Flickr / Los Alamos National Laboratories
The partially disassembled Godiva IV in 2011.
Not a clue.
The government scientists didn’t know they were breathing in radioactive uranium at the time it was happening. In fact, most didn’t learn about their exposure for months, long after they returned home from the nuclear weapons research center where they had inhaled it.
The entire event was characterized by sloppiness, according to a quiet federal investigation, with multiple warnings issued and ignored in advance, and new episodes of contamination allowed to occur afterward. All of this transpired without public notice by the center.
Here’s how it happened: In April and May 2014, an elite group of 97 nuclear researchers from as far away as the U.K. gathered in a remote corner of Nye County, Nev., at the historic site where the U.S. had exploded hundreds of its nuclear weapons. With nuclear bomb testing ended, the scientists were using a device they called Godiva at the National Criticality Experiments Research Center to test nuclear pulses on a smaller and supposedly safe scale.
But as the technicians prepared for their experiments that spring — under significant pressure to clear a major backlog of work and to operate the machine at what a report called Godiva’s “upper energy range”— they committed several grievous errors, according to government reports.
The machine had been moved to Nevada nine years earlier from Los Alamos, N.M. But a shroud, descriptively called Top Hat, which should have covered the machine and prevented the escape of any loose radioactive particles, was not reinstalled when it was reassembled in 2012.
Also, because Godiva’s bursts tended to set off multiple radiation alarms in the center, the experimenters decided to switch the alarm system off. But because the alarms were connected to the ventilation and air filter system for the room, those were shut off as well. The only ventilation remaining was a small exhaust fan that vented into an adjacent anteroom where researchers gathered before and after experiments.
On June 16, 2014, a month after the experiments were completed, technicians doing routine tests made an alarming discovery — radioactive particles were in the anteroom. They then checked the room holding Godiva, and found radiation 20 times more intense there. The Nevada site’s managers, who work for a group of private, profit-making contractors — like most U.S. nuclear weapons personnel — ordered the rooms decontaminated. But they didn’t immediately check exposures among the scientists and researchers who had gathered for the tests, many of whom had already gone back to their own labs.
None had any clue about the mishap until two months after the experiments, on July 17, when one of them — a researcher from Lawrence Livermore National Laboratory nuclear weapons lab in California — got the results from his routine radiation monitoring. His urine tested positive for exposure to enriched uranium particles.
National Security Technologies, LLC (NSTec), the lead contractor that runs the Nevada site, subsequently collected urine specimens from its own workers who’d been in the room with Godiva during the experiments. It discovered three of its technicians also had inhaled highly-enriched uranium.
News quickly spread, but only among the scientists and their bosses, who were accustomed to a shroud of official secrecy covering their work; no public announcement was made. According to an initial U.S. Department of Energy (DOE) investigative report dated April 28, 2015, calls eventually went out to test the 97 people present for the Godiva experiments. But for reasons that remain unclear the testing went very slowly, and not until 2016 did the DOE state that 31 were discovered to have inhaled uranium.
In a letter last summer to the Los Alamos and Nevada lab directors, National Nuclear Security Administration (NNSA) Administrator Frank Klotz suggested that the employees’ radiation doses were not large — at the high end, they were roughly equivalent to 13 chest x-rays. But once inhaled, uranium particles can keep emitting radiation for years, and so they pose an added cancer risk. Klotz’s letter deemed the exposures “safety-significant and preventable.” It could have been even worse, of course, given the absence of any timely warning.
Lab operations riddled with errors
The four key national facilities involved in the underlying experimentation — Los Alamos National Laboratory, Lawrence Livermore, the Nevada Test Site and Sandia National Laboratory — are among the U.S.’s premier scientific labs. They collectively employ more than 26,000 people engaged in cutting-edge and often dangerous work, governed by myriad nuclear safety regulations, with two major contract enforcement mechanisms meant to inflict financial pain when needed on the private corporations that operate them.
And yet in this case, and in others like it, not only were the labs’ procedures and responses riddled with errors, but even after attention was called to these incidents, other safety mishaps occurred. And the financial penalties imposed by the government didn’t seem to have a major impact on the labs’ conduct.
A review by The Center for Public Integrity (CPI) of more than 60 safety mishaps at 10 nuclear weapons–related federal sites that were flagged in special, internal reports to Washington, along with dozens of interviews of officials and experts, revealed a protective system that is weak, if not truly dysfunctional: Fines are frequently reduced or waived while contractors are awarded large profits. Auditors say labs and production plants are overseen by an inadequately staffed NNSA and DOE, which as a result largely rely on the contractors to police themselves.
The CPI probe, partly based on documents obtained under the Freedom of Information Act, reveals a system in which extra profit is awarded under a rating profile that persistently places higher priority on the nuclear weapons labs’ national security “mission” than on worker protections, putting production far ahead of safety. Experts say it is a practice in keeping with a culture of urgent, no-holds-barred work that took root in the nuclear weapons complex during World War II. These production pressures flow down to the highly secured rooms where workers labor with special clearances, routinely handling highly toxic and explosive materials.
The DOE typically gives the private companies involved a financial bonus when they accomplish their missions on a deadline, notes Ralph Stanton, one of 16 workers exposed to radiation in an incident at Idaho National Laboratory in 2011. “When the [bonus] milestone is in play” — meaning on the occasions corporate lab operators gain extra pay by meeting production deadlines — some of the workers feel that “safety is completely gone.”
Tracy Bower, a spokesperson for the contractors that operate the Nevada Test Site, said its performance “scores” are high and that “our primary concern has been and continues to be the safety of our employees and our community.”
When the Nevada accident occurred, the Godiva experiments had considerable urgency. After being moved to Nevada from Los Alamos, the machine was originally supposed to resume operations in 2010, near three similar machines inside a tubular-shaped building about the size of two football fields, with a striking white top barely visible aboveground because it is swaddled in compacted earth.
But in August of that year the Defense Nuclear Facilities Safety Board, an independent federal safety agency that monitors nuclear weapons operations, warned in a letter to NNSA that it was concerned about “deficiencies” at the experimental facility housing Godiva, including potentially unreliable radiation alarms. Neither NNSA nor the site’s private contractor had “conducted sufficiently detailed design reviews of the facility,” the board’s chair, Peter Winokur, wrote.
Godiva’s operation was postponed, but the controlled bursts of radiation were finally restarted in September 2013. That same month the DOE’s Office of Inspector General warned that some safety issues persisted with Godiva, including flaws in some instrumentation related to safety and incomplete documents meant to guide workers safely through their tasks. The inspector general’s report also warned that more than a third of the time, when the contractors involved said they had fixed problems flagged by the government, they hadn’t actually done so. The inspector general blamed the delays and problems, in part, on “weaknesses in federal oversight.”
A workplace hazardous to whistleblowers
But inadequate supervision from Washington was only part of the problem. According to a separate August 2013 report by the DOE’s workplace health and safety office, the leaders of NSTec were generally inattentive to workers’ safety concerns and resentful of NNSA oversight. Amid a rise in workplace injury and illnesses at the site — which the report did not explain — the contractor had depicted the government’s inquiries about it as an “overreaction.”
The report said that because the contractor did not consistently encourage open communications, frontline workers feared retaliation if they complained to their supervisors about safety issues. It said two such cases of retaliation surfaced during the safety office’s inquiry. Overall, the report said, significant stresses existed on the site’s “safety and security culture” and there were “indications of a chilling effect” on worried workers.
It was against this backdrop that the Godiva contaminations played out over several months the following year. The device is an ugly stack of high-tech machinery, shelving and wires that stands as tall as a man. Inside are rings and blocks of highly enriched uranium, a key nuclear explosive; they are made to fission, just slightly, when they are delicately moved closer by at least two technicians.
The point is to generate modest bursts of radiation useful for researching nuclear power, training nuclear safety experts, predicting weapons effects and making bomb components less vulnerable to radiation storms in a nuclear war. Those who gathered for the experiments came from Los Alamos, which oversaw the work; Lawrence Livermore; Sandia; Pacific Northwest National Laboratory in Washington State; Oak Ridge National Laboratory in Tennessee; the Nevada Field Office of the National Nuclear Security Administration; and the U.K.’s Atomic Weapons Establishment.
In a push to make up for lost time, experimentation on Godiva was placed on an accelerated schedule. As a September 2016 report detailing the NNSA’s review of the contaminations noted, “During April and May 2014, there was an increased campaign of Godiva burst operations, some of which were in the upper energy range of Godiva’s capability.”
All the while, with radiation alarms silenced, the particles of uranium were being piped into the main Godiva room and an area where experimenters thought they’d be safe. Although the contamination started showing up in June, it was not until August 7, 2014, that the DoE’s Nevada Field Office at the site shut down Godiva’s operations.
Another 11 days passed before technicians noted even higher contamination levels inside the room that housed the machine. The contractors involved — from Los Alamos, Lawrence Livermore and the Nevada test site — didn’t determine that this incident merited a formal safety report to Washington for another two weeks.
A spokesperson for NNSA, Greg Wolf, wrote in an e-mail that all the exposure levels were below regulatory limits and posed minimal health risks. He wrote Godiva’s operations are now being closely monitored and that the contractors that run Nevada and Los Alamos “take any unexpected exposure seriously.”
Still, with those types of exposures, “there is some incremental risk that you might get cancer,” said Joel Lubenau, a certified health physicist, industry consultant and former senior adviser to a chair of the Nuclear Regulatory Commission. He said the degree of risk depended on how large the particles were and whether they were soluble enough to enter the bloodstream and reach key organs. He also said he was highly surprised to hear the alarms had been turned off and that the protective shroud was never installed.
No fines for repeated safety infractions
The fallout from the episode went on for years. On October 21, 2014, four months after the original Godiva leakage, technicians detected excess radioactive particle contamination outside the Flattop critical assembly device, another open-faced radiation-burst machine similar to Godiva, located in an adjacent building.
Contamination also showed up again in the Godiva equipment room. A week later workers shut down Flattop. But oddly, “no formal causal analysis was performed,” according to a 2016 settlement order summarizing DOE investigative findings on the Godiva incident. It was assumed that it was all related to Godiva, the order disclosed.
The settlement order called the Godiva room’s monitoring system inadequate, the hazard controls “insufficiently designed and implemented,” and the management response “slow” and “less than adequate.” But because the NNSA had docked contractors’ profits — $87,000 for NSTec and $500,000 for Los Alamos National Security, LLC — due to overall operational shortcomings, it levied no fines specific to the safety infractions.
And following a common pattern, these events amounted to only a portion of the documented mishaps at both labs during this period. In September 2013, for example, security personnel at the Nevada site unexpectedly found a trailer full of Los Alamos’s low-level radioactive waste sitting in a parking lot outside Gate 100 of the Nevada site. The trailer had been sent there for disposal but the site had refused to accept it because it lacked appropriate shipping papers describing the trailer’s contents, so it was left in the parking lot for a week.
After an investigation, Los Alamos voluntarily suspended its shipment of low-level waste to Nevada. Due to continuing problems with paperwork, Los Alamos wasn’t cleared to restart its waste shipments to Nevada until September 15, 2014, a full year after the trailer was discovered.
During the same period Los Alamos was sending other, more dangerous waste to the DOE’s Waste Isolation Pilot Plant (WIPP) near Carlsbad, N.M., for underground storage. But in December 2013 one of its subcontractors added organic instead of inorganic material to one of the drums — based on a mistyped transcript of an internal meeting where the waste-packaging requirements were discussed. And in February 2014, after being sent into the bowels of WIPP with the wrong material inside, a drum exploded, sending radioactivity throughout the facility and shutting it down for nearly three years — at an estimated $1.5-billion cost to the government. For that, in 2014 the Los Alamos contractor had its profits cut by $57 million, less than 4 percent of the repair bill.
In that episode the final Accident Investigation Board report said supervisors had “fostered a culture where employees do not feel comfortable raising safety issues to management.” Ernest Moniz, secretary of Energy from May 2013 to January 2017, said shortly before leaving that the day the drum exploded at WIPP was by far the worst day of his tenure. Los Alamos, he said in a subsequent interview with CPI, had “inadequate quality control in how they were packaging the wastes. There is no sugarcoating that.”
Moniz also said that during his tenure, there was “a clear pattern” in which the nuclear weapons laboratories were annually ranked more poorly by his department on their “operations” than on their other tasks, with Los Alamos “in particular…the lowest.” He said he had “no doubt they have had some management problems.”
Even after the WIPP debacle, moreover, Los Alamos’ troubles with worker safety were not over. On May 3, 2015, when a worker sprayed cleaning fluid inside a cubicle with high-voltage connections that he thought had been turned off, he caused what’s known as an “arc-flash” explosion that blew him out of the cubicle and burned him over 30 percent of his body. “It caused him enormous suffering,” recalls Don Nichols, who was then the NNSA associate administrator for safety and health. “He made an understandable human error, where controls that should have protected him were not there.”
For these failings and more, NNSA’s then-Principal Deputy Administrator Madelyn Creedon in 2015 cut nearly $6 million from the Los Alamos contractor’s available profit in 2015, and the on-site contracting officer cut another $7.2 million just for the arc-flash. Yet the government still gave the contractor nearly $44 million in profit, or 71 percent of the total available to it.
In 2016, however, it decided to force a new competition for the roughly two billion dollar annual contract to manage the laboratory, a decision that Moniz attributed in the interview to the current contractors’ “operational failure.”
Impunity for top managers
David Overskei, a nuclear physicist and consultant in San Diego who has been hired by both the NNSA and its contractors, said, however, that in his view the award of profit based on performance is “having impact — it’s just not having the desired impact.” Instead of driving consistently good behavior, he said, performance-based profits inspire contractors to find ways to meet minimum standards to collect their bonus.
Overskei, who’s analyzed conditions at NNSA sites and led a congressionally chartered study of the NNSA’s management system in 2006, said in a telephone interview that “there need to be consequences when people don’t do their management job, and the consequence needs to be getting fired. That simply doesn’t happen.”
He pointed out that even though the NNSA withheld substantial profit from the contractors that operate Los Alamos National Laboratory as well as the Waste Isolation Pilot Plant after the 2014 accident there, the top leaders at Los Alamos, where subcontractors made the mistake that caused the radiation-exposure accident, kept their jobs. “In the commercial world, I guarantee you a lot of people would have lost their jobs, and not lower-level people,” Overskei said.
At NSTec, safety problems similarly persisted after the uranium incident. On June 13, 2014, a 55-gallon metal drum containing an explosive chemical blew up as it was being lifted by a worker, sending shrapnel into his leg and even into the tire of a truck some yards away. The worker was blown eight feet out of the shed bay where he was working, leaving him with shrapnel wounds, minor burns to his face and arms, and contusions. An investigation found that isopropyl alcohol in the drum had been transformed by the Nevada desert heat and air into organic peroxides, which are “exceptionally prone to explosive decomposition.”
An accident board determined that the explosion was “wholly preventable” and criticized the contractor team for not having a comprehensive, company-wide chemical safety program, according to the board’s report on June 13, 2014.
Then, on Feb. 26, 2015, a worker was overexposed to vapors in a degreasing operation. The chemical in question, 1-bromopropane, has been linked to neurological illnesses, cancer and reproductive disorders. An enforcement letter from Steven Simonson, director of the DOE’s Office of Enforcement, to NSTec dated Oct. 6, 2015, said the incident “revealed recurring issues in chemical handling and control” that were also evident during the drum explosion. The letter, however, was just a warning and no fines were imposed.
Four months later, on June 24, 2015, some storage drums containing other chemicals spontaneously erupted in a fire. The drums had been stored in a warehouse but moved to an uncovered pad where the blistering Nevada heat hit 102.4 degrees — “temperatures well in excess” of what is considered safe, an enforcement letter from Simonson to NSTec said. It took almost three hours for the fire to burn itself out.
That letter, dated Feb. 22, 2016, cited both the drum explosion and the fire as evidence of similar safety failures: “Both events resulted from improper storage of hazardous chemicals and a lack of understanding regarding storage and environmental factors that affect chemical stability,” the letter said. Once again, the correspondence expressed concern but levied no fines.
The Nevada contracting consortium — a partnership of Northrop Grumman, AECOM, CH2M Hill and Babcock & Wilcox — has run the site since 2006. On Nov. 14, 2014, the year of the Godiva incident, NSTec had its profit cut by $4.2 million — to $26.4 million, or 86 percent of the maximum available — due to what the NNSA called recurrent, systemic problems such as a lack of transparency and accountability, ineffective operations, and weaknesses in its worker safety and health programs, according to a copy of the government’s payment letter.
The following year NSTec again earned roughly the same amount — $26.3 million — in profits, or 90 percent of all that was available. An enforcement notice about the drum explosion, dated Aug. 25, 2015, cited NSTec for multiple failings related to chemical storage and hazards, resulting in four Severity Level I violations and one Severity Level II. But because NSTec had its profits docked for other problems in 2014, NNSA Administrator Klotz levied no safety fines in 2015.
And in 2016 the consortium earned $29.4 million, or 93 percent of what was available that year.
Despite the high performance ratings it gave NSTec, the NNSA announced on May 12, 2017, that Mission Support and Test Services LLC — a consortium of Honeywell, Jacobs Engineering Group Inc. and Stoller Newport News Nuclear, Inc. — would take over operation of the Nevada site at the end of September from NSTec, whose contract is expiring after 11 years. The new contract could span a decade with a total budget of $5 billion if the government awards all extension options, as it did each year that NSTec ran the site.
The consortium that forms NSTec didn’t bid to keep managing the Nevada site, according to Bower, its spokesperson. She did not explain why. “MSTS was awarded the contract because its proposal was determined to represent the best value to the government over the other offerers,” NNSA spokesman Wolf said.
The shift to a new manager has been delayed, however, by a legal fight between the winning bidder and a rival group that wasn’t picked, over the reasonableness of NNSA’s decision. Until it is resolved, Wolf said, NSTec will continue to operate the site.