While flying over Utah collecting air quality data in February, researchers with the Cooperative Institute for Research in Environmental Sciences encountered a day with ideal weather conditions for measuring the air over an oil and gas field. The Uintah Basin, south of Vernal, Utah, had been swept clean by high winds on Feb. 2, so the next day, with steady air patterns above the basin, CIRES researchers flew back and forth over the basin with instruments for measuring methane and other natural gases. Their data indicated that the natural gas wells in that field are leaking 6 to 12 percent of the methane they produced — numbers the researchers calculated by comparing the methane measured, some 120,000 pounds an hour, with the average hourly natural gas production rate in Uintah County as reported to the state. That percent rate of emissions is as much as 10 times higher than estimates from the EPA.
“We expected methane emissions would be detectable, but we did not anticipate levels as high as what we observed,” Colm Sweeney, CIRES research scientist and lead scientist for NOAA’s research lab aircraft program, said in a press release from CIRES.
Methane, a potent greenhouse gas, is the primary constituent of natural gas, according to CIRES. And per molecule, methane is 25 times more effective at trapping heat in the atmosphere than carbon dioxide.
The EPA National Greenhouse Gas Emissions Data, which tracks total annual U.S. emissions and removals, estimates the methane leak rate nationally at less than 1 percent of production. The U.S. Government Accountability Office reports that onshore, federal-lease, oil and gas development operators vent or flare 0.13 percent of produced gas, though EPA and Western Regional Air Partnership estimates show volumes 30 times higher — 3.9 percent, still below the CIRES findings.
The aircraft was part of a multi-agency effort to increase understanding of how emissions from oil and gas extraction affect air quality. That area of Utah, where some 30,000 people live, sees high ozone days that almost double the parts per billion levels seen on the Front Range’s high ozone days (both areas have days that are out of compliance with the EPA standard for ozone levels), says Jessica Gilman, a research chemist with NOAA and CIRES, who participated in that study. Researchers are still working out whether differences in ozone levels are a question of differences in the organisms that died in those locations and spent billions of years underground converting to natural gas with differing compositions, or a question of differences in regulations in the neighboring states.
“It’s actually hard to get some of the information that we need, such as the raw natural gas composition, and it was great here in the Front Range, the COGCC had commissioned an independent study — they went out, collected oil and gas samples from the whole of the Front Range and published it as a free public document, and we relied on that quite heavily,” Gilman says. “So one of the things we’re hoping to get is a similar style of doc from Utah in order to determine, are those differences that we saw in the composition, is it because just different sea creatures died there billions of years ago and got buried, or is it actually industry practices?”
How much to expect in the way of air emissions from various oil and gas production activities is data tracked by the EPA. But the same month that CIRES aircraft were sampling the air above the Uintah Basin, the U.S. EPA Office of Inspector General released a report declaring: “High levels of growth in the oil and natural gas production sector, coupled with harmful pollutants emitted, have underscored the need for EPA to gain a better understanding of emissions and potential risks from the production of oil and gas. However, EPA has limited directly measured air emissions data for air toxics and critical pollutants for several important oil and gas production processes and sources, including well completions and evaporative ponds.”
The land in the Uintah Basin is pockmarked with oil and gas wells. | Google Earth image
Between 1992 and 2010, about 210,000 new gas wells were drilled. The number of producing gas wells increased by 72 percent. In EPA Region 8, which includes Colorado, the number of gas wells increased by 416 percent and gross withdrawals increased by 163 percent.
States look to the EPA for data on the quantity of a pollutant released through oil and gas production activities, and use the agency’s estimates for pollutants emitted in developing emissions inventories, issuing permits and taking enforcement actions — as well as assessing air quality and subsequent human health effects. Airborne emissions of volatile organic compounds, air toxics and methanes are known to be released through vented sources (pneumatic devices, dehydration processes, gas sweetening processes, chemical injection pumps, compressors, tanks and well testing, completions and workovers) as well as through fugitive sources (equipment leaks through valves, connectors, flanges, compressor seals and evaporative sources including wastewater treatment, pits and impoundments). Volatile organic compounds, air toxics and methane have adverse health affects, particularly on lung function, asthma incidence and respiratory illnesses, in addition to contributing to climate change.
Because the data the EPA collects from and provides to states is incomplete in ways that have skewed the cumulative effects of an activity by, in one case, nearly 25 times as many tons, the inspector general’s report says, state agencies looking to that data to regulate ozone-forming pollutants may underestimate the effects the oil and gas industry’s activities have on their air quality and human health in the state.
In the online system that provides data, about half of the EPA oil and gas production emission factors for various emission sources are rated “not applicable” or unrated because of insufficient data — including produced water tanks, well completions and evaporative ponds, which previous reports have suggested have a high evaporation rate, particularly in semiarid regions with hot, dry air moving from the land.
“Limited data from direct measurements, poor quality emission factors, and incomplete [National Emissions Inventory] data hamper EPA’s ability to assess air quality impacts from oil and gas production data,” the EPA Office of Inspector General report reads. “With limited data, human health risks are uncertain, states may design incorrect or ineffective emission control strategies, and EPA’s decisions about regulating industry may be misinformed.”
The EPA did not, at that time, have a comprehensive strategy for improving that data collection and had not developed methods for directly measuring criteria and air toxic pollutant emissions from oil and gas production sources and processes, according to the report. The EPA is working to develop measurement techniques, the report says, but more research is needed.
It will hardly be the first of its kind.
In 2008, a team of NOAA researchers set out to sample air and atmospheric chemistry in northeastern Colorado after the NOAA air monitoring station in Colorado started showing levels of chemical pollutants and methane not found at the seven other stations around the United States. That team, led by Gabrielle Petron, an atmospheric scientist with CIRES and NOAA, took dozens of samples and thousands of readings along rural roads in northeastern Colorado near oil and gas equipment, landfills and animal feeding operations to try to identify the possible sources.
Researchers found oil and gas wells in the Uintah Basin were emitting higher than expected methane levels. | Photo by David Oonk
Their conclusion was that the primary source was oil and gas production in Weld County — and that gas operations in that region were leaking twice as much methane as previously estimated. Activities on well pads, including condensate storage tanks, pipelines and compressors, were leaking or venting an estimated 4 percent of the gas produced. Industry data and engineering calculations pegged that amount at 2 percent — a significant underestimation, Petron has said. The team also found emissions of benzene, a known carcinogen, from oil and gas operations at between 385 and 2,055 metric tons. Earlier estimates ranged from 60 to 145 metric tons per year.
At that point, there were nearly 14,000 operating oil and gas wells in Weld County. As of Aug. 7, there were 20,445.
More than 500 air samples taken from a Weld County test site sampled in January and February 2011 by CIRES researchers, led by Gilman, also showed high concentrations of volatile organic compounds, including propane and ethane. Gilman repeated the research in the summer of 2012 and again found an oil and gas signature in the air, though at lower concentrations — typical for warmer months.
The effects of those volatile organic compounds (VOCs) aren’t limited to Colorado residents living in the immediate vicinity of the wells.
“Propane and ethane are fairly long-lived in the atmosphere, so they travel far,” Gilman, lead author of that study, said in a January press release. “No matter where you are in the Front Range, you can still see the signature of VOC emissions from oil and natural gas operations.”
Generally not considered toxic themselves, those compounds are known precursors for ozone production.
A 2010 study by the Government Accountability Office also found that data collected by the Department of the Interior, which leases public lands for oil and natural gas development, was likely underestimating the amount of vented and flared natural gas because it failed to account for all sources of lost gas.
From that office’s view, this was lost revenue — royalties from oil and gas development generated $9 billion in 2009. The report estimates (based on EPA data) that 40 percent of the gas vented and flared at onshore rigs could be recovered, increasing royalty payments by $23 million and reducing greenhouse gas emissions by 16.5 million tons.
The EPA responded in April to the Office of the Inspector General Report on improving air emissions data with a list of corrective actions that calls for crafting a strategy for updating data by the end of 2013 and posting expected emissions for oil and gas activities that their database does not currently include by December 2014.