to CAPCOG s ozone monitoring program is limited to research purposes and not regulatory compliance purposes. The data collected at CAPCOG s monitors do not show ozone design values as calculated by EPA for regulatory compliance purposes that would be any higher than TCEQ s two regulatory monitoring stations. However, on specific days throughout the ozone season, some of CAPCOG s ozone monitors do measure the region s maximum pollutant concentrations, and the data collected at CAPCOG s monitors on those days are useful to understanding the regional ozone pollution levels. Therefore, in this report, where monitor area of maximum pollutant concentration is mentioned, it is not meant to infer that any of CAPCOG s monitors are measuring or likely to measure the region s highest 3-year ozone design value. Due to the way ozone forms as a secondary pollutant and the broad geographic area that it typically affects (25-30 mile plumes, according to a 2011 mobile monitoring study completed by the University of Texas at Austin for CAPCOG) *, accountability/performance measurements and environmental justice considerations are not applicable to this evaluation. * The University of Texas at Austin, Center for Energy and Environmental Resources. Surface Mobile Monitoring in the Austin area for 2011. April 2012. http://www.capcog.org/documents/airquality/reports/2011/austin- _Surface_Mobile_Monitoring_2011.pdf. 8
Finally, while CAPCOG s ozone monitors do publicly report data out to TCEQ s website, they are not used for official air quality index reporting. Therefore, that goal as defined by EPA is not applicable to this analysis either. 1.2: CAPCOG Ozone Monitoring Goals CAPCOG s ozone monitoring program is focused primarily on: 1) providing useful information to help protect public health, and 2) provide data that is useful for characterizing background ozone levels and local contributions to peak ozone levels. To a lesser degree, CAPCOG s monitoring program is also designed to support general scientific studies about ozone. The table below consolidates the goals listed in the previous section that would be potentially applicable to CAPCOG s ozone monitoring program, and indicates the prioritization of these goals for CAPCOG. CAPCOG has categorized these goals into the three overarching goals identified in the Federal regulations: providing public information on ambient air quality, planning, and science. Table 2: CAPCOG Ozone Monitoring Program Goals Goal Category Primary Secondary Tertiary Air Quality Model Evaluation Science Trend Tracking Planning Monitor Area of Maximum Precursor Emissions Science Monitor Area of Maximum Pollution Concentration Public Health Monitor Typical Concentrations in Areas of High Population Density Public Health Monitor Surrogate Pollutants Science Determine the General Background Pollution Concentrations Planning Determine the Impact of Significant Sources or Source Categories on Air Quality Planning Determine the Extent of Regional Pollutant Transport Planning Interpolation and Understanding Pollutant Gradients Science Measure Air Pollution Impacts on Welfare- Based Impacts Science Forecasting Assistance Science The public health imperative of obtaining an accurate picture of the ground-level ozone measurements in the region is one of the most important goals of the program. Having an accurate picture of the spatial extent of ozone within the region and being able to provide information that is most relevant to the individual communities within the region is also an important aspect of CAPCOG s monitoring program. Since the Austin-Round Rock MSA often has some of its highest ozone levels when winds are out of the northeast measuring ozone levels to the southwest of Austin is of interest to CAPCOG. 9
Providing ozone data specific to the various communities within the Austin-Round Rock MSA outside of Austin proper a priority to CAPCOG, given the significant variation in ozone levels within the MSA on high ozone days (of the 36 days when at least one monitor in the region measured an 8-hour ozone average above 75 ppb, the average difference between the highest and lowest 8-hour ozone averages was 18.97, and was as high as 57 ppb). All five of the counties and almost every city over 5,000 in population participates in the regional air quality planning effort, and having air quality monitoring data that is useful to each of those communities is important to maintaining a high level of awareness about air quality within the communities outside of Austin. For example, until last year, Caldwell County had no ozone monitor of its own. The Lockhart monitor may not cover a very large population, but it provides Caldwell County its own monitor and data specific to those communities, rather than relying on monitors in Hays, Bastrop, and Guadalupe Counties. Supplemental monitoring ozone monitoring can help improve the understanding background ozone concentrations and the nature of regional ozone transport. Understanding regional ozone transport and the extent to which local emission reduction efforts may be able to reduce the region s design value is especially important for decision-makers and policy makers within the region. Therefore, having ozone monitoring stations positioned to capture upwind ozone concentrations provides very important data in understanding ozone in the region. Tracking ozone trends over time is another important goal filled by CAPCOG s monitoring network. Some of CAPCOG s monitors have been operating for over a decade now, and several others that are somewhat newer have provided valuable data showing the success or deficiencies in the region s ozone control strategy over time. One of the more important uses of ozone and meteorological monitoring data is the development of conceptual models for the Austin area. Their presence helps provide researchers with data to get a more complete picture of ozone formation in the region, and the extent to which they are positioned correctly and there are enough of them to do so is important in determining the value of the region s ozone conceptual models. The primary customers for the monitoring data collected by CAPCOG s are the local communities that CAPCOG serves, the TCEQ, and EPA. CAPCOG s Air Quality Program believes that measuring background ozone concentrations is critical for this region because of the extent to which background levels contribute to local ozone levels. Maintaining a broad ozone monitoring network enables citizens of the CAPCOG region to have additional data on air quality conditions specific to their own communities. Finally, in the event that counties in the CAPCOG region are designated nonattainment, continued monitoring will be important to ensure that any ozone reductions strategies are accurately modeled so as to avoid unnecessary regulation and provide useful and accurate information to local decision-makers about the effectiveness of emission reduction strategies. Having sufficient monitoring data for ozone models calibration is an important application of monitoring data, although the use of monitoring data collected during 2014 and 2015 for this purpose would likely be many years in the future if it was ever used for model calibration. Therefore, it is a somewhat less urgent priority, compared to other ambient monitoring goals. 10
1.3: CAPCOG s Air Monitoring Stations CAPCOG currently holds site lease agreements for eight monitoring stations in the region, which include one in Bastrop County, one in Caldwell County, one in Fayette County, two in Hays County, one in Travis County, and two in Williamson County. These sites are owned by the Lower Colorado River Authority (LCRA), the Dripping Springs Independent School District (DSISD), the U.S. Army Corps of Engineers (USACE), the City of San Marcos, the Hutto Independent School District (HISD), the City of Lockhart, and the Austin Independent School District (AISD). The table below shows the locations and details of the terms of the site lease agreements for each monitoring station. Table 3: Locations and Site Lease Details of CAPCOG Monitoring Stations Station Address Latitude Longitude Owner Expires CAMS 601 636 Roznov Road, Roundtop, August 31, 29.9634-96.7459 LCRA TX 2014 CAMS 614 29400 Ranch Road 12, December 30.2145-98.0833 DISD Dripping Springs, TX 31, 2014 CAMS 684 1884 State Highway 71 W, August 31, 30.1408-97.4589 LCRA Cedar Creek, TX 2014 CAMS 690 500 Overlook Dr., January 14, 30.6665-97.7345 USACE Georgetown, TX 2018 CAMS 1675 599 Staples Road, San City of San December 29.8622-97.9289 Marcos, TX Marcos 31, 2018 CAMS 6602 200 College Street, Hutto, TX 30.5457-97.5424 HISD August 31, 2014 Lockhart 214 Bufkin Lane, Lockhart, TX 29.8649-97.6650 City of December Lockhart 31, 2015 Gorzycki Middle 7412 W. Slaughter Lane, December 30.2146-97.8913 AISD School Austin, Texas 78749 31, 2015 As the table above shows, three of these eight stations have site leases that expire on August 31, 2014, but CAPCOG staff is in the process of negotiating extensions through 2018 with all of these site lease holders and expects them all to be extended at least through the end of 2015. The map below shows the locations of each of CAPCOG s ozone monitoring stations (green) in relation to the two regulatory ozone monitoring stations operated by TCEQ (white). 11
Figure 1: Map of Regional Air Pollution Monitoring Station Locations (white = TCEQ, green = CAPCOG) 1.4: Equipment Available for 2014-2015 Monitoring The following table shows the number of operational pieces of equipment analyzers, sensors, and other equipment that may be available for use for 2014 and 2015 ozone monitoring, including some equipment that is owned by TCEQ which it has allowed CAPCOG to use at its monitoring stations in the past. Not included on this list are two inoperative nitrogen oxides (NO X ) analyzers and two inoperative SO 2 analyzers owned by CAPCOG that were mentioned earlier that would need further repair if they were going to be put into service. Dios Dado Environmental (DDE), who is CAPCOG s monitoring contractor, has estimated that it would cost $5,315.00 per unit to repair the NO X analyzers, $4,161.95 to repair one of the SO 2 analyzers and $7,568.35 to repair the second one. The table below shows the monitoring equipment that CAPCOG is planning on using for the 2014 and 2015 ozone seasons. The most recent equipment inventory was completed at the end of the 2013 ozone season and can be found in the 2013 Monitoring Report. * A new site-by-site equipment inventory will be available by the end of April 2014. * Dios Dado Environmental, Ltd. Annual Ambient Air Quality Monitoring Report, CAPCOG ~ CAMS 601, 614, 684, 690, 1675, 6602, The City of Lockhart Electric Department & Gorzycki Middle School. December 13, 2013. 12
Table 4: Monitoring Equipment Available for 2014-2015 Monitoring Network Operations Equipment Type Description/Model Number CAPCOG Owns TCEQ Owns Locations Ozone Analyzer Tanabyte 722 7 0 CAMS 684, 1675, 5 in storage Ozone Analyzer Teledyne 400E 2 0 Storage Ozone Analyzer Dasibi 1008-AH 5 0 CAMS 601, 614, 690, Lockhart, Gorzycki, 1 in storage Wind Direction/Wind R.M. Young Model CAMS 684, 1675, Lockhart, 6 0 Speed Sensor 05305 Gorzyki, 2 in storage Wind Direction/Wind Speed/Ambient Met One F-460 0 3 CAMS 601, 614, 690 Temperature Senor Temperature/Relative Humidity Sensor Met One 083-1 7 0 Storage Data Logger Zeno 3200 4 4 All stations, 1 in storage Trailers Various 2 2 CAMS 601, 614, 690, Lockhart Automated Calibrator Tanabyte Model 322 1 0 Storage Zero Air Generator Teledyne API Model 701 1 0 Storage Pumps Thomas Pump Model 607CA316 2 0 Storage Stainless Steel Compressed Gas Regulator Scott-Marrin Model 2SS75-660- D4T 1 0 Storage 1.5: Other Logistical Constraints At this point, CAPCOG is limiting this review to the prioritization of the eight monitoring stations for which it currently has site lease agreements. While there may be other locations within the region that would benefit from a monitoring station where one does not yet exist, given the resource constraints for 2014 and 2015 and the amount of time required to get a new site lease agreement in place, CAPCOG is only considering these eight locations for 2014 and 2015. Given the extensive amount of lead time required to relocate a site and the character of the current configuration, additional data should be collected from the current network before adding or subtracting any of these sites. This review should help prioritize these sites for future review, however.. 1.6: Overall Level of Monitoring In order to evaluate the overall number of monitors CAPCOG should field, it is useful to compare the overall level of monitoring in the Austin-Round Rock MSA and the CAPCOG region at large to other ozone planning areas and other regions around the state. Under CAPCOG s 2014-2015 Rider 8 Grant 13
Work Plan, CAPCOG committed to fielding between 6 and 8 monitoring stations. Comparing the coverage of the local network if CAPCOG fielded up to 8 sites to the coverage in other areas helps CAPCOG determine whether 6, 7, or 8 sites would be the level most comparable with other air quality planning areas and other regions of the state. For this analysis, CAPCOG used the Dallas-Fort Worth Nonattainment Area, the Houston-Galveston-Brazoria Area, and MSAs or CSAs (Combined Statistical Areas), as appropriate. Since traditionally, the Tyler-Longview-Marshall area has been considered one consolidated planning area, that is how it is reflected in the table below. If CAPCOG fielded all of the monitors it has site leases for in 2014 and 2015, the Austin-Round Rock MSA would have 0.51 monitors per 100,000 people (2 TCEQ monitors and 7 CAPCOG monitors CAPCOG s CAMS 601 is outside of the MSA boundaries), and 2.13 monitors per 1,000 square miles. This would make it ranked 8 th by monitors per person and 5 th by monitors per square mile among all of the ozone air quality planning areas in the state. If CAPCOG only fielded five monitors in the MSA, it would bring number of monitors per 100,000 down to 0.39, making it 11 th (ahead of only Waco and Laredo) per person, while the number of monitors per 1,000 square miles would drop to 1.66 (still ranked 5 th ). Table 5: Total Ozone Monitors, Population, and Land Area for Ozone Planning Areas in Texas, 2013 Ozone Planning Area Ozone Monitors Population Square Miles Monitors per 100,000 people Monitors per 1,000 sq. mi. Abilene MSA 0 168,587 2,744 0.00 0.00 Amarillo-Borger CSA 0 279,728 6,037 0.00 0.00 Austin-Round Rock MSA 9 1,772,876 4,220 0.51 2.13 Beaumont-Port Arthur MSA 9 409,363 3,034 2.20 2.97 Brownsville-Harlingen- Raymondville CSA 2 446,090 1,482 0.45 1.35 College Station-Bryan MSA 0 234,154 2,100 0.00 0.00 Corpus Christi-Kingsville-Alice CSA 8 512,636 4,989 1.56 1.60 Dallas-Fort Worth Nonattainment Area 17 6,466,598 7,830 0.26 2.17 El Paso-Las Cruces CSA (El Paso County Only) 6 832,099 5,584 0.72 1.07 Hood County 1 51,128 421 1.96 2.38 Houston-Galveston-Brazoria Nonattainment Area 43 6,073,022 7,612 0.71 5.65 Killeen-Temple MSA 2 420,703 2,816 0.48 0.71 Laredo MSA 1 262,659 3,362 0.38 0.30 Lubbock-Levelland CSA 0 320,561 3,596 0.00 0.00 McAllen-Edinburg CSA 1 873,715 2,794 0.11 0.36 Midland-Odessa CSA 0 287,975 2,713 0.00 0.00 San Angelo MSA 0 113,620 2,574 0.00 0.00 San Antonio-New Braunfels MSA 11 2,199,285 7,313 0.50 1.50 Texarkana MSA (Texas only) 0 93,559 885 0.00 0.00 14
Ozone Planning Area Ozone Monitors Population Square Miles Monitors per 100,000 people Monitors per 1,000 sq. mi. Tyler-Longview-Marshall Near- Nonattainment Area 3 571,229 5,579 0.53 0.54 Victoria-Port Lavaca CSA 3 117,798 2,241 2.55 1.34 Waco MSA 1 257,124 1,803 0.39 0.55 Wichita Falls MSA 0 143,695 1,717 0.00 0.00 TOTAL All Ozone Planning Areas 117 22,908,204 83,443 0.51 1.40 If CAPCOG were to field only five monitoring stations within the MSA, the number of monitors per person would be much lower than any other near-nonattainment area other than the next largest (San Antonio-New Braunfels, at 0.50 monitors per 100,000 people), but comparable in geographic coverage to Corpus Christi and San Antonio. If CAPCOG were to field seven monitoring stations within the MSA, the number of monitoring stations per 100,000 people would be comparable to that of the San Antonio and Tyler-Longview-Marshall near-nonattainment areas, and the geographic density would be comparable to that of the Dallas-Fort Worth area. This same analysis can be performed comparing the CAPCOG region as a whole to the State as a whole. Beyond the five-county Austin-Round Rock MSA, CAPCOG includes five additional counties outside of the Austin-Round Rock MSA, and there is one additional ozone monitor (CAMS 601) that is located outside of the MSA counties within the CAPCOG region. Table 6: Total Ozone Monitors, Population, and Land Area in CAPCOG Region and Texas, 2013 Region Ozone Monitors Population Land Area Monitors per 100,000 people Monitors per 1,000 square miles CAPCOG 10 1,886,241 8,437 0.53 1.19 TEXAS 121 25,843,786 261,233 0.47 0.46 These analyses, which include both regulatory and non-regulatory monitors, indicate that that operating eight monitoring stations would bring both the MSA and the CAPCOG region s overall monitoring coverage at a level more comparable to the rest of the state than fielding only seven or six stations would. 1.7: Ozone Monitoring History in CAPCOG Region CAPCOG has been involved in ozone monitoring in the region since 2002. CAPCOG s monitoring network has grown from just one monitoring station in 2003 to eight stations in 2013. CAPCOG s network currently includes including CAMS 601 * (Fayette County), CAMS 684 (McKinney Roughs), CAMS 690 * Note: CAMS 601 is still listed on TCEQ s website as being owned by LCRA, but CAPCOG took over direct operation of the site from LCRA on September 1, 2012, through a site lease agreement. Prior to that date, CAPCOG was 15
(Lake Georgetown), CAMS 1675 (San Marcos), and CAMS 6602 (Hutto). These monitors have played an important role in helping characterize ozone within the region and serve to collect supplemental monitoring data for Central Texas beyond the data collected by the two Federal Reference Monitor ozone samplers operated by TCEQ in Travis County CAMS 3 (Murchison Middle School) and CAMS 38 (Austin Audubon Society). Regional ozone monitoring efforts in the region began in January 1973 following the passage of the Federal Clean Air Act. The table below provides a detailed timeline of ozone monitoring milestones in the region. Table 7: CAPCOG Region Ozone Monitoring Network Milestones Date January 1, 1967 January 1, 1973 January 1, 1974 January 1, 1979 January 1, 1981 February 28, 1997 August 12, 1998 May 18, 2000 December 10, 2002 Milestone TCEQ commences operation of monitoring station at East 53 rd Street. Site collects ozone data only in 1973. TCEQ commences operation of monitoring station at Hunters Glen. Site collects ozone data only in 1973. TCEQ commences operation of monitoring stations at Lavaca & 17 th Streets and TACB headquarters. Lavaca Street site collects ozone data only from the 4 th quarter of 1976 through the 4 th quarter of 1979. TACB headquarters site collects ozone data only from the 1 st quarter of 1974 through the 3 rd quarter of 1979. TCEQ commences operation of CAMS 3 at Murchison Middle School and an analytical lab site. CAMS 3 has been collecting ozone data since the first quarter of 1979, while the analytical lab site collected data only in the first two quarters of 1980. TCEQ commences operation of CAMS 25 at Parmer Lane and Mopac. Site collects ozone data from the 3 rd quarter of 1981 to the 1 st quarter of 1997. TCEQ commences operation of CAMS 38 at the Austin Audubon Society located at 12200 Lime Creek Road in northwest Austin. TCEQ also shuts down CAMS 25 at Parmer Lane. TCEQ commences operation of CAMS 62 at the San Marcos Airport located at 2041 Airport Drive in San Marcos. LCRA commences operation of CAMS 601 located at 636 Roznov Road in Round Top. The site has measured ozone, NO X, SO 2, PM 2.5, outdoor temperature, wind speed, and wind direction. CAPCOG commences operation of CAMS 613 at the Pflugerville Wastewater Facility at 2609 East Pecan Street. Site measures ozone, NO X, SO 2, temperature, wind speed, and wind direction. indirectly supporting the site through a contract with LCRA to operate an ozone analyzer there during ozone season. 16
Date March 11, 2003 Milestone CAPCOG commences operation of CAMS 614 at the Dripping Springs Elementary School at 29400 Ranch Road 12 in Dripping Springs. Also equipped with a NO X analyzer and wind speed/wind direction/outdoor temperature sensor in a trailer. April 5, 2003 June 2, 2006 August 16, 2006 November 3, 2006 September 20, 2007 October 31, 2010 April 29, 2011 September 10, 2011 May 18, 2011 TCEQ permanently ceases operation of CAMS 62 at the San Marcos Airport. CAPCOG commences operation of CAMS 674 at 212 Commerce Street in Round Rock and CAMS 675 at 222 Sessoms Drive in San Marcos. Both sites are equipped with an ozone analyzer and wind speed/wind direction sensors. The San Marcos site is located in a trailer. CAPCOG commences operation of CAMS 684 at the McKinney Roughs Nature Park at 1884 State Highway 71 in Cedar Creek. The site is located in the Visitor s Center and is equipped with an ozone analyzer and wind speed/wind direction sensor. CAPCOG permanently ceases operation of CAMS 613 in Pflugerville. CAPCOG commences operation of CAMS 690 at Lake Georgetown, 500 Lake Overlook Drive in Georgetown. Site measures ozone, NO X, SO 2, and wind speed/wind direction/outdoor temperature sensor in a trailer. CAPCOG permanently ceases operation of CAMS 674 in Round Rock. CAPCOG ceases operation of NO X and SO 2 monitoring at CAMS 601 and CAMS 690 due to resource constraints. Under contract with CAPCOG, The University of Texas at Austin conducts mobile surface sampling in and around Austin and surrounding counties on seven separate days. CAPCOG commences operation of CAMS 6602 at 200 College Street in Hutto. Site measures ozone, NO X, and SO 2, and includes an automated calibrator, and wind direction/wind speed sensor. September 14, 2011 CAPCOG permanently ceases operation of CAMS 675. September 20, 2011 May 29, 2012 June 27, 2012 November 2, 2012 CAPCOG commences operation of CAMS 1675 at 599 Staples Road in San Marcos. Site includes an ozone analyzer and wind direction/wind speed sensor and is located inside a building at a water utility pump station. Based on the results of the mobile monitoring conducted in 2011, CAPCOG establishes a temporary monitoring site at the Liberty Hill Fire Department. This site was equipped with an ozone monitor and wind speed/wind direction sensor. CAPCOG commences operation of a temporary monitoring site at the Elroy Public Library. This site was equipped with an ozone monitor and wind speed/wind direction sensor. CAPCOG ceases operation of temporary monitoring sites at Liberty Hill and Elroy. 17
Date May 29, 2013 August 30, 2013 Milestone CAPCOG commences operation of a temporary monitoring site at the Lockhart Police/EMS Station in Caldwell County. The site is equipped with an ozone monitor and wind speed/wind direction sensor. CAPCOG commences operation of a temporary monitoring site at the Gorzycki Middle School in southwest Austin. The site is equipped with an ozone monitor and wind speed/wind direction sensor. As the table above documents, the ozone monitoring network has grown considerably over time, especially in the past ten years due to Rider 8 near-nonattainment program funding. The figure below shows the total number of ozone monitors in the region by year dating back to 1973. Figure 2: Region 11 Ozone Monitors 1973-2013 14 12 10 8 6 4 2 0 Regulatory CAPCOG + LCRA In addition to monitoring ozone, CAPCOG has also previously conducted continuous monitoring of nitrogen oxides (NO X ) and sulfur dioxide (SO 2 ), which can be an indicator of industrial pollution plumes. Currently, there is only one nitrogen oxides monitor in the region at TCEQ s CAMS 3 (April 1, 2012 present), although a new near-road NO X monitor should be deployed by TCEQ in the near future. Volatile Organic Compounds (VOC) are monitored through non-continuous sampling at TCEQ s CAMS 171, located on Webberville Road in East Austin. NO X monitoring has previously been conducted by TCEQ at CAMS 38 (1997 2012), and by CAPCOG or LCRA at CAMS 601 (2001 2010), CAMS 613 (December 2002 January 2003), CAMS 614 (2003 2011), CAMS 690 (2008 2010), and CAMS 6602 (2011-2012). SO 2 monitoring has previously been conducted at CAMS 601 and CAMS 690 (2008 2010). CAPCOG has also conducted special 1-hour, early-morning VOC sampling during ozone season in previous years for the Rider 8 program. 18
Section 2: Site-By-Site Network Analysis The EPA s guidance recommends performing a number of site-by-site analyses to compare the value of each of the existing monitoring stations. CAPCOG performed several of these as indicated in the table below. CAPCOG has added a total of three different types of analysis techniques to this list the probability of a given monitor measuring the region-wide maximum 8-hour ozone concentration on high ozone days, the probability of a given monitor measuring the region-wide minimum 8-hour ozone concentration on high ozone days, and the jurisdictional value to CAPCOG s Air Quality Program of each monitoring station. Table 8: Site-by-Site Analysis Techniques Technique Objectives Assessed Included Number of Parameters Monitored at Site Overall Site Value Model Evaluation Source Apportionment Trends Impact Trend Tracking Historical Consistency Emission Reduction Evaluation Measured Concentrations Maximum Concentration Location Model Evaluation Regulatory Compliance Population Exposure Deviation from NAAQS Regulatory Compliance Forecasting Assistance Area Served Spatial Coverage Interpolation Background Concentration Population Served Population Exposure Environmental Justice Monitor-to-Monitor Correlation Model Evaluation Spatial Coverage Interpolation Principal Component Analysis Background Concentration Forecasting Assistance Removal Bias Regulatory Compliance Model Evaluation Spatial Coverage Background Concentration Interpolation Probability of Measuring Region-Wide Maximum and Minimum 8-Hour O 3 Concentrations (CAPCOG Maximum concentration location Transport/border characterization Analysis) Background concentration Jurisdictional Value Spatial Coverage Population Exposure 19
Maximum Austin Ozone (ppb) 2.1: Number of Parameters Measured EPA recommends that sites are compared based on the number of parameters measured. The guidance states, Air quality monitoring sites hosting monitors collocated with other measurement insturments are likely more valuable than sites at which fewer parameters are measures. In addition, the operating costs can be leveraged among several instruments at these sites. This analysis is performed by simply counting the number of other parameters that are measured at a physical site. Sites at which many parameters are measured are ranked highest. Beginning with the 2014 ozone season, all seven of CAPCOG s monitoring stations within the Austin- Round Rock MSA will be equipped with a relative humidity/temperature sensor. Only CAMS 601 will not be equipped with a relative humidity/temperature sensor. The region s conceptual model * indicates relationships between both of those meteorological factors and high ozone, so adding these parameters should enhance the scientific value of the data collected at each site. Figure 3: Temperature v. Peak 8-Hour Ozone at CAMS 3, April - October, 2006-2011 100 90 80 >= 75 ppb 70 60 50 40 30 20 10 0 40 45 50 55 60 65 70 75 80 85 90 95 100 105 Murchison Daily Maximum Temperature (F) * The University of Texas at Austin, Center for Energy and Environmental Resources. Conceptual Model for Ozone for the Austin Area. July 2012. 20
Maximum Austin Ozone (ppb) Figure 4: Relative Humidity at Camp Mabry v. Peak 8-Hour Ozone at CAMS 3, April - October, 2006-2011 100 90 80 >= 75 ppb 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Average Daily Relative Humidity (%) Table 9: Number of Parameters Measured by Site CAMS O 3 PM 2.5 (TEOM) Wind Temp RH Total Parameters Score 601 1 1 5 1 0 8 100 614 1 0 5 1 1 8 100 684 1 0 5 1 1 8 100 690 1 0 5 1 1 8 100 1675 1 0 5 1 1 8 100 6602 1 0 5 1 1 8 100 Lockhart 1 0 5 1 1 8 100 Gorzycki MS 1 0 5 1 1 8 100 Since all of CAPCOG s monitoring stations will be measuring eight parameters, all monitoring stations are assigned a score of 100 for this analysis. 2.2: Trend Impacts EPA recommends ranking each monitoring station by the length of time the monitoring station has been in service. The longer a monitor has been in service, the more its data can provide a picture of air quality trends. The table below ranks each of the monitoring stations by its time in service. Since the San Marcos monitor was moved a relatively short distance (< 2 miles) in September 2011, the trend period for the purposes of this analysis is considered to start with the installation of CAMS 675 in San Marcos on June 2, 2006. CAPCOG calculated the number of ozone seasons based on the date the monitor began 21
collecting 8-hour ozone averages. Fractions of an ozone season are based on the total number of days when data was collected in the initial year of operation compared to the total number of days between March 1 and October 31, which is the region s official ozone season. The score was based on the total number of ozone seasons for each site divided by the maximum value of 35 *, which is the number of ozone seasons TCEQ s CAMS 3 has been in service. CAPCOG used CAMS 3 s number of ozone seasons as the scale for this metric because the value of CAPCOG s monitors to track trends must be put in the context of the much longer trend data available from TCEQ s monitoring stations. It would be in appropriate to assign CAMS 601 a value of 100 in this analysis, for instance, because it would tend to over-emphasize the value of this monitor relative to CAPCOG s other monitors in light of the data already available at CAMS 3 over a much longer period of time. The equation below shows how the score for CAMS 601 was calculated, for illustrative purposes Table 10: Number of Ozone Seasons in Operation for CAPCOG Region Ozone Monitor Stations and Trend Impacts Score Station Data Collection Start Date Ozone Seasons Score CAMS 601 May 18, 2000 13.7 39 CAMS 614 March 11, 2003 11.0 31 CAMS 675/1675 June 2, 2006/September 20, 2011 7.6 22 CAMS 684 August 16, 2006 7.3 21 CAMS 690 May 17, 2008 6.7 19 CAMS 6602 May 18, 2011 2.7 8 Lockhart May 29, 2013 0.6 2 Gorzycki MS September 25, 2013 0.1 0 2.3: Measured Concentrations EPA recommends ranking each monitor based on the calculated design value of each site. While CAPCOG s monitors are not used for NAAQS compliance, and so would not have a design value as defined by the EPA, statistics used by EPA in its most recent ozone policy assessment document in evaluating ozone impacts on human health and vegetation can be calculated at CAPCOG s sites in in order to assign scores to each of CAPCOG s sites. 2.3.1: Fourth Highest Eight-Hour O 3 Averages EPA uses a monitor s fourth highest daily eight-hour ozone average, averaged across three years, as the basis for the current ozone NAAQS, and staff at EPA s Office of Air Quality Planning and Standards (OAQPS) is continuing to use that statistic in the review of the ozone standard that is currently underway for impacts on human health. The table below shows the most recent three years of fourth highest eight-hour ozone averages recorded at the six CAPCOG monitors that reported to the LEADS * CAMS 3 started collecting data on January 1, 1979. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards. Policy Assessment for the Review of Ozone National Ambient Air Quality Standards. Second External Review Draft. January 2014. http://www.epa.gov/ttn/naaqs/standards/ozone/data/20140131pa.pdf 22
system and the Lockhart monitor, which started collecting data in May 2013. The Gorzycki Middle School monitor was not included in this analysis because it started collecting reliable data only late in September 2013, and would skew the scores for all of the other monitoring stations. The table includes a consolidation of data collected at CAMS 675 and 1675, since the monitor was moved a relatively short distance (1.9 miles) within San Marcos in late 2011. CAPCOG translated these data into scores by assigning a value of 100 to the highest three-year average and a value of 0 to the lowest three-year average, and calculating the scores for the remaining monitoring stations relative to the highest and lowest values. The equation below shows how the score for CAMS 614 was calculated as an illustration: ( ) ( ) Table 11: Fourth Highest Daily 8-Hour Ozone Averages 2011-2013 by Station Site 2011 2012 2013 3-Year Average Score CAMS 601 75 68 64 69 43 CAMS 614 77 73 67 72 86 CAMS 684 72 71 64 69 43 CAMS 690 73 73 75 73 100 CAMS 1675/675 78 72 70 73 100 CAMS 6602 75 69 69 71 71 Lockhart n/a n/a 66 66 0 ( ) ( ) 2.3.2: Cumulative Seasonal Ozone Exposure In the most recent draft of the policy assessment for ozone, OAQPS staff state, in considering alternative forms of the [welfare-based secondary] standard we conclude that it is reasonable and appropriate to consider a cumulative, concentration-weighted form to provide protection against cumulative, seasonal exposures to O 3 that are known or anticipated to harm sensitive vegetation or ecosystems. Such a form is specifically designed to directly measure the kind of O3 exposures that can cause harm to vegetation and would have a distinct advantage over the form of the current standard in characterizing air quality conditions potentially of concern for vegetation and demonstrating that the desired degree of protection against those conditions was being achieved. (page 6-38) The statistic that EPA staff have developed for this purpose is known as the W126 statistic, which uses 1-hour ozone concentrations measured from 8 am 8 pm (the typical daylight hours during which plants experience growth) over a three-month period in order to calculate the maximum cumulative exposure experienced by vegetation during an ozone season. A more detailed explanation of how the statistic is calculated is available on EPA s website. * CAPCOG calculated the W126 statistic for each of its monitoring stations. Excel spreadsheets calculating these data are being submitted along with this document. * http://www.epa.gov/ttn/analysis/w126.htm 23
CAPCOG assigned scores to the calculated values in a manner similar to how it assigned scores for the fourth-highest eight-hour O 3 averages, assigning 100 to the highest value, and 0 to the lowest value. The data from the Gorzycki Middle School and Lockhart sites were not included in this analysis due to the need for consecutive three-month periods of data, and the small amount of such data for both sites. The equation below shows how the score for CAMS 614 was calculated as an illustration: ( ) ( ) ( ) ( ) Table 12: Cumulative Seasonal Exposure at CAPCOG Sites Using W126 Statistic CAMS 2011 2012 2013 3-Year Average Score 601 12.77 6.16 4.76 8 0 614 14.68 9.65 4.31 10 50 684 11.02 6.54 5.36 8 0 690 14.15 9.13 11.35 12 100 1675 13.71 8.43 6.08 9 25 6602 14.16 6.10 5.83 9 25 2.4: Area and Population Served Two of the analyses EPA recommends for a monitoring network evaluation involve comparing monitoring sites based on their area covered and population covered. These analyses involve calculating the sum of the square miles of land area and population located closer to one particular site than any other site. The area located closer to one point than any other point in this analysis is known as a Thiessen polygons. This technique provides a relatively simple way of comparing the areas and populations that each monitoring station serve, but does not account for meteorology, topography, or proximity to emissions sources. Population served was one of the five site-by-site criteria EPA used in its national-scale network assessment. The geographic area CAPCOG used for the land area and population served analysis was the ten-county CAPCOG region. Although the most likely nonattainment area boundary for the region would be the five-county Austin-Round Rock MSA, CAPCOG s Air Quality Program is still accountable to the entire tencounty region and to the extent that air pollution does not respect political boundaries, even counties that are not part of the MSA are benefitting from monitoring stations located elsewhere in the region. In order to properly evaluate the benefit of CAPCOG s monitoring stations for these analyses, it is necessary to account for any other monitors operated by other entities that any part of the CAPCOG region may be closer to than a CAPCOG monitor. Failing to account for the area and population that are closer to these other sites would incorrectly assign greater value to CAPCOG monitoring stations than is warranted. Therefore, for this analysis, CAPCOG included: All eight of the stations for which CAPCOG currently holds site leases; The two TCEQ ozone monitoring stations in Travis County; 24
The two ozone monitoring stations in Guadalupe County that have produced Thiessen polygons that covered portions of the southern part of the CAPCOG region (see CAPCOG s 2013 network analysis) * ; and The two ozone monitoring stations in Bell County, one of which covered a portion of the region in the 2013 analysis, and one of which had failed to be included in the 2013 analysis. Including all of these monitoring stations was necessary in order to ensure that the entirety of the CAPCOG region was assigned to the ozone monitoring station nearest to it. For the population served analysis, CAPCOG used 2010 census block group data the lowest level of population data available and assigned each census block group in the region to a Thiessen polygon based on the centroid location of the block group. * http://www.capcog.org/documents/airquality/reports/2013/capcog_air_quality_monitoring_2013_netwo RK_ASSESSMENT.pdf Shapefile data obtained from the U.S. Census Bureau s Topologically Integrated Geographic Encoding and Referencing (TIGER) Products website (http://www.census.gov/geo/maps-data/data/tiger.html) on February 11, 2014. 25
Figure 5: Area and Population Served by Each Ozone Monitor in the CAPCOG Region 26
FY14-15 PGAFY14-1 Task 3.1 The table below summarizes the land area and population served for each monitoring station included in the analysis. Table 13: CAPCOG Land Area and Population Covered by Regional Ozone Monitors Monitor Owner Land Area (sq. miles) Land Area Rank Population, 2010 Population Rank CAMS 3 TCEQ 260 10 607,118 1 CAMS 38 TCEQ 994 4 246,257 3 CAMS 601 CAPCOG 1,355 2 46,974 9 CAMS 614 CAPCOG 2,071 1 77,929 8 CAMS 684 CAPCOG 1,138 3 100,272 6 CAMS 690 CAPCOG 627 6 143,791 5 CAMS 1675 CAPCOG 274 9 79,861 7 CAMS 6602 CAPCOG 756 5 198,547 4 Lockhart CAPCOG 585 7 39,945 10 Gorzycki MS CAPCOG 294 8 286,566 2 CAMS 504 AACOG 44 12 251 12 CAMS 1047 TCEQ 233 11 1,820 11 In comparing these data to the data presented in CAPCOG s 2013 analysis, there are several differences to note that account for the differences in the population covered by some of the monitoring stations. The first substantive is the inclusion of the Gorzycki MS and Lockhart stations, which significantly changed the land and population coverage for the southern portion of the region. As a result of the inclusion of these two monitoring station in the region, CAMS 3 will cover 30% of the population for the region, rather than the 45% of the region s population it covered before these monitoring stations were added. Similarly, the populations covered by CAMS 614, 684, and 1675 were reduced by 40%, 19%, and 45%, respectively, while the area covered by CAMS 1675 is 50% smaller and the area covered by CAMS 684 is 21% smaller than it was before the addition of these sites. The other substantive change of note is the use of census block group data for this analysis, as opposed to the use of census tract data for the 2013 analysis. This change resulted in some census block groups being assigned to different monitoring station than their corresponding census tract were assigned to in the 2013 analysis. Since the census block group is a lower level of geographic aggregation than the census tract, CAPCOG believes that it provides a better basis for performing this type of analysis. CAPCOG put the land area served and population served for each station onto 100-point scales, with the monitoring site with the largest land area or population served getting assigned a value of 100 and the other scores representing fractions of that amount. Since the monitoring station that covered the largest population in the CAPCOG region was TCEQ s CAMS 3, CAPCOG needed to use the station s population coverage as the basis for the 100-point scale. The calculated area served and population served scores for CAMS 684 are shown below.
The scores for all eight of CAPCOG s monitoring stations are shown below. Table 14: Land Area and Population Served Scores for CAPCOG Monitoring Stations. Monitor Owner Land Area Score Population Score CAMS 601 CAPCOG 65 8 CAMS 614 CAPCOG 100 13 CAMS 684 CAPCOG 55 17 CAMS 690 CAPCOG 30 24 CAMS 1675 CAPCOG 13 13 CAMS 6602 CAPCOG 37 33 Lockhart CAPCOG 28 7 Gorzycki MS CAPCOG 14 47 2.5: Correlation Analysis The EPA recommends that a network analysis includes an analysis of the correlation of monitoring values between sites to determine if any of the sites are duplicative of data collected at another site. The less two sites are correlated, the more the monitors are providing unique information relative to one another. The EPA specifically suggests that any pair of monitors with concentrations that correlate with a coefficient over 0.75 may be duplicative. CAPCOG performed an analysis for all of the monitors stationed in Region 11 using monitoring data collected from 2007 2013 to assess whether any of the monitors were duplicative when 8-hour ozone concentrations of over 60 ppb, 65 ppb, 70 ppb, or 75 ppb were measured at one or more stations in the region. These thresholds were used to correspond to the specific thresholds OAQPS used in its most recent draft policy assessment for the new ozone standard to evaluate potential public health impacts at varying eight-hour ozone concentrations. The correlation values represent the correlation between two sites maximum 8-hour ozone concentrations on any day when at least one station had an 8-hour ozone concentration recorded above one of those thresholds, irrespective of whether the two sites had ozone levels above those thresholds. The following four tables below show the correlation coefficients among paired monitoring stations. Since the Gorzycki Middle School site was only operational during one month of this period, it was excluded from this analysis, and since the Lockhart site only recorded four days of measurements on days when at least one monitor measured over 75 ppb, it was excluded from that analysis. While CAPCOG didn t include them in this year s correction analysis, CAPCOG did include any data collected on days that the 2012 temporary Liberty Hill and Elroy sites recorded ozone averages above those thresholds. Coefficient values over 0.75 are marked in red. A spreadsheet that includes the data used in this analysis is being submitted along with this report. The total number of days included in each analysis is included in each table s caption, although that number does not reflect the number of paired days for each set of 28
CAMS 0003 CAMS 0038 CAMS 0601 CAMS 0614 CAMS 0684 CAMS 0690 CAMS 1675 CAMS 6602 Lockhart CAMS 0003 CAMS 0038 CAMS 0601 CAMS 0614 CAMS 0684 CAMS 0690 CAMS 1675 CAMS 6602 Lockhart stations. The Gorzycki site was excluded due to the small number of days of data, and the Lockhart site was excluded from the > 75 ppb analysis because of the small number of days of data for that specific analysis. Table 15: Correlation Coefficients for Monitors on Days Over 60 ppb 2007-2013 (n = 317) CAMS 0003 1.00 CAMS 0038 0.76 1.00 CAMS 0601 0.45 0.38 1.00 CAMS 0614 0.63 0.66 0.48 1.00 CAMS 0684 0.61 0.39 0.51 0.51 1.00 CAMS 0690 0.57 0.74 0.27 0.36 0.24 1.00 CAMS 1675 0.68 0.59 0.72 0.71 0.63 0.39 1.00 CAMS 6602 0.79 0.71 0.68 0.67 0.59 0.66 0.60 1.00 Lockhart 0.46 0.34 0.41 0.65 0.68 0.15 0.58 0.38 1.00 AVERAGE 0.62 0.57 0.49 0.58 0.52 0.42 0.61 0.63 0.46 Table 16: Correlation Coefficients for Monitors on Days Over 65 ppb (n = 186) CAMS 0003 1.00 CAMS 0038 0.70 1.00 CAMS 0601 0.32 0.25 1.00 CAMS 0614 0.57 0.61 0.45 1.00 CAMS 0684 0.52 0.29 0.45 0.49 1.00 CAMS 0690 0.37 0.63 0.18 0.18 0.09 1.00 CAMS 1675 0.53 0.38 0.61 0.63 0.50 0.13 1.00 CAMS 6602 0.70 0.57 0.59 0.60 0.48 0.55 0.40 1.00 Lockhart 0.41 0.30 0.37 0.72 0.71 0.07 0.54 0.31 1.00 AVERAGE 0.51 0.47 0.40 0.53 0.44 0.28 0.47 0.53 0.43 29
CAMS 0003 CAMS 0038 CAMS 0601 CAMS 0614 CAMS 0684 CAMS 0690 CAMS 1675 CAMS 6602 CAMS 0003 CAMS 0038 CAMS 0601 CAMS 0614 CAMS 0684 CAMS 0690 CAMS 1675 CAMS 6602 Lockhart Table 17: Correlation Coefficients for Monitors on Days Over 70 ppb (n = 90) CAMS 0003 1.00 CAMS 0038 0.65 1.00 CAMS 0601 0.28 0.25 1.00 CAMS 0614 0.48 0.55 0.53 1.00 CAMS 0684 0.41 0.13 0.35 0.43 1.00 CAMS 0690 0.26 0.53 0.08 0.09-0.06 1.00 CAMS 1675 0.39 0.25 0.68 0.57 0.35-0.08 1.00 CAMS 6602 0.61 0.49 0.55 0.51 0.39 0.44 0.28 1.00 Lockhart 0.18 0.13 0.33 0.60 0.60 0.02 0.53 0.26 1.00 AVERAGE 0.41 0.37 0.38 0.47 0.32 0.16 0.37 0.44 0.33 Table 18: Correlation Coefficients for Monitors on Days Over 75 ppb (n = 36) CAMS 0003 1.00 CAMS 0038 0.43 1.00 CAMS 0601 0.01 0.00 1.00 CAMS 0614 0.15 0.29 0.31 1.00 CAMS 0684 0.43 0.11 0.28 0.36 1.00 CAMS 0690 0.00 0.30 0.09-0.40-0.26 1.00 CAMS 1675 0.23 0.06 0.78 0.45 0.33-0.29 1.00 CAMS 6602 0.35 0.24 0.45 0.19 0.17 0.23 0.21 1.00 AVERAGE 0.23 0.20 0.27 0.19 0.20-0.05 0.25 0.26 CAPCOG also calculated the average correlation coefficient for each monitor across all four of the 8-hour ozone concentration thresholds evaluated, excluding the self-correlation statistic, in order to provide a composite metric that could be used to rank and score each of CAPCOG s monitoring station across all four thresholds. CAPCOG then calculated a composite correlation coefficient score for each of its monitoring stations by subtracting the average correlation coefficient from 1 and multiplying the result by 100. Since a correlation coefficient of 1 means perfect correlation, the further a correlation coefficient is from 1, the 30
more unique that monitor s data are, and the more valuable that monitor s data would presumably be. An example calculation for CAMS 690 is shown below. ( ) ( ) Table 19: Average Correlation Coefficients Among Monitors and Corresponding Scores Site > 60 ppb > 65 ppb > 70 ppb > 75 ppb Average Score CAMS 0601 0.49 0.40 0.38 0.27 0.39 61 CAMS 0614 0.58 0.53 0.47 0.19 0.45 55 CAMS 0684 0.52 0.44 0.32 0.20 0.37 63 CAMS 0690 0.42 0.28 0.16-0.05 0.20 80 CAMS 1675 0.61 0.47 0.37 0.25 0.43 57 CAMS 6602 0.63 0.53 0.44 0.26 0.47 53 Lockhart 0.46 0.43 0.33 0.41 59 2.6: Probability of Peak and Minimum Ozone Measurements on High Ozone Days Since one of CAPCOG s goals for its monitoring program is to use monitoring data to estimate the background ozone concentrations and the contribution of local emissions to peak ozone measured in the region, CAPCOG performed an analysis of extent to which CAPCOG s monitoring stations measured the region s highest or lowest eight-hour ozone average on days when an eight-hour ozone average exceeded 60 ppb, 65 ppb, 70 ppb, and 75 ppb the thresholds OAQPS used in the 2 nd External Review Draft of its Policy Assessment for evaluating public health effects of ozone exposure in the current review of the ozone standard. This is an analysis that is not included in EPA s recommendations, but data of this nature has been used on CAPCOG s two most recent ozone conceptual models in 2012 14 and 2010. 15 This analysis provides a perspective on the extent to which CAPCOG s monitors can be used to perform this kind of analysis. 2.6.1: Probability of Measuring the Region-wide Maximum 8-Hour Ozone Concentration on High Ozone Days. The following table shows the calculated probability of a given ozone monitor in the region having measured the highest maximum in the CAPCOG region given at least one monitoring site in the region has measured a daily maximum of over 60 ppb, 65 ppb, 70 ppb, or 75 ppb, respectively, from 2007-2013. The total number of days a given station measured the highest 8-hour ozone averages at these thresholds was then divided by the total number of days for which that monitor had an eight-hour ozone average on that subset of high ozone days from 2007 2013. This normalizes the data to account for the differences in the length of time each station was in operation during this period. Example calculations for CAMS 601 and 614 for days with at least one regional station measuring over 75 ppb are show below: 14 http://www.capcog.org/documents/airquality/reports/2013/task_1-austin_area_conceptual_model_2012.pdf 15 http://www.capcog.org/documents/airquality/cac/2010/september2010/austin_cm_ver21.pdf 31
( ) ( ) ( ) ( ) ( ) ( ) The table below includes all monitoring stations that were operational for any period between 2007 and 2013, except for Gorzycki Middle School, which was only operational for a little over a month. The score is calculated by taking the average probability for all four thresholds and multiplying the result by 100. Table 20: Probability of a Monitoring Station Measuring the Region's Highest 8-Hour Ozone Average on High Ozone Days Station 60 ppb 65 ppb 70 ppb 75 ppb Score Fayette County C601 8% 8% 7% 3% 7 Dripping Springs School C614 10% 10% 12% 22% 14 CAPCOG Round Rock C674 10% 10% 7% 7% 8 CAPCOG San Marcos C675 8% 9% 14% 14% 11 CAPCOG McKinney Roughs C684 11% 8% 6% 0% 6 CAPCOG Lake Georgetown C690 24% 26% 19% 19% 22 CAPCOG San Marcos Staples Road C1675 20% 21% 24% 23% 22 CAPCOG Hutto College Street C6602 6% 7% 10% 10% 8 Liberty Hill 2012 38% 25% 22% 17% 26 Elroy 2012 0% 0% 0% 0% 0 Lockhart 2013 0% 0% 0% 0% 0 These data show that some locations are significantly more important in capturing the region s peak 8- hour ozone maximum at the lower end of the 8-hour ozone threshold range than they are at the higher end of the range, such as Liberty Hill, while others are much more important on the higher end of the range than the bottom, such as Dripping Springs. The importance of the relocated San Marcos site is particularly notable. 2.6.2: Probability of Measuring the Region-wide Minimum 8-Hour Ozone Average on High Ozone Days Calculating the difference between the peak ozone level measured in the region and the lowest ozone levels is a technique that has previously been used by the University of Texas at Austin in the region s ozone conceptual models to estimate the local contribution to peak ozone levels. This technique provides an estimated local contribution that is similar to the average local contribution modeled by UT using the June 2006 photochemical modeling episode (10-20 ppb). 16 Calculating the probability that a 16 http://www.capcog.org/documents/airquality/reports/2013/task_8.3-apca_analysis_final.pdf 32
given monitoring station measures the lowest peak eight-hour ozone average on a day when at least one station in the region measured a high eight-hour ozone concentration provides a useful way to estimate the extent to which that monitoring station would be useful for approximating the local contribution to peak ozone levels on high ozone days using monitoring data. The following table shows the calculated probability of each of CAPCOG s monitoring stations measuring the lowest peak 8-hour ozone average on the same subset of days used in the previous section to calculate the probability of measuring the highest peak 8-hour ozone average in the region. Gorzycki Middle School is excluded from this analysis due to the small amount of data that was actually collected there in 2013. These probabilities were then averaged across all four thresholds, and scoring was conducted the same way it was for the probability of measuring the region-wide maximum eight-hour ozone average on high ozone days. An example calculation for CAMS 601 is shown below. ( ) ( ) ( ) Table 21: Probability of a Monitoring Station Measuring the Region's Lowest Peak 8-Hour Ozone Average on High Ozone Days Station 60 ppb 65 ppb 70 ppb 75 ppb Score Fayette County C601 38% 40% 39% 41% 40 Dripping Springs School C614 14% 14% 14% 17% 15 CAPCOG Round Rock C674 7% 4% 2% 7% 5 CAPCOG San Marcos C675 26% 19% 14% 5% 16 CAPCOG McKinney Roughs C684 10% 11% 13% 9% 11 CAPCOG Lake Georgetown C690 13% 11% 11% 3% 9 CAPCOG San Marcos Staples Road C1675 5% 6% 8% 8% 7 CAPCOG Hutto College Street C6602 9% 10% 8% 10% 9 Liberty Hill 2012 6% 10% 11% 17% 11 Elroy 2012 8% 13% 20% 25% 17 Lockhart 2013 10% 14% 10% 0% 9 These rankings clearly show the importance of the Fayette County monitor in understanding the background ozone levels for the region. The importance of CAMS 614 for measuring the minimum peak eight-hour ozone average is somewhat surprising, especially when compared to CAMS 684, given their respective positions relative to the main urbanized areas of the region and the large point sources in the region. 33
2.7: Jurisdictional Value for Ozone Action Program Action Plan Counties This score is based on the jurisdictional value of a CAPCOG monitoring station to the County governments that participate in the region s Ozone Advance Program (OAP) Action Plan. In recent years, CAPCOG has successfully engaged participants in the plan to help fund the costs of monitors in their jurisdictions, and the city and county governments that host CAPCOG monitoring stations value having a CAPCOG monitor in their jurisdiction. While a concept such as jurisdictional value has not been analyzed in previous network analyses, it is the only way to consider the value to our local communities of having a monitor that they consider their own that gives their specific communities data that is unique to them. This analysis enables this benefit to be considered alongside all of the other analyses included in this document. For this scoring, CAPCOG assigned a score of 100 to any monitoring station that serves as CAPCOG s only monitoring station in a county that participates in the OAP Action plan, and divides that number by however many total monitoring stations CAPCOG fields within that County. For example, since CAMS 684 is CAPCOG s only monitoring station in Bastrop County, it has a score of 100, while CAMS 614 is only one of two CAPCOG monitoring stations in Hays County, so it gets a score of 50. Since Fayette County does not participate in the OAP Action Plan, the monitoring station CAPCOG operates there gets a score of zero for this purpose. Table 22: "Political Value" Score Assigned to Each Monitoring Station Station County OAP Action Plan Participant CAPCOG Ozone Monitors in County Score CAMS 601 Fayette No 1 0 CAMS 614 Hays Yes 2 50 CAMS 684 Bastrop Yes 1 100 CAMS 690 Williamson Yes 2 50 CAMS 1675 Hays Yes 2 50 CAMS 6602 Williamson Yes 2 50 Lockhart Caldwell Yes 1 100 Gorzycki Middle School Travis Yes 1 100 Section 3: Bottom-Up Analysis This section provides bottom-up analysis of the suitability of the monitoring network to broadly meet CAPCOG s goals. These analysis techniques not only provide an opportunity to compare the monitoring stations against one another, but also to assess the overall performance of the network and an opportunity to identify alternative configurations that may enable CAPCOG to better meet its goals. The following table lists the various techniques identified in the EPA guidance and the goals that these techniques assess. 34
One other type of bottom-up analysis which can be used for these purposes and which was used for the 2013 monitoring analysis is mobile monitoring. Since the 2013 network analysis 17 involves a more extensive discussion of that, and since it would be difficult to come up with a quantitative way to score those results for this analysis, CAPCOG is not including a discussion of those results in the current report. Table 23: Bottom-Up Analysis Techniques Technique Objectives Assessed Included Emission Inventory Emission reduction evaluation Maximum precursor location Population Density Population exposure Environmental justice Population Change Population exposure Environmental justice Maximum precursor location Suitability Modeling Population Exposure Environmental justice Source-oriented Model evaluation Maximum concentration location Background concentration Transport/border characterization Photochemical Modeling Maximum concentration location Source-oriented Transport/border characterization Population exposure Background concentration 3.1: Emissions Inventory Analysis The EPA guidance on bottom-up analyses that can be conducted to determine monitoring needs includes an emissions inventory analysis. EPA describes this type of analysis as follows: Emissions inventory data are used to find locations where emissions of pollutants of concern are concentrated. These locations can be compared to the current or proposed network. Does the network capture the areas of maximum emissions? This analysis can be scaled to various levels of complexity, depending on resources. The simplest version looks at county-level emissions of a single pollutant. More complex methods use gridded emissions and/or species-weighted emissions, depending on their importance in producing the secondary pollutant of concern. As this description indicates, this type of bottom-up analysis is usually better suited to analyzing the suitability of a monitoring network to measure primary pollutant concentrations, and that applying such analyses to a secondary pollutant like ozone is more complex. Given the complexity involved, CAPCOG 17 http://www.capcog.org/documents/airquality/reports/2013/capcog_air_quality_monitoring_2013_netwo RK_ASSESSMENT.pdf 35
evaluated whether to include this analysis or not. Ultimately, since this is the only analysis involved in this network assessment that directly looks at emissions data, CAPCOG felt that it was valuable to include some kind of analysis of the monitoring stations in the context of emissions data, even with the limits on its utility. As is described in Section 5 of this report, this analysis only accounted for 7.5% of the overall score for each station. The method CAPCOG developed to use the emissions inventory data is relatively coarse, and is therefore not given nearly the weight of the photochemical modeling analysis, which accounts for 15% of the total score. The general approach CAPCOG took to conducting this emissions inventory analysis was to compare the extent to which a given monitor may be positioned downwind of anthropogenic emissions in nearby counties in light of the wind directions that occur most frequently on high ozone days. CAPCOG focused on NO X emissions because anthropogenic local NO X emissions have 30-100 times the influence of anthropogenic VOC emissions on peak eight-hour ozone averages in Travis County, according to a modeling sensitivity analysis performed by UT-Austin for CAPCOG in 2012. 18 CAPCOG used both the 2012 Conceptual Model 19 and the 2010 Conceptual Model 20 to provide insight into the wind directions that are most important for high ozone. The 2010 Conceptual Model evaluated data collected from 2001-2009, and involved analyzing high ozone at three different levels 60 ppb, 65 ppb, and 70 ppb. The distribution of the resultant wind direction from 6 am to 6 pm on days when CAMS 3 measured at or above each of these thresholds is shown below. 18 http://www.capcog.org/documents/airquality/reports/2013/task_8.3-precursor_response_runs_final.pdf 19 The University of Texas at Austin, Center for Energy and Environmental Resources. Conceptual Model for Ozone for the Austin Area. July 2012. 20 The University of Texas at Austin, Center for Energy and Environmental Resources. Conceptual Model for Ozone for the Austin Area. July 2010. 36
Figure 6: Number of Days with Resultant Wind Direction from 6:00 18:00 CST at CAMS 3 on Days >= 60, 65, and 70 ppb, 2001-2009 WNW W WSW NW NNW N 80 70 60 50 40 30 16 9 3 4 20 10 2 0 6 2 12 45 NNE NE 55 45 30 26 ENE E ESE 70+ ppb 65+ ppb 60+ ppb SW 63 SE SSW 60 S SSE 76 These data show that the most important wind directions for ozone formation at CAMS 3 are from NNE clockwise through S for all levels, but that due NE and SSE are the single most important directions for the 60 and 65 ppb thresholds, while SE and SSE are the most important for the days at or over 70 ppb. The 2012 Conceptual Model included data from 2006 2011, and looked only at days with ozone levels at or above 75 ppb from 2006-2011. The figure below shows the distribution of resultant wind direction for both CAMS 3 and CAMS 38. 37
Table 24: Number of Days with Resultant Wind Direction from 6:00-18:00 CST at CAMS 3 and 38 on Days >= 75 ppb 2006-2011 WNW W NW NNW 8 7 6 5 4 1 3 2 0 0 1 0 0 0 1 0 N 2 4 2 1 NNE 3 4 NE ENE E CAMS 3 CAMS 38 WSW ESE SW 3 5 SE SSW S SSE 8 The data from both of the conceptual models support a general conclusion that the SSE direction is the most important upwind region on high ozone days and the NNE to ENE direction is the second most important upwind region for ozone, and that there are very few days when winds from any western direction from SSW to NNW result in high ozone at the region s two regulatory monitors. Both to the south-southeast and to the northeast, there are some significant emissions sources that modeling performed by UT in 2012 21 and modeling performed by AACOG in 2013 22 indicate may be contributing to the region s ozone levels. CAPCOG s monitoring stations may be able to help assess the impacts of emissions from these sources on local ozone levels. Oil and gas exploration and production has increased dramatically to the south of the Austin urban core. AACOG has estimated that the scale of the increase in activity in recent years has resulted in in significant increases in NO X and VOC emissions in an area that is often upwind of the Austin area on high ozone days. At this point, the Lockhart site provides the best opportunity to measure any impact from this activity on ozone levels entering the MSA. The monitoring stations in Hays County are too likely to be influenced by emissions sources in Guadalupe and Comal County unrelated to the Eagle Ford Shale emissions when winds are coming out of the SSE. CAPCOG s monitoring stations at McKinney Roughs and Fayette County are probably too far east to be detect any influence from the Eagle Ford Shale. The figure below shows the distribution of wells in the Eagle For Shale play. 21 http://www.capcog.org/documents/airquality/reports/2013/task_8.3-apca_analysis_final.pdf 22 AACOG. Future Year Photochemical Modeling, 2012 and 2018 for the Capital Area Council of Governments. December 15, 2013. <available upon request> 38
Figure 7: Wells Permitted and Completed in the Eagle Ford Shale Play The other major emissions inventory consideration for monitoring would be the location of monitoring stations relative to the locations of major point sources. The figure below shows the locations of all major point sources in the region down to 70 tons per year in 2012. Their level of emissions is differentiated by color, with red representing point sources that emitted over 1,000 tons of NO X emissions in 2012, orange points representing point sources that emitted 500 1,000 tons per year, yellow points representing point sources that emitted 100 500 tons per year, and green points representing point sources that emitted 70-100 tons per year. The locations of the two TCEQ regulatory monitors in Travis County are shown in white, and CAPCOG s eight monitoring stations are shown in green. Overlaid on this map is the wind rose from the 2010 Conceptual Model. 39
Figure 8: Positions of CAPCOG Ozone Monitors and Major Point Sources of NO X in the Region Overlaid with CAMS 3 Wind Rose 40
This map shows the utility of several monitoring stations in particular for measuring upwind ozone concentrations and sites that could be influenced by specific industrial plumes. The table below shows the NO X and VOC emissions from the 2011 National Emissions Inventory for selected counties that are adjacent to CAPCOG counties or close enough to a monitoring station and would often be upwind of one of CAPCOG s regional monitoring stations, based on the region s conceptual model. Counties that were adjacent to a CAPCOG county with a monitor in it that would be located upwind of that monitor were included in the analysis. DeWitt County was also included in the analysis because the oil and gas activity that accounts for over half of the county s 2011 NO X emissions were clustered along the border shared with Gonzales County, which is adjacent to Bastrop, Caldwell, and Fayette Counties. That county border is only 42-52 miles away from the Lockhart site, which is shorter than the distance between the Fayette Power Plant and CAMS 3. Table 25: 2011 NO X and VOC Emissions by County and Land Area - Selected Counties County NO X (tons) VOC (tons) Austin 3,655.78 3,045.16 Bastrop 4,934.39 26,655.78 Bell 8,930.91 8,692.98 Caldwell 2,650.40 10,398.62 Colorado 4,491.48 6,005.62 Comal 7,273.80 4,385.95 DeWitt 3,635.55 20,710.07 Fayette 10,863.14 5,216.79 Gonzales 2,792.48 9,041.27 Guadalupe 4,919.28 8,707.13 Hays 7,471.09 4,048.71 Lavaca 2,580.45 3,475.89 Lee 1,741.78 3,918.81 Milam 5,403.99 4,988.98 Travis 18,590.81 20,197.06 Washington 2,352.27 2,829.98 Williamson 7,257.46 7,849.02 In order to incorporate the emissions data into this analysis, CAPCOG assigned NO X emissions from the 2011 NEI from various nearby counties that could have an influence on peak ozone levels based on the distribution of wind directions on high ozone days described above and proximity of those emissions to the monitor. Counties located to the north/northeast through south (clockwise) of the monitor were assigned to each station, as was the county the monitor was located itself. 41
CAMS 601 CAMS 614 CAMS 684 CAMS 690 CAMS 1675 CAMS 6602 Lockhart Gorzycki Table 26: Assignment of NO X Emissions from Nearby Counties to Monitoring Stations for Emissions Inventory Analysis COUNTY NO X (tons) Austin 3,656 X Bastrop 4,934 X X X X X Bell 8,931 X Caldwell 2,650 X X X X X Colorado 4,491 X Comal 7,274 X DeWitt 23 3,636 X X X Fayette 10,863 X X X X Gonzales 2,792 X X X X Guadalupe 4,919 X X X Hays 7,471 X X X Lavaca 2,580 X Lee 1,742 X X X Milam 24 5,404 X X X X Travis 18,591 X X X X Washington 2,352 X Williamson 7,257 X X X Total 99,545 30,371 40,905 32,022 46,859 33,632 35,644 27,145 46,308 In a very coarse way, these data represent the potential value of each monitoring station for measuring ozone impacts from NO X emissions in and around the Austin-Round Rock MSA and Fayette County. CAPCOG acknowledges that this approach is not perfect, but was the only readily available analysis technique CAPCOG was able to develop that was could incorporate the most recent NEI data without conducting new photochemical modeling, which would have been a costly and time-consuming endeavor that could not have been completed in time to be included in this report. Readers are cautioned not to only use this type of analysis in conjunction with a number of other analyses, and assign an appropriate weight to an analysis like this given its limitations. CAPCOG welcomes any and all suggestions on how to improve this analysis technique or alternative techniques to incorporate new emissions inventory data into an evaluation of the ozone monitoring network for future analyses. CAPCOG translated the total emissions assigned to each monitoring station into a score on a 100-point scale based on the percentage of the total emissions from all counties included in the analysis. An example calculation for the Gorzycki Middle School site is shown below: 23 DeWitt County boundary is 42-52 miles away the Lockhart site 24 Milam County boundary 21-24 miles away from CAMS 6602 42
The scores for each site are listed below. Table 27: Emissions Inventory Scores Assigned to Monitoring Stations Station NO X Emissions Assigned Score CAMS 601 30,371 31 CAMS 614 40,905 41 CAMS 684 32,022 32 CAMS 690 46,859 47 CAMS 1675 33,632 34 CAMS 6602 35,644 36 Lockhart 27,145 27 Gorzycki 46,308 47 3.2: Photochemical Modeling Analysis Modeling results can provide a useful way to assist in the review of the current monitoring network configuration, estimate upwind and downwind ozone concentrations in the region, and estimate the physical extent of the ozone plumes generated from the urban core of Austin. To the extent that the modeling results are consistent with the region s ozone conceptual model, the two sets of data provide more robust conclusions about the monitoring network and can serve as a more robust basis for making recommendations for the network. In 2012, CAPCOG contracted with the University of Texas to perform various photochemical modeling tasks with the June 2006 ozone episode. For CAPCOG s 2013 monitoring network analysis, CAPCOG obtained the data that was used for UT s modeling and obtained graphical displays of the maximum 8- hour ozone concentrations for each 4 km x 4 km grid cell covering Central Texas. UT plotted the location of all of the ozone monitors in operation in 2012 within the region on each tile plot for comparison. The plots include CAMS 3, 38, 601, 614, 684, 690, 1675, 6602, a temporary site in Liberty Hill, and a temporary site in Elroy, all of which were operational in 2012, but did not include the Lockhart and Gorzycki Middle School sites that were operated in 2013. The Gorzycki Middle School site in particular was set up in order to address deficiencies that were identified in the review of these photochemical modeling results in CAPCOG s 2013 monitoring network review. 25 A simple visual review of some of the tile plots suggested that the highest eight-hour ozone averages in the region may at times occur in an area of Southwest Austin that may not be measured properly by the network as it was configured in 2012, as the next two maps show. 25 CAPCOG. Capital Area Council of Governments Ozone Monitoring Network Assessment. February 2013. http://www.capcog.org/documents/airquality/reports/2013/capcog_air_quality_monitoring_2013_netwo RK_ASSESSMENT.pdf 43
Figure 9: Modeled Daily Maximum 8-Hour Ozone Concentrations, June 3, 2006 Liberty Hill CAMS 690 CAMS 6602 CAMS 38 CAMS 3 CAMS 614 CAMS 684 Elroy CAMS 1675 44
Figure 10: Modeled Maximum 8-Hour Ozone Concentrations, June 13, 2006 Liberty Hill CAMS 690 CAMS 6602 CAMS 38 CAMS 3 CAMS 614 CAMS 684 Elroy CAMS 1675 The new Gorzycki Middle School site is well-positioned to measure ozone within the two plumes shown above. Those types of conditions are quite common later in the summer when the region tends to see its highest ozone levels. Other plots prepared for the 2013 analysis showed that the Lake Travis Area may also occasionally have elevated levels of ozone, and also showed that the Fayette monitor is still likely a good monitor to measure estimate background from the Southeast, as the next few maps show. Several of the maps also showed the distinct influence of the Alcoa/Sandow plant s emissions on ozone levels in eastern Williamson County (see the plot for June 3 and June 14, for example) the only source of significance in 45
that part of the region is the Alcoa/Sandow plant located in southern Milam County, but following 2006, the emissions from the plant decreased dramatically. The sensitivity modeling conducted by UT in 2012 included an analysis of the potential impact of that change on ozone levels in the region. 26 Figure 11: Daily Maximum 8-Hour Ozone, June 29, 2006 Liberty Hill CAMS 690 CAMS 6602 CAMS 38 CAMS 3 CAMS 614 CAMS 684 Elroy CAMS 1675 26 http://www.capcog.org/documents/airquality/reports/2013/task_8.3-precursor_response_runs_final.pdf 46
Figure 12: Daily Maximum 8-Hour Ozone, June 14, 2006 Liberty Hill CAMS 690 CAMS 6602 CAMS 38 CAMS 3 CAMS 614 CAMS 684 Elroy CAMS 1675 For the 2013 analysis, CAPCOG created a map that showed isopleths of the values of the 4 th highest eight-hour ozone average for each grid cell. The image below shows a interpolated map of Bastrop, Burnet, Caldwell, Hays, Travis, and Williamson Counties, along with the fourth highest eight-hour ozone average. For reference, CAPCOG has added the approximate locations of the Gorzycki Middle School and Lockhart sites, which are denoted by a star. 47