Calgary is the largest city in the western Canadian province of Alberta, about 230 miles north of the U.S. border with Montana. Fish Creek Wastewater Treatment Plant is the smallest of three wastewater treatment plants in the metro area of about 1.4 million people.

For 40 years, citizens in surrounding neighborhoods have complained of odors they believed to be coming from the plant. City managers weren’t convinced the plant was the actual source of the odor, which they believed to be caused by H2S. For years, they had tried to discern where the undeniable miasma was coming from, with frustrating results.

They of course wished to respond to their neighbors’ concerns, since these people were also their ratepayers. But it also became clear that — at just 9 mgd — the Fish Creek plant would soon need to expand to keep pace with its siblings; Bonnybrook at 91 mgd, and Pine Creek at 23 mgd. They realized there was no way popular sentiment would lean toward approving enlargement of a plant that already had an odor problem.

“There’s a significant amount of public engagement required to upgrade and expand a plant like this,” says Kevin Colbran, the city of Calgary’s wastewater treatment plant manager. 

The decades-old mystery would need to be solved, and soon.

Attempted mitigation

Fish Creek shares a catchment with Pine Creek. A structure at Fish Creek diverts most of the flow in that system to the Pine Creek plant. But Pine Creek doesn’t currently have capacity to treat all that wastewater. “There’s still that 35 mgd going into Fish Creek on any given day,” explains Colbran, “but there are large trunks upstream of Fish Creek as well. And as the smallest plant, Fish Creek also gets the bulk of our industrial wastewater.

“There’s suspicion that a lot of the industrial waste doesn’t get mixed, and gets directed into the Fish Creek plant. You’ve got everything from animal processing, food and beverage manufacturing, industrial machine shops, you name it,” on a gravity-feed system with a few lift stations.

Fish Creek uses a pure oxygen process. It does not remove ammonia, causing occasional noncompliant effluent into the Bow River. This is one of the main drivers of the upcoming plant upgrade, whose new process has yet to be identified. “Right now, it looks like aerobic granular sludge is preferred, but that’s debatable,” Colbran says. “We might end up with traditional BNR. We’re working through technology selection.”

Built in 1960, Fish Creek’s primary clarifiers were initially uncovered, and the headworks discharge was not scrubbed. Development encroached, and by the mid-90s, complaints about odor led to the covering of those clarifiers. There was only about one a month, but it was consistent, especially in warm weather when many more people were enjoying the outdoors.

When it comes to odor, however, it’s incredibly difficult to pinpoint a source without solid data. When an odor report comes in, someone is dispatched from the plant to investigate, but by then, the odor is typically gone. 

Colbran describes the frustration: “We can tour the plant to see if there’s any sort of odor issue, but most of the time without technology and data, the cause is inconclusive. There is always more than one source of odors in any particular area.” One preventive measure the plant takes is dosing the line from Fish Creek to Shepherd with hydrogen peroxide, to prevent odor generation in that line itself, but Colbran believes that’s probably unrelated to the issues at Fish Creek.

Technology to the rescue

At no point were plant personnel able to satisfactorily eliminate odor for long, which is unsurprising. It’s difficult to spend money on significant action without knowing the odor source and its proper mitigation. Still, due to its beautiful location just upwind of high-ticket homes on a nearby escarpment and in a provincial park frequently used for recreation, the Fish Creek plant got regular visual exposure. And with that came the not-unreasonable assumption that it was the source of the frequent unpleasant smell.

Fish Creek management issued a Request for Proposals in 2014 to get to the bottom of the problem. “I think it was for a pretty broad scope, just saying, ‘This is our issue. Who’s got the technology, and what can you come forward with, to help us pinpoint our issues?’’

Chris Madland, lead for the Prairies and Northern section of WSP’s (then Golder Associates) Air Quality team, saw the RFP and responded to the city’s team members Josh Lamarre and Rino Fedato. WSP had previously worked with Airdar Inc., deploying their unique air sampling and analysis technology to detect odor sources on clients’ oil and gas extraction sites. For this project, his idea was to set up an Airdar array to identify emission sources.

“They were asking for a real-time modeling solution, just to know how big of a problem their sources were causing,” says Dennis Prince, CEO of Airdar. “Chris responded to that RFP with, ‘We’ll watch those eight places and everywhere else; we’ll watch beyond the fence and identify any sources, then determine their impact in the community in real time.’ That got up to Kevin, and they all were open to this innovative approach, which I think is so critical.”

The equipment deployed at the end of 2014, went online in 2015, reported to the plant all year, then got renewed every year for seven years, giving them real-time alarms. It just shut down in January 2022.

Innovation is key

When the project began, it wasn’t yet possible to track sources using air concentration data. Efforts had been made by others using inverse dispersion modeling, but this approach suffers from biases that limit its capabilities. The plant’s odor issues continued unresolved, because there was uncertainty about the location of the sources. An approach had yet to be found that could ascertain actual sources. 

Madland believed the Airdar technology they’d used in the oil fields could be applied to this challenge. Airdar is a form of limited artificial intelligence that solves odor sourcing problems by mimicking animals’ location capability. Some animals track food sources using air concentrations in the wind. For example, a polar bear can track a seal off the ice 50 kilometers away, because its brain can identify patterns in air concentrations through its nose. The company calls this a form of air detection and ranging, or “Airdar.”

Airdar is similar to sonar and radar in that all three technologies use data patterns to locate things remotely. Sonar uses patterns in sound waves, radar relies on patterns in radio waves, and Airdar uses patterns in air concentrations traveling in the wind to locate sources of any compound including odor, smoke and particulate. It’s a powerful capability that can identify plumes in data from a single observation position. 

In this project, Airdar analyzed patterns in air concentrations from 32 observation positions deployed to the site, all tracking sources and sharing information with each other, 24/7. 

Airdar’s most important breakthrough is that it can delineate plume signatures from sources in ambient air concentration data measured at great distances from the sources. Its other innovation is the design of a system that can measure ambient air concentrations of H2S with a 1 ppb detection limit (amply accommodating human sense of smell) at multiple locations. Both innovations allowed Airdar to remove any uncertainty regarding the true odor emission sources.    

Overcoming obstacles

Obtaining these continuous measurements requires an expensive ambient air monitoring instrument that can require power, frequent calibrations and support gases. Deploying several of these instruments would be costly. An innovative way of measuring H2S concentrations at 32 different locations was needed to reduce equipment costs. 

This challenge was overcome by deploying a single instrument cabinet with access to power, in a tunnel at the site. It contained two high-end H2S instruments, with supporting calibration gas cylinders and diluters on board. This provided calibrations daily, eliminating the need for an on-site technician. 

The cabinet also housed valve manifolds and vacuum pumps, which drew samples through 32 small-diameter tubes, connected to remote sample inlets located throughout the plant and along the fence line. 

This system was able to pull a sample from up to 1.2 km away, while adjusting for sample travel time. The advantage was that 32 expensive instruments were not needed; only two were used, eliminating the need for a power source at each remote sampling location. 

H2S emissions from faraway sources can have important impacts on people because we can detect it at low concentrations, and because of the plumes’ ability to travel great distances before they disperse below these concentrations. Stubbornly complex odor problems may be due to distant sources rather than those suspected locally, and this was proven by the Airdar testing’s hard data results. It identified that plant personnel were doing everything right and that the source of the odor, while remaining unidentified, was not the Fish Creek plant.

One significant suspect is the local collections system itself, but there is insufficient data and investigations to conclude anything definitively. The plant management is grateful to have been absolved of any responsibility for the odor issue, which seems to have resolved itself after some routine changes to the collections system. If it arises again, attention will immediately be focused on the collections system and outlying infrastructure that feeds into the area.