Monica Maceira. Image credit: Carlos Jones, ORNL
An ORNL team has tackled the challenge of detecting nuclear materials in vehicles by combining radiation detection with seismology, typically used to study earthquakes.
Putting these two detection methods together, they were able to not only detect radioactive materials in passing vehicles but also determine what direction those vehicles were traveling. This unique combination of data could provide authorities with a crucial tool for enhancing nuclear safety.
This project grew out of a previous experiment at ORNL’s High Flux Isotope Reactor. In that project, Monica Maceira, seismology portfolio manager in the National Security Sciences Directorate, noticed that vehicles generated enough energy at certain points around the facility to produce seismic signals — for instance, when encountering a large metal plate.
This observation planted an idea: What if seismology could be used to detect all vehicles driving around a facility, regardless of contact with an obstacle? So, Maceira and her team set out to prove that seismic sensing could be used for nonproliferation efforts that account for vehicle movement.
For 11 days, including a two-day targeted experiment, researchers deployed 24 three-component seismic sensors along roads around HFIR. Sensors detected vibrations from vehicles’ engines and the force of tires hitting the road.
“Those signals are very small,” Maceira said. “They are part of the background noise recorded by your sensor, and it’s nothing compared to an earthquake.”
Such small seismic signals could become lost among the heterogenous noise of a facility like HFIR. Despite this hurdle, the seismology team successfully monitored vehicle movement and verified results with GPS sensors and video surveillance.
“You still can see the vehicles throughout the HFIR facility where you don’t have a metal plate or a speed bump or anything like that in the road,” Maceira said.
Polarization indicated vehicle direction; about every 0.5 seconds, the three-component sensors mathematically rotated. The angular change between each rotation indicated the direction a detected vehicle was traveling, explained Omar Marcillo, an ORNL geophysicist who works alongside Maceira.
“That is a huge advantage,” Marcillo said. “The level of resolution is not great, but it gives you enough information to hint the direction of movement.”
On its own, the ability to detect and monitor a traveling vehicle is a valuable asset for high-security situations — so Maceira’s efforts were only strengthened by a collaboration with Dan Archer, section head for Nuclear Science and Advanced Technology in the Physics Division.
Traditionally, radiation detection systems travel while searching for radioactive material, but in Archer’s team’s system, the radiation detectors remain stationary to monitor moving vehicles.
“If your source is moving and your detector’s moving, there’s a low probability of overlap,” Archer said.
Six two-sensor radiation systems around HFIR detected moving vehicles, charted their trajectories and detected radioactive materials — including nuclides such as actinium-225, a rare cancer-fighting isotope that ORNL produces — among cargo.
As vehicles pass by, sensors record gamma radiation photon-by-photon, allowing researchers to identify both the radioactive material and the speed and distance of the vehicle. The captured photon energy histogram is then used to denote the specific radioactive nuclide.
“The different nuclides emit photons with different energies, and so they deposit in different portions of your spectrum,” Archer said. “You have more counts in some parts of the spectrum than in other parts of the spectrum, and based on that you can fingerprint the nuclide.”
The data volume accumulated during this experiment was so large — researchers dedicated a full petabyte of storage to the project — that the team plans to analyze the data for additional, unnoticed patterns.
“It’s a data-rich set,” Maceira said. “It would be great for us to keep looking [for more information].”
Researchers are planning additional experiments to further characterize the detection method’s potential. — Alexandra DeMarco
