Edge Localised Modes

ELM Cycle

Edge Localised Modes (ELMs) are repetitive bursts of the edge plasma. Because of their periodicity (albeit irregular), one way to imagine the ELM phenomenon is to picture a single ELM cycle. The most rapid changes occur during an ELM crash which is usually significantly shorter than the time between the ELMs. The plasma cross-section and the radial plasma pressure profile (i.e. plasma pressure as a function of distance from the plasma centre) are shown at four different time points during an ELM crash.

The first column in the figure corresponds to the situation before the ELM crash. The plasma is stable and has a steep pressure gradient at the edge. The gradient is maintained by the edge transport barrier that is always associated with the high confinement mode (H-mode) of tokamak operation.

The second column shows the onset of an ELM, which can be imagined as an onset of many small turbulent eddies at the edge due to the pressure gradient having exceeded a critical value for stability. The instability is not necessarily triggered by the pressure itself, but, for instance, by the so called "bootstrap current", an electric current driven by the pressure gradient. In the third column, the edge plasma is lost to the Scrape-Off Layer (SOL) where it flows along the magnetic field lines towards the divertor.

The lost plasma ends up on the divertor plates producing the distinctive peak in the D-alpha radiation (visible light emitted by excited atoms of deuterium fuel) as indicated in the fourth column.

ELM activity may evolve as e short, intense heat load on the plates causes erosion of the divertor materials. During the instability, the edge pressure gradient is reduced until the plasma becomes stable again. Then the pressure gradient starts recovering to the level where it reaches the stability limit so that another ELM occurs. If the conditions stay constant, the cycle can continue indefinitely. Depending on the ELM type and the details of a plasma device, each ELM removes 1 - 7 % of the plasma energy and particles.

Sound corresponding to density fluctuations due to ELM bursts can be downloaded here. The signal was obtained from a probe inserted into the plasma edge of the TCV tokamak. Note that the signal has been slowed down so that ELMs appear at a rate of five per second whilst in reality they are nearer two hundred per second. Courtesy of CRPP EPFL.

ELM Classification

Another way to examine ELMs is to study the global behaviour of the plasma during ELMs. While some of the features are common to all ELMs, there are also distinctive differences. Consequently, it has become standard to use the following classification of ELMs:

• Type I ELMs: The D-alpha radiation shows large isolated bursts and, therefore, Type I ELMs are also called 'large' or even 'giant' ELMs. The plasma edge is close to the theoretical ("ideal ballooning") stability limit or even beyond it. The instability is pressure driven, and as the heating power is increased, the ELM repetition frequency also increases. The degradation of the plasma confinement is smaller than with other ELM types.

• Type II ELMs: These are observed only in strongly-shaped plasmas, i.e. with high elongation and triangularity of plasma cross-section. Also the plasma density needs to be rather high. The magnitude of the ELM bursts is lower and the frequency is higher than that of type I ELMs, while the confinement stays almost as good. Sometimes, type II ELMs are called 'grassy' ELMs.

• Type III ELMs: The bursts are small and frequent. Therefore, another name for type III ELMs is 'small' ELMs. The instability is driven by electric current, and appears when plasma resistivity is rather high (i.e. edge temperature rather low). The ELMs repetition frequency is found to decrease with the increasing heating power. The plasma confinement is degraded more than with other ELMs.

In the presence of the edge transport barrier, i.e. in the tokamak H-mode operation, ELMs are instrumental for maintaining a stable density of confined plasma. In other words, without ELMs the plasma density in the H-mode increases above the overall stability limit, leading to sudden loss of the plasma confinement in a major instability called plasma disruption. However, two ELM-free operating modes with stable density have been observed in high confinement mode (H-mode) of tokamak operation:

• The Alcator C-MOD tokamak exhibits an "Enhanced D-Alpha" mode or EDA. In an EDA, while the plasma behaves as in the ELMy H-mode (steady-state density achieved, no accumulation of impurities), there are no periodic bursts of plasma, but the D-alpha-radiation remains at an increased level throughout the EDA period. The particle and energy confinement is reduced in comparison with a real ELM-free H-mode.

• In the DIII-D tokamak, with neutral beams injected in the direction opposite to the plasma current and with a large distance between the plasma and first wall, low density H-modes have been observed and named "quiescent H-mode". In this mode the ELMs become suppressed, and replaced by harmonic oscillations in the plasma edge. The oscillations are a sign of turbulent transport that keeps the particle transport high. Consequently, the plasma density does not increase as in a normal ELM-free H-mode which would normally lead to a disruption. From the fusion reactor operation point of view, the drawback of the quiescent H-mode is that the effective charge Zeff increases due to the accumulation of impurities into the core plasma.

• pace making of ELMs by injecting small pellets of frozen fusion fuel into the plasma edge at a high frequency, see e.g. JET's capabilities in support of ITER

• plasma edge ergodisation by resonant perturbations of the magnetic field. Studies at the DIII-D tokamak demonstrated an unexpectedly strong ELM suppression via resonant magnetic field perturbations. This is considered to be a very promising result for a reactor-relevant operation. However, both its understanding and its validation on other tokamaks is still at an early stage (in 2006).

ELM Model

Several models for ELMs have been suggested. Most models use plasma instabilities to explain ELM behaviour of plasmas. The plasma goes through a cycle where it is destabilized and then stabilized again.

The ELM cycle starts with a low pressure gradient as a result of the previous ELM crash that has removed the edge pressure "pedestal". Due to the edge transport barrier, the edge pressure pedestal develops quickly (1). The growth of the pedestal stops at the so called "ballooning stability" limit (2). Due to the pressure pedestal, the above mentioned bootstrap current - which is proportional to the pressure and temperature gradients - starts to grow. Eventually, the bootstrap current destabilizes an effect known as "ideal peeling" which leads to an ELM crash (3) and the loss of the edge pressure pedestal (4). The cycle then restarts from the beginning.

For general information on plasma modelling see Focus On: Computer Modelling of Fusion Plasmas.