Oak Ridge officials are puzzled by failure of targets at Spallation Neutron Source
Oak Ridge National Laboratory’s Spallation Neutron Source, the premier source of pulsed experimental neutrons in the world, has been operating at reduced power due to the premature back-to-back failures of two of the target vessels that are the source of the facility’s experimental neutrons. Now down to a single spare target, the SNS will continue to operate at a reduced power level of 850 kW until April, soon after a new target is delivered, says Kevin Jones, SNS operations manager. At full power, the SNS operates at more than 1 MW.
But Jones doesn’t believe there is a significant risk of a shutdown. “We think we have a strategy for managing this problem,” he says. “We’ve been in that position before; in late 2012 and early 2013, we had a similar problem of two unexpected failures in a row.”
The sequential failure of the two $1 million targets in the fall, and an unrelated technical problem, resulted in the loss of 1000 hours of operating time, Jones says. But about half of that has been recouped by rescheduling the experiments for the remainder of the fiscal year. On balance, the machine’s available operating time for the fiscal year has been reduced by 10%, he says. Growing demand—a record 600 proposals were received from users for an upcoming four-month-long experimental period—means that more researchers will be left out in the cold.
Although the SNS had a “difficult fall,” Jones says, “I don’t see a risk to us shutting down.”
Plans call for the SNS to resume full-power operation soon after the new target is delivered, Jones says. “We’ll evaluate the performance of the current target, look at its exposure history, and make a determination of whether we will increase the power in the latter part of April.” The impact of the reduced power operation varies depending on the type of experiments performed on SNS’s 17 neutron-scattering instruments. Some have seen “a slight degradation in throughput,” but others have experienced only minimal changes in performance.
Still, it isn’t entirely clear what caused the recent failures. The stainless steel target vessels contain the mercury, from which neutrons are knocked off, or spalled, by the accelerator’s impinging proton beam. Although the calculated lifetime for each should be about 5000 megawatt hours, none of the 11 targets that the SNS has gone through since its 2006 debut have lasted that long. The two that failed in the fall went out after just 100 and 600 MWH. Both failed at welds, one at the same weld that caused both of the 2012–13 failures. But the quality of the welds has been significantly improved since 2013, Jones notes.
With pulses of 20 kilojoules of energy deposited on them 60 times per second, the targets experience lots of thermal and mechanical fatigue, he says. Those cyclical stresses can cause a tiny defect in a weld to develop into a crack. The laboratory is working with the manufacturer to determine what can be done to prevent future failures. But Jones doesn’t think that the electron beam welding process, which uses no filler material at the joint, can be improved on. “It’s a very difficult manufacturing problem,” he says.
There are two other lifetime-limiting factors, Jones says. “One is the sheer number of displacements per atom induced by the primary and secondary radiation fields of the beam impingement and then the neutrons, pions, muons, and rescattered protons that are coming out of the mercury in the target.”
The other source of damage is internal cavitation from the shock of the energy deposited in the mercury, which results in a vacuum bubble at the interface of the mercury and the steel vessel. When that bubble collapses, he says, it can form a jet of mercury that chews the vessel away from the inside out.
In the SNS targets, mercury also acts as the working fluid in removing the heat generated by the impinging beam. Lower-power devices such as the Los Alamos Neutron Scattering Center don’t develop as much heat and can use targets made of such other neutron-rich elements as tungsten and tantalum.