determining tropospheric ozone trends in pennsylvania from ...significance in variation among three...
TRANSCRIPT
DISCUSSION Looking back at the four focus questions: 1) Sufficient data: There may look to be sufficient monitoring stations, but the actual consistency
of CP data is sparse, inaccessible, and variable. Data in rural areas (where fracking is heavy) is very limited.
2) How to improve: More sites in rural areas and those closer to fracking activities would give a better picture of air quality in relationship to natural gas extraction.
3) Significant ozone: Location and time interval both play roles in tropospheric ozone variation. Ozone trends tend to follow national patterns.
4) Association with fracking: Active wells do seem to play some part in O3 trends. The entire process of hydraulic fracturing needs to be assessed and included in further studies to determine if fracking as an entity leads to higher tropospheric ozone levels. Wells alone do not account for all patterns.
Determining Tropospheric Ozone Trends in Pennsylvania from 2007 to 2012
€
Kayla Kelly - Slatten1, Kenneth Davis2
€
1) California State University, Fresno; Dickinson School of Law, Pennsylvania State University 2) Department of Meteorology, College of Earth and Mineral Sciences, Pennsylvania State University
ABSTRACT There has been little research conducted and published on hydraulic fracturing’s effect on air quality, specifically on the six NAAQS Criteria Pollutants. This preliminary study attempts to assess current air quality monitoring systems and utilize tropospheric ozone data to determine whether or not there is a significance in variation among three sites in Pennsylvania from 2007 to 2012. It also looks at the possible trend between tropospheric ozone levels and the number of active well sites within the state during the same time period. Although consistent, agreeable, and easily accessible data is difficult to come by, a 2-Way ANOVA shows significance in variation based on location and time. Furthermore, Pearson’s Correlation Coefficient depicts some positive relationship between tropospheric ozone levels and the number of active fracking well sites. A completely comprehensive air quality study would take into account all aspects of the fracking process, not just wells.
BACKGROUND & FOCUS QUESTIONS
The State of Pennsylvania is among the leading producers of natural gas via conventional and unconventional hydraulic fracturing mechanisms. “Fracking” is a highly controversial alternative fuel extraction process with little scientific data published in regards to its effects on general air quality.
Under the 1970 Clean Air Act (CAA), Congress granted the Environmental Protection Agency (EPA) authorization to regulate air quality and emissions. The National Ambient Air Quality Standards (NAAQS) provides national thresholds and monitoring systems for harmful and/or hazardous air pollutants.
Fracking is thought to be associated with an increase in various pollutants, including ozone. Many ozone precursors, such as volatile organic carbons (VOCs), are believed to be emitted throughout the fracking process. Thus, the following questions have been posed in order to attempt a better understanding of air quality, taking particular interest in ozone, in a state heavy with fracking activity: 1) Does there seem to be sufficient air quality data available to assess trend patterns in the State of
Pennsylvania for a specific time frame? 2) If there is not sufficient data, what steps could be taken to improve availability within the state? 3) Utilizing available ozone data, is there a significance in ozone patterns from 2007 to 2012 among the
testing sites: Scranton, Reading, and Erie? 4) Is there a trend in tropospheric ozone levels in association with fracking well sites?
METHODS Using maximum 8-hour daily values from the EPA Air Quality System (AQS), tropospheric ozone data was downloaded and input into Excel for Scranton, Reading, and Erie, Pennsylvania. The three sites are controlled by the Pennsylvania Department of Environmental Protection, providing daily ozone readings from 2007 to 2012. The testing period within each year is from April 1st to October 31st, leaving out winter and spring months when solar activity is low, which results in minute photochemical ozone production. The selection of the research sites was largely in part to eliminate variability in interannual weather conditions. Furthermore, when considering Pennsylvanian wind patterns, Erie should typically be upwind, while Scranton tends to be downwind. In regards to fracking, all have different proximities to active wells, with Scranton being directly in the hydraulic fracturing belt along the Marcellus Shale Beds and Reading lying in the lower portion of the state, farthest from fracking activities.
With the maximum 8-hour daily ozone values, the sites having multiple air quality monitoring stations were averaged across the board for each day. Maximum values within each month of data were also taken. Normalized data was used to conduct a two-way ANOVA statistical analysis to determine if there is a significance in variation among the three sites over the years. Maximums and averages were graphed by month/year/site to evaluate any abnormalities and/or trends. Lastly, ozone levels were graphed in relation to the number of active fracking wells (both conventional and unconventional) on an annual basis to depict any patterns over the years of increased methane production. The well numbers were acquired from the DEP Office of Oil and Gas Management SPUD Data.
RESULTS CONTINUED
ACKNOWLEDGMENTS My genuine thanks goes out to the entire SCRiM family for granting me this opportunity. Thank you to Dr. Beth Weinman for presenting and encouraging me to take on this summer research experience. Lastly, my love and gratitude go to my amazing California family and supportive boyfriend for standing by me on my journey to the East Coast.
RESULTS
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
2007 2008 2009 2010 2011 2012
[O3] ppm
Year
Monthly Ozone Averages 2007-‐2012 Scranton
April May June July August September October EPA
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
2007 2008 2009 2010 2011 2012
[O3] ppm
Year
Monthly Ozone Maximums 2007-‐2012 Scranton
April May June July August September October EPA
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
2007 2008 2009 2010 2011 2012
[O3] ppm
Year
Monthly Ozone Averages 2007-‐2012 Reading
April May June July August September October EPA
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
2007 2008 2009 2010 2011 2012
[O3] ppm
Year
Monthly Ozone Maximums 2007-‐2012 Reading
April May June July August September October EPA
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
2007 2008 2009 2010 2011 2012
[O3] ppm
Year
Monthly Ozone Averages 2007-‐2012 Erie
April May June July August September October EPA
0
0.02
0.04
0.06
0.08
0.1
0.12
2007 2008 2009 2010 2011 2012
[O3] ppm
Year
Monthly Ozone Maximums 2007-‐2012 Erie
April May June July August September October EPA
Clean Air Act (Federal Government and EPA)
Office of Air Quality Planning and
Standards (Criteria for States to meet)
NaOonal Air Monitoring StaOons
(NAMS)
More stringent siOng, equipment,
reports, etc.
Quarterly and Annual Reports to
the EPA
State and Local Air Monitoring StaOons
(SLAMS)
Less stringent than NAMS
Annual reports only to EPA
Special Purpose Monitoring
Short term monitoring for
specific inquiries/requests
Photochemical Assessment
Monitoring StaOons (PAMS)
Regulates O3 precursors; monitors
60 VOCs and Carbonyl
Figure 3: Map of the EPA air monitoring stations in PA. http://www.epa.gov/reg3artd/airquality/r3monitors.htm
Figure 4: Map of DEP air monitoring stations in PA. http://www.dep.state.pa.us/dep/deputate/airwaste/aq/aqm/copams.htm
Figure 5: Hydraulic fracturing well sites and research locales.
Figure 1: NAAQS standards for criteria pollutants.
http://www.epa.gov/air/criteria.html
Figure 2: Flowchart of hierarchal air monitoring in US.
Figure 7: Scranton averages are well below the EPA threshold, and all months tend to follow the same trend.
Figure 8: Scranton maximums vary but 2008, 2010, and 2012 have high peaks for most months.
Figure 9: Reading averages are well below the EPA threshold, but with more variability than Scranton.
Figure 10: Reading maximums are often well above the threshold, with April and October not following the common trend.
Figure 11: Erie averages are well below the EPA threshold, with July deviating a little.
Figure 12: Erie has some of the highest maximums, with 2007 and 2012 having the largest peaks.
0
1000
2000
3000
4000
5000
6000
0.03
0.032
0.034
0.036
0.038
0.04
0.042
0.044
0.046
0.048
0.05
2007 2008 2009 2010 2011 2012
Num
ber of AcHve W
ells
[O3] ppm
Year
Yearly Ozone Averages in RelaHonship to AcHve PA Fracking Wells
Scranton O3 Reading O3 Erie O3 AcOve Wells
y = 2E-‐06x + 0.0353 R² = 0.60011
0.039
0.04
0.041
0.042
0.043
0.044
0.045
0.046
0.047
0.048
0 1000 2000 3000 4000 5000 6000
[O3] ppm
Number of AcHve Wells
RelaHonship Between Average Ozone and Number of AcHve Wells (Without Outliers)
y = 0.0006x2 -‐ 0.0048x + 0.0512 R² = 0.46428
y = 0.0014x2 -‐ 0.0107x + 0.0852 R² = 0.59961
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
0.07
0.075
0.08
4954 4989 2842 3335 3232 2370
[O3] ppm
Number of Wells Based Upon Year (2007-‐2012)
Pennsylvania Ozone Average and Maximum in RelaHon to Number of Statewide AcHve Wells
Ozone Average Ozone Maximum Poly.(Ozone Average) Poly.(Ozone Maximum)
Figure 6: National tropospheric ozone trends from the EPA. http://www.epa.gov/airtrends/ozone.html Figure 13: Site averages tend to follow yearly
well numbers except for in 2012. Figure 14: Without the 2012 outlier, there is a relatively strong positive
correlation between number of wells and tropospheric ozone.
Figure 15: Averages and maximums follow overall national trend pattern as seen in Figure 6.
Figure 16: The 2-Way ANOVA conducted using yearly site ozone averages as the DV; the IV were broken into treatment (city/site)
and block (year). An ANOVA is a statistical analysis of variance among a dataset as broken down by IV.
Based upon time and location, there is a significance in variation among ozone levels.