This story is part of our Clearing the Air series about shale gas drilling in Penn Township. We monitored the air quality at five homes near the township’s first well pad in late 2015. This is what we found.
James Moore had grown so accustomed to the sound of shale drilling, that when it stopped, it weirded him out.
“When it was quiet, the crickets were not loud enough” to make up for the absence of noise from drilling, he said. “I actually couldn’t sleep without the noise.”
Moore lives in a bright green and yellow house — dubbed the “John Deere” house by neighbors — about 500 feet from Penn Township’s first shale gas wells. His house (which he rents) is closest to the gas well pad.
The pad is relatively quiet now. For about a year, two wells there have been feeding into a pipeline that jets natural gas to Pittsburgh. But in 2015, Apex Energy was drilling and fracking those wells. Trucks, rigs and incinerators stampeded the site.
Adding to the noise issue, Moore suspected the gas site was emitting pollution that he couldn’t see, hear or feel.
Because he’s the closest person living to the pad, Moore was concerned about the exposure, so he decided to participate in PublicSource’s air quality monitoring project. Four other residents living fewer than 3,000 feet from Quest Realty Central Pad 7 joined him.
In August 2015, researchers at the Southwest Pennsylvania Environmental Project* [EHP] and I set up the project in the township, a rural suburb about 20 miles east of Pittsburgh. Over about two months, we took air quality measurements at the participants’ homes.
One of our goals was to track short-term peaks of air pollution. We looked for fine particulate matter — called PM 2.5, which can release from various sources at gas sites, including diesel trucks and other machinery.
Natural gas operations can have episodic spikes in emissions that last for a few minutes or hours. Federal and state standards for PM 2.5 are based on 24-hour averages and can miss these short-term spikes or “peaks” because the data gets averaged out, said David Brown, EHP’s lead toxicologist. Studies have shown that short-term exposure to increases of PM 2.5 can cause health issues.
EHP researchers found that peaks of PM 2.5 at all five houses were high enough to potentially cause upper respiratory problems in sensitive populations including children, the elderly and people with asthma.
However, the baseline levels of PM 2.5 we recorded at each house stayed below the daily 24-hour threshold set by the U.S. Environmental Protection Agency.
We also looked for an array of volatile organic compounds (yes, they’re as toxic as they sound), and our monitoring showed that their levels did not exceed health standards either.
To measure PM 2.5, we used low-cost air monitors called Speck sensors that take minute-by-minute measurements.
The analysis showed that the Specks we placed outside the homes recorded about two to four peaks of PM 2.5 per day at all five houses over a 32-day period during our monitoring. 
Peaks were substantially higher closer to the well pad, at Moore’s house. The Speck on his porch recorded peaks more often and at higher levels than Specks farther away.
“This fits with the expectation that the source of PM 2.5 is closer to this residence than it is to the remaining four homes,” EHP researchers concluded.
There were times when all the outdoor Specks showed increased PM 2.5 at the same time, notably just before fracking reportedly began and also during the combustion phase when incinerators were used to burn excess natural gas.
Based on the results, EHP researchers recommended the use of indoor air filters in homes near this site and other similar industrial sites.
The group has used Speck sensors at more than 200 homes in Pennsylvania, New York, Ohio and West Virginia.
The Penn Township location “is not the worst site we’ve seen” but “is one of the most accurate characterizations of exposures to a gas well that we’ve gotten so far,” Brown said.
Why look at short-term peaks?
A shale gas well pad is not static.
Different equipment is used for varying lengths of time to drill and frack wells. Truck traffic ebbs and flows. On Apex’s Quest pad, two large furnace-like devices were eventually plopped on top of well heads to burn off gas. Those ran on and off for about a week.
Each phase of a shale gas operation can release air pollution at various amounts, for different lengths of time. A few minutes. An hour. A half day. These short-term peaks of air pollution can trigger health symptoms within a few minutes of exposure, said EHP’s Brown, who previously worked for the Agency for Toxic Substances and Disease Registry at the Centers for Disease Control and Prevention.
“That is important because some of the health effects we’re seeing around sites and people are reporting have to do with respiratory diseases, such as asthma in children and chronic obstructive pulmonary disease [COPD] attacks,” he said.
In 2014, Brown wrote a paper criticizing common state and federal methods that he says underestimate public health threats because they fail to account for short-term peak emissions at shale gas operations. Regulators monitor the air in Pennsylvania, but generally use methods for determining regional air quality over longer periods of time, such as 24 hours.
What the driller and DEP had to say
I shared our findings with Mark Rothenberg, chief executive officer of Apex Energy, the company that operates Quest Realty Central Pad 7.
“Most of our questions revolve around doubt that the findings are related to our activities as opposed to the very busy roads and industrial work that goes on in and around our location,” Rothenberg said in an email.
We cannot definitively say that what showed up on the monitors was caused by the shale gas operation because there are other possible sources of local and regional air pollution near these houses, and we didn’t take baseline air measurements before the start of the gas operation.
Apex officials said to reduce emissions, they took many steps above what’s required by state and federal standards (you can see a list of some of these in this email).
However, Brown said he’s concluded that the monitors show pollution spikes originating from the shale gas pad, largely because peaks were higher and more frequent in the house closest to the site.
Apex officials also questioned the accuracy of the Speck monitors, and if high humidity and unfavorable weather patterns could have caused the monitors to register inaccurate figures. They pointed to Environmental Protection Agency research that found accuracy issues with the Speck when it was used outdoors compared to a regulatory-grade monitor.
Researchers at New York’s Clarkson University recently tested the Speck and other low-cost air monitors indoors. They found that the Speck performed with “adequate precision” compared to more expensive monitors, but stated that the monitors could lose precision when used outdoors. Carnegie Mellon University’s CREATE Lab, where the Speck was invented, also found the Speck performs well against a federal-grade monitor.
Researchers familiar with the device told me that Specks are considered to be “indicative” monitors, meaning they’re good for spotting trends and patterns, the ups and downs of particulate matter, even if they may not show exact concentrations.
Still, indicative monitors are often used in exploratory surveys (like ours) to get an idea if there is a pollution problem and if more robust air monitoring is needed.
Staff at the Pennsylvania Department of Environmental Protection’s Bureau of Air Quality had the same concerns as Apex about the accuracy of the Specks and other potential pollution sources near the houses.
“The findings definitely show a correlation — but not necessarily a causation,” Neil Shader, an agency spokesman, wrote in an email.
On Dec. 5, Penn Township officials agreed to settle litigation with Apex after denying the company drilling permits for other proposed pads based on environmental concerns. The settlement allows Apex to construct the pads if it meets certain conditions, including that no trucks will idle on the sites and that air monitoring will be conducted.
What else did we find?
In addition to measuring PM 2.5, we used summa canisters to measure volatile organic compounds (VOCs) — a group of hundreds of chemicals that include known carcinogens such as benzene.
We took nine summa canister samples that detected a range of VOCs, some of which have been found at other shale sites. No chemicals exceeded any regulatory health standards.
But Kim Schultz, a research scientist at the Endocrine Disruption Exchange, said there’s very little research on what exposure to low levels of multiple VOCs at the same time can do to the human body.
“Just because you found [VOCs] beneath health thresholds doesn’t mean they are necessarily safe,” said Schultz, whose organization studies health issues caused by exposure to low doses of chemicals.
Also, our results could potentially have been watered down because we took only nine samples and the results were 24-hour averages. Certain chemicals could have been present at higher concentrations over shorter periods of time or on different days.
More VOCs were detected at the Penn Township pad when the wind was calm or light. Therefore, researchers concluded that during times of low wind speed, potential exposure to air pollution could increase because the wind doesn’t disperse pollution as much.
VOCs can mix with nitrogen oxides and create ground-level ozone, or smog, which can trigger asthma attacks in children. VOCs also have the potential to make exposure to PM 2.5 worse because they can latch on to dust particles and be carried into the deep lung, Brown said.
What other researchers had to say
EHP’s analysis was structured in a way that could give residents experiencing health issues and their physicians an understanding of potential pollution exposures, Brown said.
None of the residents participating in this project reported health problems to me. However, in October 2015, when the drillers were using incinerators to combust gas, several people living within 1 to 2 miles of the gas site reported sore throats and coughing fits to Protect PT, a local citizens group against fracking near homes.
I showed the analysis to public health and air quality experts, many of whom have monitored air quality around shale operations.
“I thought the way you approached looking at the maximum exposures of people is of interest to a lot of the scientific community and very forward thinking,” said Michael McCawley, an environmental scientist at West Virginia University.
He said researchers and regulators shy away from looking at short-term exposures because they are hesitant to use Specks and other monitors that might have accuracy issues. So there’s little data showing how much pollution people living near gas sites are exposed to.
“[Specks] are telling you something about the relative trends … so doing this is a very good first step,” McCawley said. “Some data is better than no data.”
Wilma Subra, a Louisiana-based environmental scientist, said she thinks our Speck data shows pollution coming from the well pad, in part because peaks correspond with the different phases.
“You’ve got the hits when they were drilling,” Subra, a chemist, said, “you’ve got the hits at [Moore’s] house when they were preparing for fracking, when all of the vehicles were coming in preparing it, hauling in sand… it went up with fracking and then once they pulled all of the equipment off, it went down.”
While Apex and DEP had concerns with Specks picking up pollution from other sources, Illah Nourbakhsh, who helped invent the Speck, said the sensors don’t register the tiniest particulates, known as ultrafines. So it’s possible the Specks didn’t detect all the particle pollution near the houses.
My thoughts
Here’s what I learned about doing a citizen science air monitoring project: It’s tough. But it can be a powerful tool.
It’s tough because as much as you plan, things can always go wrong. People are not computers or robots.
For instance, one day I noticed the weather was overcast and calm. That’s an ideal time to open a summa canister because pollutants hang out in the air. I tried to contact one of the residents to open up a canister, but couldn’t. He’d run out of cell phone minutes.
It’s tough because air pollution is fickle and hard to track.
An emission can occur, a person can have a health problem, but then the evidence can quickly disappear. Dust and chemicals move with the wind and end up far from where they originated. Some chemicals sink into valleys or combine with other chemicals to make new ones.
It’s tough because, if you use low-cost air monitoring devices or have citizens involved in the data-gathering process, industry officials and regulators will be skeptical of your results.
“You have this huge inequity issue,” said Nourbakhsh, a robotics professor and CREATE Lab director, “where the monitors that most accurately measure outdoor air … are so expensive that only organizations that have power own them.”
Nourbakhsh said that while one set of Speck data (or one project using several Specks, like ours) may not be a “slam-dunk case” by itself, it can be “a stone on a mountain of evidence… that altogether, crushes doubt, because it’s just too much to deny.”
For the shale gas industry, that mountain of evidence seems to be building.
A 2016 analysis of seven years of peer-reviewed studies that assess the impacts from shale gas drilling found that 87 percent of 46 air quality studies showed elevated air pollutant emissions and/or atmospheric concentrations. The analysis was done by PSE Healthy Energy, a research institute that studies the energy sector.
Citizen science projects have also contributed. In a 2013 community-led study, trained volunteers took air samples near oil and gas sites in five states and found eight volatile chemicals that exceeded health standards. It was one of the studies included in the New York health commissioner’s review of public health studies that led to the state’s fracking ban.
Still, major data gaps remain. Only 31 of the 685 studies the PSE analysis examined dealt specifically with how fracking affects people’s health. Most of those studies indicated a need for more research, especially large-scale epidemiological studies, which will likely take years to do.
In the meantime, if low-cost monitoring technology keeps improving, could community-led projects make citizens more aware of risks and help them protect themselves near gas sites?
I’ll explore that question next week. I’ll focus on how citizen science could evolve, the ongoing efforts to expand its uses and how federal regulators and academics are trying to validate and improve low-cost sensor technology.
*PublicSource is funded by some of the same foundations as the Southwest Pennsylvania Environmental Health Project. We are not for or against shale gas drilling and we did this project purely as a fact-finding mission.
This project was funded by Arizona State University and the Knight Foundation. The Heinz Endowments also helped support this project.
Comics by Em Demarco for PublicSource.
Reach Natasha Khan at 412-515-0063 or natasha@publicsource.org. Follow her on Twitter @khantasha.
About the data
In early 2015, I received a grant from Arizona State University and the Knight Foundation to bring something “digitally innovative” to the PublicSource newsroom. In recent years, journalists have been using newly developed low-cost sensor technology to generate data about the communities they cover. I asked the grantmakers for funds to set up a network of sensors around a new shale gas site.
I worked with the Southwest Pennsylvania Environmental Health Project (EHP) to help residents test for PM 2.5 and volatile organic compounds (VOCs) near the first shale gas well pad in Penn Township.
The goal of the project was to see if short-term peaks of PM 2.5 had the potential to affect the health of nearby residents and to identify what VOCs were present in the air. Claims of negative health impacts from shale gas operations are often dismissed as anecdotes and speculations. Often there is little evidence to back up these claims because regulators generally don’t do localized monitoring near gas sites.
In March 2016, EHP sent each resident individualized health assessments based on the data we gathered from each of their homes. Public health researchers at EHP also wrote an overall health assessment based on the data from all the monitors and all the VOC samples we collected. Here’s a copy of EHP’s full statistical analysis, which includes a detailed breakdown of how the group calculated a “peak” measurement. For questions about the analysis, please contact EHP toxicologist David Brown at dbrown@environmentalhealthproject.org.
Testing for PM 2.5
To test for PM 2.5, we used low-cost Speck sensors at each of the five homes whose residents participated in our project. In mid-August 2015, Ryan Grode, an EHP environmental health educator, helped me to conduct a training session of these for Penn Township residents.
Each resident received two Speck sensors to place inside and outside their homes. All but one of the outdoor sensors were kept under covered porches to protect from weather. One resident without a porch used a metal box designed by CMU’s CREATELab researchers (the group that invented the Speck) to protect it from outdoor elements.
Recently, Specks and other low-cost air sensors have exploded in popularity with people who recognize a lack of government monitoring occurring and want to know more about their personal air quality.
Here are some technical specifications about the Speck.
Testing for VOCs
To test for VOCs, we took nine summa canister samples at two houses. James Moore, who lives about 500 feet from the well pad, took five samples, and another resident about 2,000 feet away took four samples. We used summa canisters because they are often used by regulatory agencies and are accepted as a reliable testing method.
All canisters were opened over 24-hour periods by these two residents and then handed off to me. I packaged and sent all nine canisters to ALS Global laboratory in Simi Valley, Calif., where the captured air was tested for VOCs using the U.S. Environmental Protection Agency-certified TO-15 method. Sometimes residents didn’t sign chain-of-custody documents in the correct places, but I can attest that these canisters were opened at the residents’ houses, given to me and shipped correctly.
A caveat
We cannot definitively say that our findings were caused by the shale gas operation, largely because of other possible sources of air pollution near our monitoring location. Other sources of potential air pollution near the homes include local neighborhood traffic, a state road a quarter to a half mile away, a toll road roughly a mile and a half away, nearby gas compressor stations, a nearby industrial park and regional air pollution from Pittsburgh and Murrysville.
Monitoring in the future
After monitoring for our project ended, PublicSource donated 10 Speck sensors to Penn Township citizens living near areas pegged for new well pad construction. EHP staff are helping citizens collect data and will provide them with health assessments as shale gas drilling develops in this community.
Apex Energy has plans for new well pads in Penn Township and will be required to conduct air monitoring at the sites during construction, drilling and completion phases. Penn Township officials said the specifics of that air monitoring (including what pollutants will be tested and if the data will be publicly available) are still being worked out with Apex.
