When a malfunction at Suncor Energy’s oil refinery spewed sulfur dioxide into the air over Commerce City in the predawn hours of April 12, a machine quietly sampling air quality in the community detected the pollution.

But other instruments tracking the air around Suncor did not.

Air monitors operated by the Colorado Department of Public Health and Environment, Suncor itself and Cultivando, a nonprofit that serves as a watchdog over the refinery, all recorded different data on what had been blown into the air as people were just beginning their day. In fact, some instruments did not record the spike in sulfur dioxide at all.

And when the state health department made the sulfur dioxide incident public later that evening, its report included a data set that did not reflect the peak emission point, but instead provided a less-alarming five-minute average. None of the data collected found Suncor had violated federal air quality standards. But the company has filed a report with the state that indicates it violated the conditions of its air permit, and that remains under investigation.

The situation surrounding Suncor’s April 12 sulfur dioxide discharge illustrates just how complex and confusing air quality monitoring can be, especially for the average resident who is simply trying to figure out exactly what pollutants are in the air, how long they are being exposed to them, and what the health impacts might be.

Jeremy Nichols, climate and energy program director for advocacy group WildEarth Guardians, said releasing complicated data muddies the public’s understanding of what’s in the air.

“Polluters like to leave it complicated,” he said.

Detlev Helmig, the scientist who runs Boulder Air, the company that monitors Commerce City air quality on behalf of Cultivando, said his instruments recorded sulfur dioxide peaks of 491 parts per billion and 233 parts per billion between 5:17 a.m. and 8:50 a.m. on April 12.

In the health department’s news release, the agency reported “short-term spikes” of 155 parts per billion and 186 parts per billion, but those so-called spikes were five-minute averages, Helmig said. The state said the emissions did not violate federal standards for sulfur dioxide, which are 75 parts per billion on a one-hour average.

Still, the short-term spikes in sulfur dioxide levels were high enough — even without a sustained one-hour average that exceeded federal standards — that the health department was compelled to issue a public notice, albeit about 12 hours after the incident occurred.

As for Suncor’s air permit, the amounts of sulfur dioxide it is allowed to release are measured two ways — in parts per million by volume on a 12-hour rolling average and pounds per hour.

In a report filed on April 13 by the company, Suncor said its sulfur dioxide emissions likely reached 4,527 parts per million by volume, while its permits say it should not exceed 250. And it also reported those emissions likely reached 545 pounds per hour, while the permit allows 15.68 pounds per hour, according to the documents.

That day, word was spreading in the neighborhoods around Suncor that some kind of incident had happened, but the company had not issued a public notification about it. (Suncor did not respond to questions from The Denver Post about why no emergency notification was issued.)

A representative from the Adams 14 school district called Cultivando to ask what had happened and how it might affect students, who were taking state standardized tests that week, said Guadalupe Solis, Cultivando’s director of environmental justice.

“Without any of that guidance, it was hard for them to determine whether to pause their tests or postpone,” Solis said.

The lack of consistency is a big enough problem that the head of the state’s Air Pollution Control Division said he is trying to figure out a way to “harmonize monitoring and messaging” so public and private air monitoring data can be synchronized in an understandable, consistent way.

“It’s something that has been a specific interest of mine for years now,” said Michael Ogletree, the Air Pollution Control Division director.

Ogletree said work is being done to figure out how that would come together, but it’s not an easy project.

Different monitors, different readings

On the morning in question at Suncor, Boulder Air’s equipment recorded the first spike in sulfur dioxide at 5:17 a.m. The computer analyzing air samples determined sulfur dioxide levels peaked at 491 parts per billion at that moment in time, Helmig said.

“It was huge,” he said.

At 8:50 a.m., the reading spiked again at 233 parts per billion, Helmig said.

Sulfur dioxide is a byproduct of refining fossil fuels, and even short-term exposure can cause respiratory illnesses, including asthma attacks. High concentrations also can damage trees and other plants, the Environmental Protection Agency says.

The spikes on April 12 were so high that Helmig and the Boulder Air staff thought that something must have gone wrong with their instruments.

“We’ve never had anything close to those values,” Helmig said.

Boulder Air’s air-monitoring equipment sits inside a trailer parked on private property about 1.3 miles — as the crow flies — northeast of the refinery. Inside, hundreds of thousands of dollars worth of high-tech instruments are pulling air samples from outside and analyzing them.

An employee went to the air monitoring station around 9:30 a.m. that day and unplugged the computer so he could recalibrate it. Nothing was wrong. The spikes were that high. Once Boulder Air sent its findings to Cultivando, the nonprofit notified the Colorado Department of Public Health and Environment’s Air Pollution Control Division, Helmig said.

The state also has an air monitoring station within the vicinity of Suncor — known as the Colorado Air Monitoring Mobile Laboratory. It was parked 600 yards south of Suncor that morning. It did not detect abnormal sulfur dioxide levels at all.

Meanwhile, Suncor’s community air monitoring program, which is run by Montrose Environmental Group, had 10 air monitoring stations collecting air samples at that time. But only one detected a high level of sulfur dioxide, Loa Esquilin Garcia, a Suncor spokeswoman, wrote in an email.

The company reported that its reading was recorded for a single five-minute period. But Suncor did not provide specific emissions data.

So what happened?

Wind and other variables

To start with, the wind played into the detection.

That morning, winds were steady and blowing in a south/southwest direction, Helmig said. So Cultivando’s path was downwind of the refinery and the emissions were blown over its instruments that are pulling in air samples.

The state’s air monitoring van was on the south side of the refinery so the winds blew the sulfur dioxide away from it.

And Suncor’s single air monitoring station that detected the sulfur dioxide was on the northeast side, Garcia said.

But other factors can alter how much and how long a pollutant flows into the air.

The Boulder Air monitors, for example, alternate which pollutants they are surveying every 10 minutes. So the instrument that was picking up high levels of sulfur dioxide the morning of April 12 switched to a search for hydrogen sulfide shortly after recording its first spike, Helmig said.

Along with the 10 monitors stationed in the community, Suncor also has five monitors around its perimeter as part of a state-mandated fenceline monitoring program. The company did not report whether any of those monitors detected a high level of sulfur dioxide.

The Air Pollution Control Division in August ordered the refinery to continuously monitor its emissions in all directions but Suncor is suing the state over the requirements.

Meanwhile, other variables make air monitoring an inexact science.

For example, Suncor has permits to release emissions from 34 smokestacks on its property, and those stacks range in height from 30 feet to 200 feet. The height of a stack would influence how a pollutant would be dispersed, and plumes are not uniform as they move into the atmosphere, Helmig said.

“We don’t know which stack it came from,” he said the day after the April 12 sulfur dioxide release.

Suncor later filed a report with the state health department specifying which stack released the sulfur dioxide, but that report does not say how high the stack is.

On top of the nuances of wind, stack height and monitoring mechanics, there is the complexity of how emissions are reported, what is considered a violation of federal air quality standards, what violates Suncor’s air permits and how much of a pollutant is dangerous to human health.

The state’s Air Pollution Control Division and Suncor said in statements that the April 12 incident did not violate federal National Ambient Air Quality Standards, which define the maximum amount of pollutants that can hang in the air over a period of time without impacting people’s health.

Still, state health officials found the spike high enough to send a news release.

“As we think about how we get information out, we look at the best available science and data about health impacts, and the best we have right now is looking at one-hour averages,” Ogletree said. “During that time period, we did see some that were approaching some of the ambient air quality standards and we wanted to be proactive. But no time during that period did we go above a National Ambient Air Quality Standard.”

The state air quality regulators said the short exposure could have exacerbated asthma and made breathing difficult, especially during exercise or physical activity. They offered tips about staying limiting outdoor activities and closing windows and doors.

Nichols of WildEarth Guardians said he “was floored” when the health department sent out a public notification about the sulfur dioxide emissions even though federal standards were not violated.

“That email never would have happened five or six years ago,” Nichols said. “That’s a big shift in culture in the agency.”

Michael D’Agostino, a spokesman for the Air Pollution Control Division, said one reason the agency issued its news release is that Suncor didn’t send a community alert on its own.

The state’s investigation into whether Suncor violated conditions allowed by its Title V air permit remains ongoing.

Which numbers matter

Cultivando and the state health department have had this disagreement over which numbers matter before.

In March, Cultivando held a news conference to release findings from its first year of the air monitoring project.

The presentation included a report of high levels of benzene and particulate matter, but in response, the state health department said those were snapshots and did not present a complete picture of air pollution.

And Suncor has said that its air monitoring has never found any pollutant levels that exceed acute and chronic health-protective guidelines used by federal and state agencies.

Federal violations or not, those who live in the neighborhood fear even short-term spikes, especially when they learn about all the other pollutants mixing in the air at the same time, Cultivando’s Solis said.

“It is safe to say any type of exposure, whether it’s short term or long term, it is bad for the community,” she said. “Ultimately, all exposures and all adulterations of our air are bad.”

At the end of the day, the big question is what is being done with all of the data gathered, Nichols said. The state needs to use it to keep Suncor in check and prevent future releases of sulfur dioxide and other substances that make people sick, he said.

“People can nickel and dime the emissions readings,” he said. “We know it’s there, so what do you do with it?”

Updated 5:50 p.m. May 2, 2023: This story has been updated to correct information about Suncor’s fenceline monitoring program. Each of the five monitors stationed along its perimeter continuously monitor air flowing in one direction.

Get more Colorado news by signing up for our Mile High Roundup email newsletter.