Focus on the future during Fusion Energy Week

"Ever since I first heard about fusion, I’ve been fascinated by the idea that we humans might one day master the same type of energy process that powers the stars. It’s an ambition with potentially huge benefits for society, and I wanted to be part of it. And I happen to enjoy physics and programming – which are both central to this field," says Andreas Gillgren, PhD student in plasma physics at the Department of Space, Earth and Environment.

In a star like the sun, vast amounts of energy are released when light hydrogen atoms fuse to form new, heavier atoms. That may sound simple, but recreating fusion on Earth requires heating the atoms to temperatures above 100 million degrees. At that point, the atoms transitions from gas form to plasma – a state of matter that must be controlled long enough to extract more energy than is used to heat it up. The energy potential of this technology is enormous, since neither the fuel nor the resulting “ash” is in any way hazardous. The promise of clean, abundant energy is what drives global efforts – and it is also a big part of what drew PhD students Andreas Gillgren and Ida Ekmark into the field of fusion energy.

"It’s a combination of exciting physics and its potential benefit for the future. The physics in a fusion plasma involves large quantities of charged particles, which interact through electromagnetic fields and display collective behaviour, and this leads to very interesting dynamics. Moreover, if we manage to realise fusion as an energy source, it will have a highly positive impact on our society and the environment, and it is inspiring to be part of the development that aims to make that happen, says Ida Ekmark, PhD Student in Plasma Theory, at the Department of Physics at Chalmers.

A century later

On 10 May 1900, British astronomer Cecilia Payne-Gaposchkin was born. In 1925, she became the first to show that the sun is composed of hydrogen. Her discovery laid the groundwork for our understanding of fusion – the process that keeps the stars shining. Fusion Energy Week is held during the second week of May each year to mark her birthday.

But it’s not just the centenary of Payne-Gaposchkin’s discovery that makes this year’s Fusion Energy Week extra special. For the first time this spring, all European fusion experiments have run intensive campaigns simultaneously – a coordinated effort demonstrating the strength of the European collaboration Eurofusion, of which Chalmers researchers are a part.
Read more in the press release: European labs lead the way: Europe’s fusion energy research in full swing, from Eurofusion.

In recent years, several new milestones have also been reached. The recently retired UK reactor JET set a record for the amount of energy produced in a fusion experiment, while the French reactor WEST set a record for the longest sustained plasma.

As for the future of fusion technology, both researchers are optimistic. When asked whether she expects to see working fusion power in her lifetime, Ida Ekmark replies:

“Fusion is famously always thirty years away – we’ve been saying that since the dawn of fusion research, so yes, of course! Joking aside, I think we will have achieved break-even – meaning we get more energy out than we put in – within a couple of years, and I am hopeful that fusion power will be an established energy source contributing a significant share of our energy supply within my lifetime,” says Ida Ekmark.

When Andreas Gillgren gets the same question, his answer is short and to the point:

“Yes!”

More info: 

Fusion Energy Week 2025 is celebrated every year May 5-9, initiated by US Fusion Energy.

Read more about Ida's and Andreas' research below: 

Ida Ekmark: Preventing runaway electrons in fusion plasmas

Ida Ekmark is a doctoral student at the Department of Physics and a member of the Plasma Theory research group, led by Professor Tünde Fülöp. Ida Ekmark’s research focuses on how to prevent damage to fusion reactors caused by so-called runaway electrons during the operation. Two years ago, she was also awarded the Bert-Inge Hogsved Prize for best entrepreneurship for developing a modelling tool, together with Esmée Berger (who is also currently a doctoral student at the Department of Physics). The tool enables researchers to study how fusion reactors can be safely started.

Hi Ida! Can you tell us more about your research?

– I study a subgroup of electrons in the plasma, more specifically the fastest ones. In a fusion plasma, electrons are accelerated by an electric field. At the same time, they are slowed down by a friction force caused by collisions with other particles in the plasma. However, this friction force decreases with increasing velocity, which means that if an electron reaches a sufficiently high velocity, the friction will no longer balance the acceleration from the electric field, and the electron will accelerate without restraint – it will “run away” to relativistic speeds.

– These runaway electrons can accumulate into a highly energetic beam, and if this beam hits the reactor wall, it can cause severe damage to the fusion reactor. We want to avoid that, and that is what my research is about; how to prevent or mitigate runaway electrons in fusion plasmas.

You’re about to present your licentiate thesis. What is it about?

– My licentiate thesis focuses on how we can model runaway electrons during the start-up and disruptions of plasma shots in tokamaks, which are a type of fusion reactor, and how disruption mitigation can be optimised to minimise the risk of damage.

What do you hope to achieve in the fusion field in the future?

– I hope to be part of the positive progress that takes us towards functioning fusion, and to contribute at least one piece to the puzzle of how to handle runaway electrons in fusion reactors. In the short term, I hope to provide insight into whether runaway electrons might appear in stellarators, a type of fusion reactor concept where relatively little research has been done on runaway electrons. They are mainly expected to be a problem in tokamaks, another reactor concept, but could also be generated and cause damage in stellarators. In the long term, I hope to stay in academia and continue doing research regarding fusion plasmas, but exactly how is difficult to say at this point.

Andreas Gillgren: AI as a problem solver in fusion research

Andreas Gillgren is a PhD student at the Department of Space, Earth and Environment. He is part of the Plasma Physics and Fusion Energy research group, led by Professor Pär Strand. The group provides theoretical support and helps test and validate models and tools for the major fusion experiments taking place at various European research facilities. For this purpose, Andreas and his colleagues have developed their own tool, NeuralBranch – a neural network framework designed to be more interpretable and easier to understand.

Andreas, what is your research about?

I'm exploring how machine learning and AI can be used to solve various challenges in the field of fusion. I have a particular interest in interpretable machine learning, which means we’re trying to overcome the so-called "black-box" problem. In other words, it’s about using methods that help us understand how machine learning models justify their predictions or decisions. The goal is to build trust in these models within the field – and to actually learn something from them.

What do you hope your research will achieve for the future of fusion?

I hope my research will help increase the use of interpretable machine learning models in fusion science. We’re already seeing growing interest: researchers who were previously skeptical of machine learning are now curious and want to compare their expertise with the patterns revealed by our interpretable models – which is exciting. In the long term, I’d love to see a future where all machine learning models are interpretable, both in fusion and beyond. To me, that’s a better future than one dominated by black boxes we don’t fully understand.

Do you think your work could be useful beyond fusion research – and if so, how?

Absolutely. The methods we’re developing in interpretable machine learning aren’t limited to fusion. I believe transparency and interpretability will be crucial for AI applications in high-risk fields like healthcare and law. But really, any field can benefit from gaining a deeper understanding of its data – and that’s exactly what interpretability helps with.