Science in the spotlight

Taming the ELMs

MAST ELM teamIt's well known that fusion research copies the power of the Sun – but the connections don't end there. Eruptions at the edge of the tokamak's plasma are remarkably similar to solar flares, and can be just as harmful to anything that gets in their way. A team of CCFE physicists is studying these phenomena, known as ELMs, and how to stop them from damaging fusion machines.

Keeping plasma under control inside the tokamak's magnetic field is one of the major challenges in fusion science. To extract energy to run powerplants, it will be necessary to confine the extremely hot plasma long enough for fusion reactions to occur and for the process to become self-sustaining. However, plasma does not respond well to being kept in a magnetic cage, and reacts by trying to escape wherever it can. The instabilities and energy losses this causes are a headache for researchers – particularly as plasma, despite forming over 99% of the known universe, remains a mysterious form of matter.

“Controlling a plasma in a fusion device is not an easy thing to do. It's a bit like trying to keep a jelly still in your hand,” says Dr Andrew Kirk, who leads the ELM research team. “Advances in plasma physics mean that we are beginning to understand a lot more about how it behaves, but there is still much that we don't know – it's a fascinating branch of science.”

ELM in MASTThe stability problem is most severe at the edge of the plasma, where pressure builds up as the ionised gas attempts to break its shackles. This results in particularly forceful instabilities, known as Edge Localised Modes (ELMs for short). ELMs are a focus of intense research in the MAST tokamak at CCFE, and at similar experiments around the world. Not only do they take energy out of the plasma and affect our ability to confine it efficiently; the ELMs unleash hot material which can damage machine components. This will become even more concerning when larger, higher-power machines such as ITER are built – the bigger the ELM, the more harm it can do.

Physicists developing theories of the ELMs' behaviour were struck by the comparison with solar flares – the spectacular bursts of material thrown out by the Sun into space where its magnetic field is strongest and most turbulent. Typically tens of times the size of the Earth, the flares can contain as much energy as a billion hydrogen bombs, and can be responsible for interruptions to our communications signals and even our power supplies.

Howard Wilson (then of CCFE) and Steve Cowley (then of Imperial College, London) developed a theory for ELMs that predicted that filaments of hot plasma would erupt from the surface of the tokamak in much the same way as solar prominences. These filaments were first observed at MAST in 2004.

Solar flareA look at the images confirms the similarity between the two. Whereas a solar flare (shown right) lasts for an hour or so, there is only a window of about 50 microseconds to capture the effect on camera in MAST. What can be seen in this short time is that the hot ELM filament projects beyond the surface of the plasma in the same way as a solar arc is ejected beyond the Sun's surface.

Solar flares are linked to the Sun's 11-year magnetic activity cycle, which will reach its latest peak in 2013. But while the strength of the solar flares reaches a crescendo, fusion scientists are working on ways to produce the opposite effect on ELMs in tokamaks.

“The aim for ITER will be to make small ELMs or to stop ELMs from occurring at all,” explains Dr Kirk. “We call this ‘ELM mitigation'.”

At CCFE, researchers are using an ELM mitigation technique known as resonant magnetic perturbation. This involves applying targeted magnetic fields around the tokamak to punch holes in the edge of the plasma, allowing pressure to be released in a measured way. This results in weaker ELMs which do not pose a threat to the plasma-facing materials in the wall of the machine.

ELM lobe structureThis method has proved to be successful in curbing ELMs, and new images observed in plasmas during MAST's most recent campaign are giving the CCFE team valuable clues on how to perfect it. In December 2011, MAST's high-speed cameras recorded the formation of finger-like lobe structures near the base of the plasma (see photo opposite). Although predicted by theoretical physicists since 2004, this was the first time they had actually been seen in a tokamak.

The lobe structures are caused by the resonant magnetic perturbation, which shakes the plasma and throws particles off course as they move around the magnetic field lines, changing their route and destination. Some particles end up outside the field lines, forming offshoots near the base of the plasma.

Changing the shape of a small area of the plasma in this way lowers the pressure threshold at which ELMs are triggered. This should therefore allow researchers to produce a stream of small, low-energy ELMs.

Looking ahead, the next phase of the research will involve developing computer codes and simulations to map how particles will be deposited and how the lobes will be formed around the plasma. Dr Kirk is optimistic about the results of recent experimental findings and their long-term contribution to controlling the plasma instabilities:

“Measurements from the recent MAST experiments will allow us to refine our plasma models to take the lobe effect into account. We'll also be able to create advanced codes to give more accurate predictions for the performance of future machines like ITER.

“The lobes we've identified at MAST could be an important discovery for tackling one of the biggest concerns for physicists at ITER. We're confident of taming the ELM.”

Lifting the lid on ELMs

  • PanThe processes that cause ELMs are similar to those that occur every day in the kitchen. A saucepan of water is put on to boil. The heat is turned up until the water starts boiling and steam escapes from the top of the saucepan. Heat is being lost; this can be compared to a plasma with a low confinement (known as ‘L-mode').


  • The heating efficiency can be improved by putting a lid on the saucepan. The water comes to the boil much more quickly because the loss processes have been reduced and the pressure in the saucepan increases. This is what physicists call ‘H-mode' – a high-confinement plasma.


  • As the pressure inside increases, there is enough energy to lift the lid of the saucepan and release some of the pressure – this is similar to an ELM. As the pressure is released, the lid falls back and allows the pressure to build up again, resulting in a cycle of pressure building and releasing; this is an ELM cycle.


  • In the kitchen you have two choices if this clattering of the lid gets on your nerves. You can either go for a heavier lid – but in the end the pressure inside would overcome this extra weight and boil over, throwing out a large amount of the contents of the saucepan. In fusion, this is what we call a ‘disruption', when control of the plasma is lost. The other alternative is to either stick a spoon under the lid, or open a valve in the lid to regulate the pressure so that it never gets too high; this is what researchers are attempting with ELM mitigation. The aim is to keep the pressure in the plasma as high as possible to improve the efficiency of the ‘cooking' whilst not letting it get high enough to trigger an ELM.