Statistical Characterisation and Classification of Edge Localised Plasma Instabilities

TL;DR

This paper demonstrates that the waiting times between edge-localised plasma instabilities (ELMs) in fusion devices follow a Weibull distribution, confirmed by data from JET, enabling better classification and understanding of ELM dynamics.

Contribution

It establishes a first-principles-based statistical model for ELM waiting times and provides a quantitative classification method, advancing understanding of ELM processes.

Findings

ELM waiting times follow a Weibull distribution.

Empirical validation with JET data confirms the model.

Enables classification and analysis of ELM types.

Abstract

The statistics of edge-localised plasma instabilities (ELMs) in toroidal magnetically confined fusion plasmas are considered. From first principles, standard experimentally motivated assumptions are shown to determine a specific probability distribution for the waiting times between ELMs: the Weibull distribution. This is confirmed empirically by a statistically rigorous comparison with a large data set from the Joint European Torus (JET). The successful characterisation of ELM waiting times enables future work to progress in various ways. Here we present a quantitative classification of ELM types, complementary to phenomenological approaches. It also informs us about the nature of ELMing processes, such as whether they are random or deterministic.