GREENSBORO — The city water department is embarking on a long, costly effort to counteract pollution from firefighting foams that apparently have accumulated over the years in part of the city’s watershed.
The first act played out last week, when officials at the Mitchell Water Plant christened a new treatment device that the city is renting for $9,000 a month and that costs about $1,000 per day to run.
They plan to buy one of these “PAC Feed” systems to replace the rental unit at a cost of $80,000 to $120,000 within six months.
The game plan calls for only running the machinery when weekly tests show that the concentration of firefighting chemical PFOS and a sister compound reach the federal health advisory level of 70 parts per trillion (ppt) in the plant’s Lake Brandt water supply.
But testing for PFOS, the related chemical PFOA and other perfluorinated compounds costs another $1,500 per week.
And with test results coming in at that pace, it means that once cranked up, the PAC Feed system likely will run for days at a time using roughly $1,000 per day in filtering materials before a new round of chemical testing gives plant operators the all-clear to shut it down. Or not.
“This figure is very close to four times the current Mitchell daily cost of all other chemicals combined,” city Water Supply Manager Dell Harney said of PAC Feed’s operating cost in a recent report.
The newly leased PAC Feed system’s goal will be to keep the combined total of PFOS and PFOA at least 10 percent below the nationwide health advisory, meaning that the maximum concentration would be 63 ppt, Harney said.
Looking ahead, the off-and-on PAC technology could define the Mitchell plant’s operating routine — and its financial drain on the city — for the next four or five years.
That’s the rough time frame for planning, designing and installing a permanent replacement system that city water administrators see as the ultimate PFOS solution — at an estimated cost up to $30 million.
They plan to hire an engineering firm to design that permanent replacement system in the next budget year, said Mike Borchers, assistant director of the Water Resources Department.
“Construction could start as early as two years out,” Borchers said of the permanent replacement. “Due to the complexity of the project, it will likely take two years, so that places us in fiscal year 2022-23 for having the full-scale system in place and operational.”
The replacement system would use a different, carbon-based technology to screen out PFOS and other pollutants. That replacement technology is equally effective, but less costly to operate day-to-day than what the city unveiled Wednesday at the Mitchell plant.
The leased PAC equipment that the city debuted last week relies on powdered activated carbon to chemically attract PFOS and many other pollutants that form tight bonds. PAC is the acronym commonly used to describe the technology.
The leased machinery was added to a concrete slab beside the Mitchell plant in recent weeks, and city officials gained speedy approval from state regulators to include it as part of their treatment process.
The system works by crushing small bits of carbon into a powder, which is fed into a pressurized stream of water to form a thick fluid that is injected into untreated water during the early stages of purification.
The carbon in that injected fluid binds with PFOS and many other types of contaminants in the “raw” water, forming a sludge that is removed for incineration at the city’s Osborne Wastewater Treatment Plant.
The city’s longer-range plan would use a similar technology known as “granular activated carbon,” or more commonly by its GAC acronym. This technology adds a step to the treatment process by running future drinking water across filters of charcoal-colored carbon that do not have to be replenished as frequently as the bulk bags of PAC Feed carbon.
Currently, thanks to all the rain brought last month by Tropical Storm Florence, PFOS and PFOA are not a problem at the Mitchell plant because the compounds have been so heavily diluted. In test results received Tuesday, PFOS came in at 11 ppt and PFOA at 3.1 ppt in Lake Brandt, which supplies the Mitchell plant.
The neighboring Lake Townsend reservoir is larger, so the city’s other treatment plant on that body of water does not have the same PFOS-PFOA problem because the compounds are more heavily diluted and dispersed by the time they reach that plant.
The two chemicals are classified as “unregulated contaminants,” meaning they have not been formally branded as hazardous and subject to strictly enforced limitations. But they are on the U.S. Environmental Protection Agency’s short list for future action in that regard.
The two chemical compounds have been used for decades in a variety of industrial and commercial products ranging from stain-resistant carpet to the linings of microwave popcorn bags, nonstick cooking surfaces and other heat-resistant coatings. PFOS gained the most attention for its use as an additive to firefighting foam that helped to squelch fuel-based fires, so it has been heavily used at both civilian airports, military air bases and in industrial buildings that have internal fire suppression systems.
Domestic production of the two perfluorinated compounds has ceased because of health concerns, but they still can be found in durable products made years ago and in goods imported from countries that still permit their use.
Greensboro’s difficulties trace back to 2014, when PFOS was detected at levels significantly above the current EPA advisory level as part of nationwide testing ordered by the federal agency. That level is not a hard and fast safety limit, but a calculation of health risks based on the lifetime effects of consuming water every day with that concentration of a certain chemical.
PFOS has been the city’s major problem all along. Sister chemical PFOA has been present at concentrations that would not raise eyebrows but for the larger footprint of its close relative.
Borchers and other Water Resources staff spent months tracing most of the problem back to the Horse Pen Creek branch of the city’s watershed, which is one of Lake Brandt’s major sources.
They say that they have tested extensively at Piedmont Triad International Airport and the industrial area around it, where the problem seems to be centered, leading them to believe so far that the genesis is not any commercial or manufacturing enterprise.
Rather, they attribute it mainly to decades of firefighting and fire training exercises using PFOS-containing foams at PTI, nearby businesses and industries in the area.
“There are multiple sites, in and around the airport, that are problematic based on historical activities over the past 50-plus years,” Borchers said last week.
Some tests in the Horse Pen Creek tributary have found the combined levels of PFOS and PFOA at concentrations greater than 300 ppt at the point where the stream enters the lake, said Harney, the city’s water supply manager.
Those levels decline by the time that creek water flows farther into Lake Brandt and reaches the Mitchell plant’s downstream water intake. But they have at times been over EPA’s 70-ppt health advisory level, most recently in late July when PFOS registered 80 ppt and PFOA 10 ppt, for a combined reading of 90 ppt in the lake.
A week later, samples of water treated at the Mitchell plant and about to enter the distribution system registered even higher, a combined total of 100 ppt. Then, within seven days, further tests showed levels of the pollutant pair had dropped back by two-thirds without any known human intervention, to a combined total of 33.3 ppt.
Water managers have at least one hope on the horizon for getting a better handle on the erratic situation. They want to develop a reliable model for how PFOS and its cousins move through the watershed.
That would help them predict how long it takes a chemical spike detected in Horse Pen Creek to make its way to the Mitchell plant and, on the back end, how long that spike might last before the contaminants shrink to “normal” levels.
In the short term, that would help them anticipate when they need to turn on the Mitchell plant’s new PAC Feed system and for what length of time. That knowledge could save the city from running the PAC Feed equipment more than needed — conserving roughly $1,000 with every unnecessary day they can safely avoid.
But their “predictive model,” as Borchers terms it, will be charting new territory. They recently conferred with a high-ranking EPA official who told them that “Unfortunately, he was not aware of any modeling efforts that have been developed” anywhere for predicting the travel patterns of such perfluorinated compounds, Borchers said.
The city is working with the private engineering and consulting firm HDR to create such a model, which involves detailed analysis of stream flow, drainage areas, water quality and a variety of other factors in a study process that could still take “many months,” he said.
Water treatment professionals would prefer to deliver a clean, safe product to customers’ faucets without facing such challenges, said Steve Drew, interim assistant city manager and director of the Water Resources Department.
But there is also satisfaction in tackling a problem that many other communities across the nation are only beginning to recognize.
“From a professional standpoint, it’s very interesting because we are out here on the cutting edge,” Drew said. “The things we are learning here may help other utilities elsewhere across the country.”