Targeting mercury

SNS’s high energies call for a pioneering target source design

Mercury, the liquid metal, will be key to one of the boldest design elements of the Spallation Neutron Source, the state of the art neutron science facility to be built at Oak Ridge. The SNS will be the first accelerator-based neutron source that features mercury as a target material.

Mercury is one of the most fascinating elements. It is the only metal that in pure form is liquid at room temperature. It is a superior conductor of heat and electricity, making it useful for thermometers, switches and a host of other mechanical devices.

In the SNS, protons that have been accelerated to enormous energies will be directed into flowing mercury, knocking away neutrons from the atoms’ nuclei for use by researchers for a wealth of analytical tasks.

“It’s never been done before,” says Tony Gabriel, who is leading the SNS target development effort, which is a responsibility of ORNL in the five-lab SNS collaboration. Despite the fact that designers are turning to mercury as a target for the first time, Gabriel says the element is a natural for the job.

“About three years ago I heard a discussion of target designs for the proposed European Spallation Source,” Gabriel says. “The Europeans were pushing for a mercury target. I was a little skeptical, but we gave it a serious look. The more I looked, the more logical it seemed for a high-powered target.”

Mercury became attractive largely because the SNS is going to be such a big gun that will likely get even bigger. Conventional targets are solid, usually tantalum or tungsten. In the SNS, each pulse of protons will strike the target with 17 kilojoules of energy. In watts, that represents enough energy to illuminate 17 million homes for a microsecond. Those pulses will arrive 60 times a second. The SNS, in its startup configuration, will operate at an average of one megawatt of power (17K joules × 60 pulses per second = 1M joules per second or 1 MW).

At those energies, a solid target would be subject to destructive thermal shock and radiation damage. If the SNS is eventually upgraded to four megawatts, as planned, there is a risk that a solid target would quickly be reduced to burnt toast.

Gabriel and his colleagues on the design team—John Haines, deputy senior team leader; Tom McManamy, target lead engineer; and Jack Carpenter, a senior consultant from Argonne National Laboratory—point out that mercury, as a circulating liquid, won’t encounter the radiation damage that afflicts solids. It will also better conduct away heat from the process and be self-cooling in case of a shutdown. At the same time, mercury has a high atomic number, which means it will yield lots of neutrons.

Approximately 14 tons of mercury—enough to fill a cube about the size of a washing machine—will be pumped continuously through an enclosed, piped system. Protons will strike the target area, which is about the size of a computer keyboard, and a resulting torrent of neutrons will be channeled into beamlines for researchers. Fast, high-energy neutrons will be slowed by cooling—either with water or a liquid-hydrogen cold source, depending on energies desired—to make them more usable for experiments.

As for the mercury target, it barely takes a lick. “Over a long period of time, only a few percent of the mercury atoms are destroyed. The target system will operate for years with the same amount of mercury,” Gabriel says.

Gabriel, McManamy and Haines are investigating design issues that could arise. One concern, Gabriel says, is how the system will respond to the thermal shocks resulting from the pulses. That particular issue is being investigated in tests at Brookhaven National Laboratory’s AGS facility and Los Alamos National Laboratory’s LANSCE facility. How the system will respond to the combined effects of radiation and corrosion also will be investigated.

“The target must be shielded,” McManamy says, adding, “The radiation levels also require designing a remote handling system to service the system.” A prototype target system, complete with piping and pump, soon will be constructed at Building 7600, home of ORNL’s Robotics and Process Systems Division. Another Lab division, Metals and Ceramics, is investigating how mercury will react with the system’s materials. The most likely candidate material for the piping, Haines says, is stainless steel.

The system will be built with multiple barriers to contain both radiation and any mercury that might leak out. “We’re taking every possible precaution,” Haines says.

Interestingly, the shielding around the target area will be made of recycled steel, or steel that has come from other nuclear facilities and is slightly contaminated and unmarketable. It’s a much “greener” approach, Haines says. “We’re not contaminating new steel.”

Argonne’s Carpenter, who has advocated pulsed neutron sources for three decades, says the design of SNS will offer more neutrons with less waste heat to deal with. “Pulsed-mode accelerators instantaneously produce large fluxes of neutrons in a fraction of a second. In fission reactors, where 200 million electron-volts of heat are generated for each usable neutron, the SNS will generate only 40 million. It’s five times more efficient for the heat-removal buck,” Carpenter says.

The SNS is “the way toward higher power,” from an initial one megawatt to an eventual upgrade to four megawatts. In comparison, England’s ISIS accelerator, currently the “world’s brightest pulsed neutron source,” runs at 160 kilowatts. “For short-duration pulsed sources, one megawatt is a big jump,” Carpenter says.

One by-product of the spallation process, Carpenter adds, is gold, which resides right next to mercury on the periodic chart. Unfortunately, the gold occurs in such minute amounts that its only value will be amusement.

The best deal with the SNS may be the mercury itself. The SNS will obtain its mercury from the Y-12 Plant, which once used it for weapons production, and the supply is virtually unlimited. Best of all, says Haines, “It’s free. They said we could have all we want.”—B.C.