New DEQ Air Toxic Rules Would Allow Disturbing Emissions of Carcinogenic Chromium

By Daniel Forbes

Change continues apace at the Oregon Department of Environmental Quality.

On the heels of receiving a clarifying letter from the U.S. Environmental Protection Agency that Bullseye Glass and Uroboros Glass actually do fall under existing federal regulation, [See Portland Mercury, here] the DEQ has issued its own, significantly revised temporary rules governing emissions from the two Portland glassmakers at the heart of a months-long air toxics scare

DEQ is waiting for information from the glassmakers on how many furnaces they have whose annual capacity exceeds 50 tons before it considers implementing the EPA standards. In the meantime, it’s issued a new version of temporary rules governing Uroboros and Bullseye, rules that DEQ’s overseers, the Environmental Quality Commission, will accept or reject at a public meeting in the Convention Center Friday April 21.

Those rules bring with them a host of concerns—particularly what they have to say about one carcinogen: hexavalent chromium (Cr VI).

Under the rules no facility would be allowed to emit an annual average of more than the current air benchmark of 0.08 nanograms per cubic meter. (In 2005, an EPA contractor found median annual Cr VI readings of 0.03 ng/m3 in 25 cities nationwide. Detroit was highest, at 0.066—still less than DEQ’s annual limit for the glass companies.)

What’s more, from now until September 1st, should EQC approve the rules, the glass factories would be allowed a “daily acceptable source impact level” of Cr VI emissions from uncontrolled furnaces of 36 ng/m3 of Cr VI. That’s 450-times Oregon’s 0.08 ng/m3 stated annual benchmark, yet DEQ insists it’s safe. The agency arrived at the daily, 36 ng/m3 concentration in consultation with Oregon Health Authority state toxicologist, David Farrer.

“Whatever is in the rules is protective. That’s the whole idea of putting the temporary rules in place,” said DEQ spokesperson Marcia Danab, who noted that Farrer gave his “approval.” This though a recent, single day’s hex-chrome ambient concentration of 2.23 ng/m3 that DEQ’s monitors found in southeast Portland elicited public comment and concern.

Additionally, the 36 ng/m3 “source impact” concentration pertains to “any off-site modeled receptor.” Given natural dispersion of emissions from a facility’s stack to its fence-line and then to an off-site receptor — home or school or day-care center — apparently the amount at the stack might be greater than 36 ng/m3. While DEQ likely will continue at least for a time with its current air monitoring program centering on Bullseye and Uroboros, the rules themselves require a stack test only; dispersion modeling will estimate the impact on any receptor.

The proposed 36 ng/m3 daily limit is a disturbing amount of hex chrome to release into the air over the course of a single day. In Bullseye’s case, it would be released within 100 meters or so of a daycare facility. Studies of Cr VI cancer risk consider adults working with chromium in some form—not young children. “I personally know of no studies extrapolating human dose/response and any related risk estimates specific to early life exposure (in infants) to life-time cancer risk,” says Sharan Campleman of the American College of Toxicology.

That view was echoed by a medical school senior toxicologist who asked not to be named but said, “I am not aware of any dose-response data specific to infants. After all, we don’t give them this stuff.”

So. There’s little to no good research on Cr VI’s threat to toddlers. But there’s also scant research extant on the risks of exposure over the course of a single day.

The federal Agency for Toxic Substances and Disease Registry, which sets Minimum Risk Levels for the nation as whole, threw up its hands at determining Cr VI’s risk for “acute-duration” exposure of anywhere from one day to two weeks. “[T]he data are insufficient,” it says.

ATSDR did derive an MRL of 5 ng/m3 for exposures of more than two weeks—on up to years. (Hex-chrome is such nasty stuff that, in a departure from ATSDR’s normal protocol, Cr VI’s “intermediate” MRL for exposures over two weeks but up to a year, is the same value—not less as is normally the case—as the MRL for “chronic” exposure lasting a year or more.)

Authored by eight credentialed professionals, the peer reviewed ATSDR toxicological profile for chromium is 592 dense pages long, including 70 pages of references, and the product of years of work. When it says there’s no data to arrive at a one-day MRL threshold, it’s fairly definitive.

So how did DEQ derive the daily Cr VI limit of 36 ng/m3?

A storage shed at the Bullseye Glass site. Photo by Wm.™ Steven Humphrey, Portland Mercury.

Both Farrer and DEQ toxicologist Sue MacMillan felt the need for a 24-hour reference-figure to contextualize the air-monitoring data their agencies now release. DEQ has its Ambient Benchmark Concentrations (the aforementioned “safe-air goals”) for chronic exposure to chemicals; it’s meant to measure annual data. But MacMillan said DEQ and OHA needed a daily value to offer perspective on the more elevated readings DEQ/OHA was releasing to a worried public.

She found what she was looking for in a New Hampshire state regulation that specified a 36 ng/m3 daily limit. She declared the New Hampshire threshold “a safe level … it’s what’s fair.” But when asked how the Northeastern state arrived at those levels, she only said: “As to the toxicity studies, they must have looked.”

MacMillan and Farrer also looked at Texas and Ontario’s reference levels, but they were higher than N.H., she asserted. “New Hampshire was the lowest of the three.” That’s actually not true. The Ontario Ministry of the Environment’s limit of 0.35 ng/m3 is 103-times more protective than New Hampshire’s.

Yet MacMillan and Farrer term New Hampshire’s the lowest, most protective levels. An OHA/DEQ release says, “The most stringent of the three short-term values obtained from New Hampshire, Texas, and Ontario, CA guidance was used as the short-term concentration….” And here’s why the agencies can claim to use the lowest figure in New Hampshire’s: “Only non-cancer values were used.”

Though Farrer has for weeks refused interview requests, we found him at a recent Environmental Law Education Center seminar. And three times from the podium and a fourth time when interviewed privately, Farrer said that the N.H. Cr VI levels were based not on the risk of cancer—all anyone really cares about with hex chrome—but on “respiratory irritation.” Echoing his agency’s statement that “only non-cancer values were used,” he said that New Hampshire was guarding against “coughing and decreased lung function.”

But if New Hampshire was supposedly ignoring cancer, and DEQ/OHA threw a dart at the wall to pick its level on that basis—that doesn’t fly. Tom Niejadlik, an air toxics administrator for the New Hampshire Department of Environmental Services, says the state level is based on lung cancer. What’s more, the private group, the American Conference of Governmental Industrial Hygienists, that N.H. derived its figure from, also refers to lung cancer. Ryan Peltier, its science and education manager, says the ACGIH sets its “threshold limit values” to be protective against both lung cancer and upper respiratory irritation.

Farrer said he rejected Ontario’s conservative, 0.35 ng/m3 level since, in his view, it extrapolates mathematically from long-term exposure. “I don’t believe a single, 24-hour exposure can influence life-time risk.” Many toxicologists agree. But Nancy McCarroll, an EPA geneticist, has published research on Cr VI that states, “Within the time-frame and [Cr VI tumor-causing] dose range for early events, genetic changes in mice (single/double-stranded DNA breaks) commence within 24 hours.” Such DNA damage is a contributing factor to cancer.

Consider New Hampshire’s annual limit for some perspective on how conservative and protective its daily limit might be. For a year—a figure akin to Oregon’s ABC—New Hampshire’s limit is 24 ng/m3; that’s 300 times higher than Oregon’s annual safe-air goal of 0.08 ng/m3. That might give one pause when plugging a N.H. number into Oregon regulation.

And what does Ontario, with its very low 0.35 ng/m3 level, know about 24-hour exposure that N.H.—and now Oregon might learn? As to the “dose” of Cr VI that might cause lung cancer, the World Health Organization states, “no safe level of chromium(VI) can be recommended.” Without specifying exposure duration, the WHO asserts that as little as 2.5 ng/m3 might cause an additional 1 in 10,000 cancer cases. (In epidemiology, 1 in 100,000 is considered significant.)

One final note on the 36 ng/m3 daily limit: neither MacMillan nor Farrer knew how many days a year Uroboros—right down the hill from Tubman School—or Bullseye might approach this single-day level before exceeding the much lower annual limit. In other words, how many days a year might nearby neighbors be awash in levels of Cr VI approaching 36 ng/m3? Farrer said, “I don’t know, it’s a mathematical question.” MacMillan said, “If it’s not in the rules, I don’t know…. I don’t know how that applies in the practical world.” This from the two professionals signing off on this provision in the proposed rules.

Despite all this, the revised rules may be an improvement of their original iteration presented to the public on 3/14/16, a day before they were slated to be voted on by the EQC. They were changed in large part in response to some 1200 public comments on EQC’s website.

One difference: the prior version relied on a technically unfeasible stack-test of its permissible level of Cr VI, which was 1.6 ng/m3 (itself a twenty-fold increase over Oregon’s ABC). What’s more, the prior rules displayed a lack of understanding, which DEQ freely admitted, that significant portions—a third, a half or more, research varies, as will be discussed—of relatively harmless trivalent chromium (Cr III) converts to Cr VI when heated in a 2500-degree Fahrenheit glass-melt furnace.

To DEQ’s credit, the new rules control lead and manganese, hazardous air pollutants that are regulated by the EPA; the prior version ignored them. But there’s still no mention of fluorine, a toxic that both Uroboros and Bullseye use extensively to produce their opalescent glass.

However, under the prior proposal—its first version—glassmakers would have had to stack-test for hex chrome at that 1.6 ng/m3 level. And some air testing experts say that’s just short of impossible to achieve in a glass-furnace stack, where temperatures within the stack might reach, the EPA estimates, from 600 to 1000 F. “There’s no way to get there from here,” said Bill Mullins, technical director of Alliance Source Testing.

Agreeing that DEQ would have set the glass makers an impossible task testing for 1.6 ng/m3, Darrell Doerle, senior staff scientist at Air-Tech Environmental, said, “It sounds like the regulator needs to sit down and talk to someone about how this is done.” (Both Mullins and Doerle boast 40 years each of air monitoring experience.) A basic check of EPA standards on measuring Cr VI indicates that 1.6 ng/m3 is five times lower than the 8 ng/m3 set by the EPA as its lowest detection limit.


Even in the new version of the rules, the one to be voted up or down by the EQC on Thursday, the two glass companies will be allowed to use Cr III in an uncontrolled furnace potentially for four months provided they don’t yield more than 36 ng/m3 a day of Cr VI. Given that DEQ has admitted it has little quantitative understanding of how—and how much—Cr III might convert to Cr VI; given the uncertainties and technical challenges of measuring Cr VI on a glass furnace’s hot-stack; given that the proposed rules require no monitoring off-site, just modeling—why the rigmarole to allow chromium to be melted in a glass furnace with a naked stack for the four months before the September 1st deadline for emission controls on all furnaces?

George Davis, a DEQ environmental engineer, said it came down to the time needed, a matter of months, for Bullseye to get a large, comprehensive baghouse designed and built, one that can cut pollution from mutliple furnaces. And the company wants to make green glass from chromium, some 30 percent of its product line, in the meantime. (Though much smaller, Uroboros’s chromium-green output is roughly the same percentage of its total.)

According to Davis, the one, modest baghouse Bullseye has completed is “a pilot project too small to control multiple furnaces.” And it was DEQ’s “understanding,” Davis said, that the company wants to use Cr III in the meantime with no emission controls. That’s why some two-dozen supporters came to plead with the EQC on March 15 to let Bullseye use chromium in an uncontrolled furnace. “Based on what we’ve been told by Bullseye, they want to use trivalent chromium, and so it went into the temporary rule,” Davis said.

Davis said provisions regarding “the uncontrolled furnace came about when we were given direction to develop the rule.” He added that the signed Uroboros agreement was “hammered out” between its lawyer, Leonard DuBoff of the DuBoff Law Group, and DEQ and its legal counsel, Paul Garrahan of the state Department of Justice. There were also negotiations, currently stalled, with Bullseye’s attorneys, Jeffrey Hunter and Ivan Gold of Perkins Coie. Bullseye apparently decided to wait on EQC’s decision.

DEQ compliance and enforcement manager Leah Feldon said that while the DEQ team didn’t sit with its opposite numbers at a table discussing working drafts of the rules and agreements, there were “a lot” of emails back and forth with the glassmakers’ lawyers.

Hunter sent one such email to Feldon and Garrahan on Friday, 3/4/16. Referring to production Bullseye wished to start up the following Monday, Bullseye’s lawyer wrote: “Due to prior contractual commitments, Bullseye must make glass with Cr III…. In order to make this glass, Bullseye will need to use a total of 5.1 lbs. of Cr III per day.” The number of days is not specified.

Hunter did refer to 2,420 pounds of green and violet glass, 1,590 pounds of it being two shades of “Opalescent” green—the sort of glass that typically uses fluorine. And thus toxic fluorine gas Bullseye wanted to loose upon the world from a naked stack—just as it had been doing for many, many years.

Bullseye lists between 16,000 and 64,997 pounds of three different fluorine compounds on file with the Oregon Office of State Fire Marshal—enough for a whole lot of opalescent glass. Yet neither the signed Uroboros agreement nor either version of DEQ’s proposed rules mention fluorine even once. There’s no discussion of controlling its emissions. (A single public comment on the initial proposed rule asks about fluorine, and DEQ fobs regulatory responsibility off on the Oregon Occupational Safety and Health Division.)

As hydrofluoric acid, fluorine has etched Uroboros’s windows as well as the lungs of anyone in the vicinity. [See Portland Mercury

Despite a published account of Uroboros’s extensive use of fluorine, DEQ says it’s aware of its use, but has not yet examined the issue.

As to the violet glass that Bullseye wants to make with Cr III, three glassmen have told this reporter that such violet glass also employs manganese dioxide to yield attractive purples. And Bullseye lists between 1,000 and 4,999 pounds of maganese dioxide in its fire marshal data-base. According to published research on the conversion of Cr III to Cr VI: “Trivalent chromium could also be oxidized to hexavalent form through interaction with manganese dioxide….” That finding was echoed by Robert Strongin, a professor of chemistry at Portland State University.

Hunter’s email continued: “Prohibiting Bullseye from using a small quantity of Cr III before Thursday provides no public benefit but will cause substantial economic hardship to Bullseye.”

Then, referring to Feldon and Garrahan’s top boss, came the email’s most trenchant sentence: “We would appreciate you sharing this with the Governor’s Office.” Of course, this shot over the bow to these state employees didn’t preclude Bullseye from contacting Governor Kate Brown itself.

Said one senior government official familiar with DEQ and the governor’s office’s dealings with the glass companies, “I think there was a lot of pressure from the governor’s office on DEQ concerned about the jobs at Bullseye and Uroboros.” This official, who asked not to be named, said that Gabriella Goldfarb, the governor’s Natural Resources Policy Adviser, was involved. “But the governor’s office doesn’t realize how strong the fear and anger people have. When 700 people gather at Cleveland High School, they’re not the tin-foil hat” brigade, this official said.

Goldfarb declined to be interviewed (as did Garrahan). Asked generally about the new initial rules, on March 15, the day EQC met to potentially pass them, Melissa Navas, Brown’s press secretary, emailed to say, “Governor Brown heard communities loud and clear that they wanted swift action to address their concerns, and that’s what she’s asked DEQ to do.” So, according to Navas, Brown gave the DEQ its marching orders.

But a week later, after several attempts to reach Goldfarb, plus questions to Navas about Goldfarb potentially spurring on the rules and agreement on CR III in an uncontrolled furnace, Navas responded differently: “From the time the Governor learned about potentially concerning industrial air emissions in Portland, her direction to the agencies was to act expeditiously to inform the public and ensure that companies were operating in a manner protective of public health…. DEQ initiated the temporary rule making process and generated the content of the rules, informed in part by consultation with the Oregon Health Authority.”


Davis noted that his agency’s October 2015 air monitoring program outside Bullseye—the one prompted by the U.S. Forest Service’s initial, high heavy-metal readings found in moss—didn’t differentiate between different types of chromium. “We don’t have specific information regarding hexavalent chromium,” he said. “Because of that, we’re very cautious in our approach. We could have forced pollution controls if we knew about any high levels” of Cr VI.

In effect, DEQ felt it didn’t have the cudgel of high Cr VI readings to force Bullseye’s hand. This though there were readings in October for total chromium of 407 ng/m3 and of 440 ng/m3, the latter 5,494 times the state’s safe-air goal for Cr VI. This though the average for the month was 71.5 ng/m3. It’s worth noting that California air quality regulators recommend “assuming all total chrome is hexavalent.”

More from Hunter’s email: “In Bullseye’s discussions with DEQ, which have included experts from EPA, there has been no scientific evidence proving that Cr III will convert to Cr VI in Bullseye’s furnaces.” This contention was echoed numerous times by Bullseye partisans before the EQC.

Bullseye lists up to 5,498 pounds of Cr VI compounds in the fire marshal’s data-base. Asked about the amount of hex-chrome Bullseye has on hand according to this public record, Feldon said, “I won’t speculate about that. I’m not aware of that.”

Several DEQ officials have publicly stated their personal and institutional lack of understanding of the fact that Cr III can oxidize to lose electrons and become Cr VI. On the flip side, Cr VI can reduce—or gain electrons—to convert to Cr III. (Think of an OIL RIG: oxidation is loss; reduction is gain.)

At the 3/15/16 EQC hearing, DEQ officials said they weren’t sure about all this. Acknowledging that it might happen, Feldon told the commissioners, “We don’t have a lot of information about this conversion. We don’t have good data about this conversion.” Asked about chromium shedding and gaining electrons, Sue MacMillan said, “We’re still trying to figure that out, the potential for it.”

Yet published research is available on what percentage of total chromium might convert to Cr VI when subjected to high temperatures such as in a 2500-degree Fahrenheit glass furnace.

The EPA published a “Health Assessment” on chromium back in April 1990 that starkly declared in a table that “Glass Manufacturing Cooling Towers” emit a small amount of “Chromium Emissions” compared to power plants and specialty steel manufacturers and the like. But of those emissions, the percentage for glass makers of “Estimated Cr(VI)” is 100 percent. Yes, all of it, according to the EPA.

Robert Strongin, the PSU professor, emailed to say, “I have attached evidence from two peer-reviewed journals showing that heating Cr(III) in the presence of oxygen can lead to significant amounts of Cr (VI).”

Bullseye has been using liquid oxygen in its furnaces as a fuel additive to save on natural-gas costs and cut greenhouse gas emissions since 2005.

One study Strongin referenced found that chrome oxide, a form of trivalent chromium Bullseye stated to this reporter that it will be using to make its green glass, “could be readily converted to hexavalent chromium when heated in the presence of oxygen at temperatures of 200–300 C[elsius]. The conversion rate was up to 50% at 300 C [or 572 Farenheit] in 12 hours.”

The second paper Strongin offered examined the burning in a laboratory of the sort of wild grass typically consumed by brush fires. Researchers “confirm[ed] the possibility of Cr III to Cr VI oxidation with atmospheric oxygen. It was found that the initial content of Cr VI  in grass, 2.5%, in respect to the total amount of Cr, increased to 23% after 1 hour of heating at 500 C and to 58% at 900 C.”

Now, burning grass is pretty far afield from molten glass, true. But the observed increase in Cr VI from 2.5% of total chromium to 58% after heating at a high temperature is still disturbing in the context of DEQ allowing Cr III in uncontrolled furnaces.

David Kalman, an environmental sciences professor at the University of Washington, emailed, “I found one article describing modeling simulations of the potential for producing hexavalent chromium in incineration at temperatures up to 2000°F. It finds that Cr VI in a gaseous form … is the dominant species formed at these temperatures with abundant oxygen….This is true even if the feedstock chromium is the trivalent form.”

Apply heat, and add the oxygen Bullseye has been using for over a decade, and voila: Cr III in, Cr VI out—or so it would seem, according to Kalman.

The North Carolina Department of Labor discusses chromium exposed to high temperatures: “In these situations, the chromium is not originally in the hexavalent state but at sufficiently high temperatures undergoes oxidation (i.e., loses electrons) to yield the hexavalent form.”

Finally, consider oxygen’s effect in an EPA study of burning sewage: the hexavalent to total chromium ratio increased from 11.9% to 29.9% with added oxygen. The explanation: [T]he higher hearth temperatures and excess oxygen levels recorded during the improved combustion condition favor the formation of hexavalent chromium.”

So, estimates of Cr VI’s percentage of total chromium fugitive emissions exiting an untreated glass stack range from the EPA’s 100 % to 58% and 50% on down to 29.9%. Apply that lowest figure, call it 30%, to the monthly average DEQ found around Bullseye for October 2015: 71.5 ng/m3. That works out to an estimated 21.45 ng/m3 of hex-chrome for the month. You live around there, don’t even do the math for the two off-the-charts days of over 400 ng/m3, not if you want to sleep at night. Not if you got kids. Speaking at that seminar, David Farrer said the people living there have likely suffered high levels of exposure in the past. “Based on the October levels we found, that’s possible.”

Daniel Forbes is the author of Derail This Train Wreck found here. He lives in Portland, and can be reached at Other articles on air toxics found here.

Scroll to Top