May/June 2005
Despite radiation detection equipment that scanned incoming scrap, a Timken minimill in Ohio melted a radioactive source in May 2004. An 11-day shutdown and cleanup followed, along with major changes to the way Timken handles its scrap stream.
By Robert L. Reid
Early on the morning of May 4, 2004, a truck leaving The Timken Co.’s Faircrest steel plant in Canton, Ohio, set off a radiation alarm at the minimill’s gatehouse. That in itself wasn’t unusual. The truck was carrying baghouse dust from the steelmaker’s electric-arc furnace (EAF), bound for a contracted disposal site. Such dust often contains radon, a naturally occurring radioactive material that decays rapidly. While radon in baghouse dust is generally not harmful, it will set off radiation detection equipment, notes Andy Black, Timken’s manager of technical services for the steel group.
Unfortunately for Timken, radon wasn’t the cause of the May 4 alarm. Instead, the mill had melted a radioactive source in its scrap charge.
Ultimately, the mill had to shut down for 11 days to clean up the affected areas. Customers were told that their scheduled steel deliveries—more than 5,000 orders representing 60,000 tons of steel—would be delayed indefinitely. Contractors wearing protective clothing from head to foot and breathing from air tanks removed and safely disposed of more than 4,000 dust-containing bags from the mill’s baghouse. A veritable city-within-the-site had to be set up for the cleanup crews to provide meals, restroom facilities, protection from inclement weather, and mobile showers for their contaminated clothing.
In the end, a load of contaminated scrap valued at less than $40,000 cost Timken more than $15 million in cleanup costs and forced the company to review many of its policies and procedures—both internally and externally regarding its scrap suppliers. Timken also bought new and additional radiation detection equipment, retrained scores of employees, and helped launch a new effort with ReMA and other groups to keep radioactive sources out of the scrap stream.
Despite the trials of this experience, Timken considers itself lucky—lucky, first of all, because no one was injured in the incident. “Our baghouse did its job and captured all the radioactive material, so no one was exposed to this deadly material,” noted Tim Timken, president of the steel group, at a scrap supplier conference five months after the incident.
The company also feels lucky because the source itself—believed to be an industrial gauge—contained radioactive cesium-137, which had a low-enough melting point to simply vaporize in the mill’s furnace. While the cesium did contaminate the baghouse dust, ductwork, and other parts of the pollution control system, it didn’t contaminate the steel itself, Black notes. Had the radioactive source been something like cobalt-60, he says, it would have created worse problems, melting into the steel and then contaminating everything the molten steel touched, from the furnace to the slag.
A Sick Feeling
On the morning of the alarm, Black and others were in a normal scheduling meeting at the Faircrest plant when the vice president of operations and melt shop manager were suddenly called out of the meeting.
“I don’t remember if it was a nice day, a cold day, a warm day, or anything along those lines—all I remember is that it was very strange that our meeting was interrupted,” Black recalls. “I can just remember the sick feeling that went through everybody when they realized what had happened.”
It wasn’t supposed to happen, of course. At the time, Timken thought it had adequate defenses against radiation. After all, the firm had detection equipment that scanned incoming truck and rail shipments of scrap. Also, there was no record of an alarm around the time the source was melted. (Due to the nature of Timken’s operations and its scrap inventories, the company does not know exactly when the radioactive source arrived or when it was melted.)
Once Timken determined that the truck alarm meant a radioactive source had been melted, a set protocol kicked into action. The company’s radiation safety officer was notified, along with the appropriate authorities—in this case, the Ohio Department of Health (ODH) and the Ohio Environmental Protection Agency. Working with an ODH team and outside experts, Timken first determined that the incident—identified as involving cesium-137—posed no health risk to the plant’s employees, the general public, or the environment. After examining the immediate area around the furnace and the slag processing facility, Timken found that the contamination was limited to the baghouse and dust collection system. The radiation levels were less than the radiation used during a chest X-ray, the mill found, but the cesium had attached itself to the dust particles, which consisted of lead and cadmium. Though harmful if inhaled, lead and cadmium were concerns that Timken knew how to safely address, which helped lower employee anxiety, notes an internal Timken report.
Fortunately, since the radiation was concentrated in the dust collection system, Timken didn’t have to shut down the mill immediately. “It was decided that if we continued to run the furnace, the system would clean itself out, with the contamination settling in the 4,080 bags that collect the dust,” explains Harriet Dutka, manager of steelmaking-bar manufacturing, according to the Timken report.
This cleaning-while-running phase lasted several days, during which all contaminated dust was contained on-site, adds Andy Black. Then the real work began.
A ‘Herculean’ Cleanup
“Cleaning up radioactive debris takes a Herculean effort,” notes the Timken report. Indeed, the plant quickly found itself hosting a small city of several hundred contractors from 50 firms across the United States, led by Duratek Inc. (Columbia, Md.).
“A tent city was built to handle sanitary facilities, showers, food, testing, storerooms,” and more, Timken says. “The workers filled all of the hotels in Canton and some had to be sent as far away as Akron, 20 miles north.”
The Faircrest mill officially stopped operating at 7 a.m. on May 11, one week after the radiation problem was discovered. Working around the clock, the contractors and Timken employees vacuumed out the contaminated dust, replaced the 4,080 baghouse bags, and scrubbed all contaminated surfaces. Then everything had to be tested to make sure it was free of radioactivity before being “released” to go back into operation, Black says.
“You can’t imagine how many garbage barrels you have to dump in a situation like that,” notes Harriet Dutka. In the end, more than 100 large containers holding more than 1,000 tons of dust and debris had to be shipped to licensed radioactive material disposal sites throughout the United States. The final shipment went out in September 2004, notes an ODH timeline of the incident.
Though the Faircrest remediation was scheduled to take three weeks, Timken and its contractors finished in 11 days. The company restarted its furnace shortly after 7 a.m. on Sunday, May 22.
If the cleanup effort was Herculean, the final costs to Timken were positively gargantuan—somewhere between $10 million and $15 million, plus another $750,000 in new and improved radiation detection equipment, notes Black.
Revamping Procedures and Practices
Two days before the alarm, the Faircrest furnace’s exhaust emissions control system recorded a spike in the lead level. Timken believes that spike represents the roughly 300 to 400 pounds of lead that likely shielded the pencil eraser-sized radioactive source itself.
To date, neither Timken nor ODH has been able to identify the supplier of the radioactive source, though the mill’s tracking system for scrap loads has narrowed the possibilities down to four dealers. Each of those firms had radiation detection equipment at the time of the incident, and each is still supplying scrap to Timken, notes Randy Ehret, general manager, strategic sourcing. One of the firms, however, did broker material that might not have been checked with radiation detectors.
“We don’t believe anything was malicious on anybody’s part,” Ehret says. “We all thought we had procedures and practices that were good enough to prevent this sort of thing, but something still got through.”
Following the cleanup, Andy Black, who normally has the engineering responsibilities for capital improvements in the steel group, was charged with putting together a new radiation detection program to prevent a similar problem at Faircrest or Timken’s other minimill in Canton, the Harrison steel plant. Prior to the May 2004 incident, Timken expected its scrap suppliers to screen their material for radiation before shipping it, but such screenings were not verified and, hence, not guaranteed. The mill also ran all inbound scrap shipments through a single radiation scan at its gate. “We came to realize that that simply is not sufficient,” Black says.
It’s important, he adds, to periodically evaluate the capabilities of radiation detection equipment as newer and more effective models are developed. Scintillators tend to age and new algorithms are being developed to distinguish real problems from background radiation. Also, better data collection is now available to record alarms caused by everything from a truck or railcar that moved too swiftly past the sensors to problems with background radiation or other false alarms. That data is important because a shipment of scrap might sit in inventory for several months before being melted, Black notes.
To prevent another such incident, then, Timken updated its equipment from Exploranium GR-256s panels to newer Exploranium AT-900 units. The steelmaker also switched from a two-panel to a four-panel system and essentially tripled the number of times prior to melting it scans scrap for radiation. Plus, Timken began checking “high-risk” loads of scrap with new handheld monitors, Black says. Scrap from a supplier that lacks adequate radiation equipment or that can’t show that it properly trains its employees on radiation issues is generally designated “high risk,” he explains.
Now, all trucks and railcars undergo two inbound scans before reaching the melt shop, where more detectors scan each magnetload of scrap prior to charging. “We believe that only through multiple scans can we assure ourselves of not repeating a similar incident,” Black asserts.
If a radioactive source somehow manages to slip by, Timken has plans to install new detectors in the mill’s baghouse. The company is also adding equipment and procedures to more precisely identify the supplier of any future radioactive source. That would enable Timken to charge the costs of any resulting cleanup or other business disruptions back to the scrap supplier.
Retraining and Reinforcing
Timken’s new approach to radiation protection didn’t stop at technology. The company also “basically started over with training,” Black says. The firm sent some 60 employees—including all the detection equipment operators, melt shop workers who load scrap into charge buckets, and supervisors—through a three-day radiation safety course. The training included classroom instruction, practical exercises, and a real-time exercise in the plant.
Topics ranged from the proper operation of Timken’s new stationary and portable detection equipment to fundamentals of radiation, exposure limits, emergency procedures, how to safely investigate potential radioactive sources, contamination/decontamination issues, and more. The program also included refresher training on site-specific radiation safety in the Faircrest plant and offered certain employees the opportunity for additional OSHA-specific training.
In addition, Timken spent considerable time calibrating its new equipment to distinguish between background radiation and real sources. The firm reinforced existing procedures such as the 3-mph speed limit for vehicles and railcars passing by the sensors, and it stressed that no alarm should be treated as a false alarm.
Thus, Timken adopted a “no assumption” policy that treats every alarm as a potentially serious radiation source. It established procedures to document that operators and supervisors either confirmed that scrap deliveries were alarm-free or that they properly investigated the cause of any alarm. For instance, they must document that a truck was rerouted through the detectors if it exceeded the 3-mph limit and set off a speed alarm, Black notes. Likewise, operators must sign off that they performed a final screening prior to the scrap’s delivery to the melt shop.
Thanks to the new equipment and procedures at Faircrest, plant personnel have stopped and rejected scrap shipments that contained “radioactive materials,” Black says, though he hesitates to identify the suspect material as actual radioactive sources. Some alarms were likely caused by naturally occurring radioactive materials, he notes. Also, all of the suspect shipments involved material with low radiation levels, enabling Timken to simply reject the loads rather than quarantine them on-site for examination by state officials. Timken does notify ODH of all such incidents, which are then investigated by the state agency, Black says.
Assessing Risks and Seeking Solutions
Though Timken has considered a blanket refusal to buy scrap from any dealer without radiation detection equipment, the company hasn’t taken that step yet. The company has created a “risk matrix” of scrap suppliers, however. This matrix, Black explains, compares every supplier’s radiation equipment and procedures to determine what additional actions—such as unloading the material on the ground and checking it with handheld monitors—might be necessary to ensure the safety of each supplier’s material.
“We’re consciously factoring into our value equation the amount of radiation screening that suppliers do, and we are identifying those suppliers that don’t have good procedures and practices,” Randy Ehret notes, adding, “I’m going to buy all the material I can from people who do have good radiation detection equipment.” Whenever Timken buys from “second-tier” suppliers whose material requires additional screening, it factors that extra cost into the price it pays for the scrap.
Timken has even begun auditing some suppliers to verify the information they provide on their radiation screening equipment and procedures—something the mill has never done before, notes Ehret.
The steelmaker also held a scrap supplier conference in Canton in October 2004 that focused largely on its radiation incident and the measures it has taken to prevent a similar occurrence. More than 150 people attended the event, which included talks by steel group president Tim Timken, other Timken officials (including Randy Ehret and Andy Black), and a consultant who assisted with the site cleanup.
Participants received a 140-page binder detailing the May 2004 incident and Timken’s response.
Ehret describes the feedback from the company’s scrap suppliers as “very supportive ... I think the [scrap] industry as a whole didn’t realize the risk still existed even with radiation protection equipment in place.” As such, Timken’s new approach emphasizes that “radiation detection equipment alone is not enough,” he says.
Toward that end, the company helped create a new radioactivity task force last June with ReMA and other groups—including the American Iron and Steel Institute, the aluminum industry, and stainless steel producers—to develop a better approach to detecting and disposing of radioactive sources.
Most mills send suspect materials back to the supplier and hope that the materials “will be disposed of properly,” noted Andy Black during the scrap supplier conference. “I believe we need to capture these materials and assure ourselves that they are taken out of the recycling stream and disposed of in a safe and proper manner.”
The task force’s proposals include establishing a national program for “orphan sources” (radioactive sources whose owners can’t be identified), providing financial incentives for scrap companies to install radiation detectors and offering a “bounty” to reward dealers that recover radioactive sources at their facilities. Given the national security issues surrounding lost radioactive material, the Department of Homeland Security might even be approached for support, some believe.
Andy Black also hopes that ISRI’s new Recycling Industries’ Operating Standard—called RIOS—will include information on the proper procedures and equipment for keeping radioactive sources out of the scrap stream.
At presstime, ReMA planned to explore the task force’s proposals at its annual convention in New Orleans. Both Black and Ehret hope the task force will move decisively and quickly on the issue. Given Timken’s radiation incident, their sense of urgency is understandable. As Ehret asserts, “You have to experience the problem firsthand to understand the risk, both in dollar terms and the potential health risks to your employees.”
For those reasons and others, he notes, the issue of radioactive sources in scrap must gain a higher priority for both scrap processors and their consumers.
Robert L. Reid is managing editor of Scrap.
Despite radiation detection equipment that scanned incoming scrap, a Timken minimill in Ohio melted a radioactive source in May 2004. An 11-day shutdown and cleanup followed, along with major changes to the way Timken handles its scrap stream.