New troubles are bedeviling ITER, the massive international fusion experiment, on two fronts: at the reactor site in southern France and at the Barcelona, Spain, office that coordinates the 45% share of in-kind contributions the massive project receives from the European Union (EU).

In late January, France’s Nuclear Safety Authority (ASN) ordered a halt to assembly of the ITER vacuum vessel after finding misalignments between the welding surfaces of the first two 440-ton stainless-steel vessel sections. The pair, the first two of nine segments that when joined will form a torus that contains the fusion plasma, had been damaged in transit from South Korea, where they were built.

ASN further told the ITER Organization (IO), which manages ITER construction and is not affiliated with the French government, that the two-meter-thick concrete radiological shielding that’s to be installed around the reactor is inadequate to protect personnel once experimental operations get underway in 2025. ASN also expressed concern that any increase to the shielding could cause the total weight of the reactor to exceed the 140 000-ton capacity of its earthquake-resistant foundation. The three issues will need to be satisfactorily addressed before ASN will lift its construction hold.

The EU’s executive branch, the European Commission, now estimates that completion of ITER construction will be delayed by at least 17 months beyond its official 2025 completion date, not counting any additional delay that could result from the ASN-ordered hiatus. Massimo Garribba, deputy director general of the commission’s energy directorate, told a 28 February hearing of the European Parliament Committee on Budgetary Control that the stretch-out is attributable to delays in the manufacture of one-of-a-kind components and to shortfalls in contributions from several of the member states. The ITER parties are China, the EU, India, Japan, Russia, South Korea, and the US.

A re-baselining of the ITER construction timeline will be presented to the governing ITER Council in June, says Alain Bécoulet, head of ITER’s engineering domain. The new schedule won’t be official until the council’s ratification, expected in November. But the “tsunami of COVID” has delayed the project on the order of 20 months, Bécoulet acknowledges, and has created havoc with construction scheduling. Components being built in South Korea and Italy have been delayed by late deliveries from the manufacturers’ own suppliers. “On top of that, each time you have a sector and you try to get a transporter, you are told, ‘I’m sorry—I have no boats available for six months to a year, and it’s 10 times more expensive than you expected,’” he says.

Since the seven ITER partners use different accounting practices to determine the value of their in-kind contributions, an official cost estimate for ITER construction has never been compiled. The EU in 2017 estimated its 45.6% share of the project will total €18.1 billion ($19.6 billion) from 2017 to 2035, the year experiments with tritium are currently scheduled to begin. By extrapolation, the total ITER cost during that period would be €41 billion if the entire project were to be undertaken in the EU. In 2018, using the US’s 9% share of ITER as a basis, the US Department of Energy estimated ITER’s total cost would be $65 billion if all the work were to be done in the US. (See “ITER disputes DOE’s cost estimate of fusion project,” Physics Today online, 16 April 2018.)

Bécoulet says he expects that ASN will lift its hold by early fall, allowing assembly to resume in November, just a couple months behind schedule. That shouldn’t further delay project completion, he says. “ASN is used to seeing fission reactors one after another, and they know whether they’re good or not immediately. This is a learning process. And we are learning with them.”

The vacuum vessel sections are 17 meters tall and 6–7 meters wide. “You have to fit them together to a fraction of a millimeter,” Bécoulet says. When the sections are assembled, slight adjustments have to be made to account for dimensional nonconformities. The IO was told by ASN last year that it hadn’t satisfactorily demonstrated its procedure for making nuclear-grade welds.

ASN deemed the two-meter-thick concrete radiation shield planned to encapsulate the fusion to be insufficient. But Bécoulet says the regulator’s concern was actually caused by the IO’s “overscrupulousness.” Instead of the “envelopes and margins” that ASN requested, the IO provided a highly detailed three-dimensional model. “The problem is that if you say you will give them a very detailed map, then they will follow the details,” he says. “We are trapped by giving too many details, and the details are still evolving as the design is being finalized. And they don’t like that—they just want something stable. So we are considering modifying slightly our level of communication and commitment.”

A section of the ITER vacuum vessel, shown here at the ITER site in France, after its shipment from South Korea. When all nine sections are assembled, they will form the toroidal chamber that will contain the fusion plasma. The French nuclear regulatory agency halted assembly of the first two sections in January, after finding that misalignments caused by damage during shipment would prevent them from being properly welded.

A section of the ITER vacuum vessel, shown here at the ITER site in France, after its shipment from South Korea. When all nine sections are assembled, they will form the toroidal chamber that will contain the fusion plasma. The French nuclear regulatory agency halted assembly of the first two sections in January, after finding that misalignments caused by damage during shipment would prevent them from being properly welded.

Close modal

In Spain, meanwhile, executives of Fusion for Energy (F4E), the EU entity that manages Europe’s share of in-kind contributions to the project, are pledging to hire additional staff, provide counseling services, and take other steps to address employee complaints of its poor management and excessive workloads.

Discontent at the agency came to a head last November after a three-member investigatory panel appointed by F4E’s governing board found no direct link between job pressures and the suicide of an F4E engineer in May 2021. In a note he left, Mario Gagliardi called his suicide his “final gesture,” saying in reference to his job, “I found myself to be the scapegoat of an unsustainable situation for years.”

Most of the 440-member F4E workforce staged a half-day strike in response to the panel findings. In their strike notice to F4E management, the three trade unions at the site complained of “the continuous degradation of the F4E working environment including management style, workload, lack of respect of the formal process and paying lip service to the social dialogue processes.”

F4E management and the unions both requested a more thorough investigation into the causes of the suicide by the European Anti-Fraud Office, or OLAF. Cristiano Sebastiani, a representative for one of the trade unions, says employees must be given full protections against retaliation for them to testify. OLAF has yet to agree to open an investigation.

F4E’s director, Johannes Schwemmer, told the European Parliament hearing that a “psychosocial-risk-assessment survey” found that excessive workload is a “risk factor” for the F4E workforce. An F4E spokesperson says that the assessment was not specifically linked to the suicide. Since joining F4E in 2016, Schwemmer stated, the amount of work under his management has doubled from €500 million to €1 billion annually. F4E is currently managing 98 separate ITER component projects, each having an average value of about €100 million.

F4E management has requested its governing board and the European Commission for approval to hire 34 additional employees. The request is pending, says an F4E spokesperson.

F4E has committed to other morale-boosting measures, including additional medical and psychological services, and regular meetings and discussions among management, staff, and trade unions. Yet Schwemmer defended F4E’s management record at the 28 February hearing, saying that annual personnel turnover is below 3% and job security is high.

Sebastiani, however, attributes low turnover to the dedication of F4E staff to the ITER project. “Building the reactor is more than just a job,” he says. “It’s what they have been living for. They are strongly motivated, and they only want to work there.”

As for management promises for workplace improvements, Sebastiani says, “It’s always the same. First you deny, and when you can’t deny anymore, you start saying you weren’t aware and then promise to change everything. What is clear is that the director must be told that the address for F4E is .eu, not .com. It’s not a commercial entity. It’s a European institution, and the rules are to be respected.”

Robert Winkel, the IO’s expert on beryllium, resigned his post in March, after supervisors ignored his recommendations, he says, and failed to consult him on occupational exposure to the toxic metal. The ITER vacuum chamber walls are to be lined with 12 tons of beryllium to protect them from damage by fusion neutrons and to absorb stray oxygen molecules. Kathryn Creek, also a beryllium expert and Winkel’s wife, left her IO job two years ago, blaming pressure from managers to water down her recommendations on beryllium protections.

Aerial view of the ITER site in May 2021. The cryoplant, magnet power conversion facility, and electrical switchyard are located to the left of the tokamak complex and assembly hall at center. The heat-rejection plant and central offices are to the right.

Aerial view of the ITER site in May 2021. The cryoplant, magnet power conversion facility, and electrical switchyard are located to the left of the tokamak complex and assembly hall at center. The heat-rejection plant and central offices are to the right.

Close modal

At 0.2 micrograms per cubic meter, the US and EU standard, beryllium’s occupational exposure limit is one-fifth that of hexavalent chromium, the next lowest limit for a toxic metal. But France observes its own limit of 2 micrograms, Creek says. Of a projected 450 ITER staffers who will be exposed to beryllium throughout the project’s lifetime, 100 could be expected to develop chronic beryllium disease if adequate protections aren’t implemented, Creek says. That statistic is based on experience from workers at US DOE facilities that handle beryllium, which is used in nuclear weapons.

Chronic beryllium disease is a debilitating condition that develops many years after exposure and goes undetected or can be misdiagnosed with x rays and CT scans. Diagnosis requires a specialized lymphocyte proliferation test. About 18 000 such tests are performed in the US each year, but France has no labs to perform them, Creek says.

“The fundamental problem is that beryllium isn’t recognized in France or Europe as a serious threat, like it’s known to be in the US,” says Winkel. “Consequently, there are no people with experience in how to safely manage it.” Beryllium, however, has long been used in Europe for space applications and in particle detectors.

Creek and Winkel’s specific concern involves exposures that will occur during maintenance in ITER’s operational phase, when ports to the vacuum chamber that hold diagnostic instruments will be opened and beryllium particles will flow into enclosed workspaces known as hot cells. They say that 4–20 air changes in the hot cells per hour are necessary to adequately filter the particles. But the IO’s current design will provide less than one change per hour, Winkel says.

Bécoulet says that as a nuclear facility, ITER will provide the same level of protection to beryllium exposure as it will to radiation. “The machine is made not to release anything,” he says. “If you remove anything from the machine, as it is a nuclear machine, it will be done in a completely sealed environment.”

Bécoulet says that the original choice to use beryllium in ITER was “nonsense” for multiple reasons, including its toxicity and cost. ITER’s design was based on the Joint European Torus in the UK, which also uses beryllium. Already, plans call for a different material, likely tungsten, to replace beryllium during ITER’s tritium phase.

But Bécoulet would like to see beryllium eliminated altogether at ITER, and he says that could yet happen. “It is an option; it is possible, but I don’t want to oversell it.” A change at this point could further delay construction, he says. “We should have considered this earlier.”

1.
D.
Kramer
, “
ITER disputes DOE’s cost estimate of fusion project
,”
Phys. Today
,
16
April
2018
.