An algorithm that reveals missing spent nuclear fuel pins awarded a prize

a cross-sectional image of a fuel assembly can be computed from measurement data

A cross-sectional image of a fuel assembly can be computed from measurement data.

An algorithm developed in a collaborative project of mathematicians and physicists from the University of Helsinki won second place in a challenge organised by the International Atomic Energy Agency (IAEA), in which solutions were sought for interpreting gamma radiation originating in spent nuclear fuel assemblies.

A problem faced by the organisation tasked with controlling the safety and security of nuclear energy is that even though gamma radiation emitted by spent uranium can be measured, it is not possible to determine whether all uranium pins in a specific fuel assembly are included as reported, based solely on the amount of radiation. For this purpose, the IAEA has developed a device known as a passive gamma emission tomograph (PGET), with which the internal structure of fuel assemblies can be examined. The technology is brand new, and with the help of the challenge the IAEA wanted to find out whether expertise exists to better interpret data collected with PGET equipment. In Finland, the findings are of interest to the Radiation and Nuclear Safety Authority (STUK) which monitors the implementation of the permanent disposal of nuclear waste.

With the mathematical algorithm, a cross-sectional image of a fuel assembly can be computed from measurement data, displaying the radiating uranium pins as bright spots and any missing pins as darker gaps.

“The method utilises inversion mathematics, as the result is used to calculate the original radiation source,” says Professor Samuli Siltanen from the University of Helsinki.

These so-called inverse problems are the focus of a research group active in the Department of Mathematics and part of the Academy of Finland Centre of Excellence of Inverse Modelling and Imaging.

The proposal by the University of Helsinki scientists approaches the problem from an original perspective compared to methods previously employed in this context. Due to the strongly absorbent nature of uranium, that is, its ability to absorb gamma radiation, the algorithm takes into account such absorption in addition to the radiation emitted, thus providing a more accurate image of the radiation source.

On the right track

Professor Peter Dendooven and Postdoctoral Researcher Camille Bélanger-Champagne from the Helsinki Institute of Physics (HIP) already started working on this imaging problem in 2015, with their collaboration expanding into a joint project of the Department of Mathematics and HIP in 2018. Tapio Helin, who moved to LUT University to take up an associate professorship during the project, emphasises the significance of cross-disciplinary cooperation.

“In this collaboration, the tools of modern mathematics meet the radiation expertise found in the field of physics. That’s a sure-fire recipe for novel and exciting research,” Helin says.

The challenge was held at the right moment, affirming the Finnish researchers in their choice of direction and in the competitiveness of their solution compared to other parties active in the field.

Another member of the group is student of mathematics Rasmus Backholm who is putting the finishing touches to his master’s thesis and working part-time at the Radiation and Nuclear Safety Authority. This month, he will travel to Vienna to accept the €3,000 prize.

“We haven’t yet decided what to do with the prize money,” Backholm reveals.

In addition to a ceremony held in honour of the winners, the research groups ranked highest in the challenge will introduce their algorithms in more detail at the event organised by the IAEA, providing an opportunity to learn from each other. After all, the work is not finished.

“Now that simulated data has yielded good results, we have been able to start working with authentic data. The results will also be reported in scientific publications of the field,” says Postdoctoral Researcher Tatiana Bubba.

/Public Release. View in full here.