Prediction and analysis of JET fusion performance based on reduced first principle transport models

The design of a Tokamak device is carried out initially with a 0D approach aiming at defining the plasma engineering parameters estimated with the help of empirical scaling laws, and the technological limits of the device components. The assessment of local parameters (1D) is then required to define the optimal plasma performance during the entire time evolution of the discharge. In this contest, the transport of energy and particles in fusion plasmas is one of the main actor in determining the evolution of a plasma scenario both in present experiments and in future reactors. The Joint European Torus (JET) experiment has operated in deuterium (D) and tritium (T) main ion plasma composition in 1997 (DTE1) and in 2021 (DTE2). The most important differences between the two experimental campaigns are related to the plasma facing components, carbon (C) in DTE1 and Be/W in DTE2, the increased additional heating power, and the presence of improved diagnostics, especially at the plasma edge which is determinant in the global plasma performance. After DTE1 the high levels of T retention in the C-wall have been considered unacceptable for a reactor, leading to the substitution of the C-wall with a metallic wall in the design of the International Tokamak Experimental Reactor (ITER). DTE2 campaign at JET aimed at studying D-T plasmas in the closest conditions to ITER operations. Differently from DTE1, the recent campaign focused on the stationarity of the performance and on addressing ITER-relevant aspects such as α-particles physics, plasma wall interactions and plasma heating schemes. In preparation to D-T operations, a wide experimental and modelling activity has been performed at JET in order to optimise the plasma scenarios. The focus of this thesis is the extrapolation in D-T main ion plasma composition of the JET baseline scenario. The latter is a high confinement mode (H-mode) plasma, characterized by the presence of Edge Localized Modes (ELMs), where the confinement relies on high plasma current. In ITER D-T operations, the baseline scenario is envisaged to achieve a gain factor, defined as the ratio between the fusion power and the input power, Q = Pfus/Pin ≈ 10. The objective of the thesis has been achieved through extensive integrated modelling, based on the reduced first principles transport models QuaLiKiz and TGLF employing different assumptions, and in a wide range of plasma operating conditions. QuaLiKiz and TGLF transport models have been validated in reference D plasmas, and their extrapolation capability with different plasma parameters has been tested by performing blind predictions. The results of the predictive modelling have been compared with the experimental data and analysed in order to address the sensitivity of the plasma scenario to the experimental boundary conditions. The QuaLiKiz transport model has also been validated against the experimental results produced at JET in DTE1. Before the start of the DTE2 campaign, an estimate of the particle sources required to sustain a 50-50 D-T baseline plasma has been obtained. This result has provided inputs to the JET control team in the preparation phase of the baseline fuelling scheme. This contribution boosted JET D-T operations without spending experimental time, neutron and T budget. The results of the predictive modelling performed in preparation to DTE2 are presented and discussed. The sensitivity of the predictions to plasma parameters v vi such as current, toroidal magnetic field, pedestal confinement and impurity content are analysed together with the sensitivity to the available amount of auxiliary heating power. The experimental results obtained in DTE2 by the baseline scenario are also presented and discussed. In the last part of this thesis, the implications of the modelling assumptions performed on D pulses will be compared with the assumptions done on D-T discharges with the experimental boundary conditions. The key parameters needed for reliable predictions of future experiments are discussed both in D and D-T main ion plasma composition. The estimate of particle sources obtained before the DTE2 campaign are adjusted to reproduce the experimental conditions, leading to an estimate of the different fuelling channels and an evaluation of the wall sources. The thesis is organised as follows: • In Chapter 1 we introduce fusion as a potential energy source. • In Chapter 2 we describe the Tokamak configuration and the JET experi- mental device, and we present a first comparison between the different D-T experimental campaign, and between the different scenarios prepared for DTE2. • In Chapter 3 we introduce the issue of energy and particle transport in Tokamaks, we present the theoretical background and the state of the art of transport analysis. The models used in this work are presented together with the different assumptions implemented in JINTRAC. • In Chapter 4 we present and discuss the validation of the reduced first principle transport models performed on D pulses. The extrapolations in D-T plasma mixture are presented with their sensitivity to the operating conditions. • In Chapter 5 we present and discuss the baseline results obtained in DTE2. The predictive simulations are improved by adopting the actual boundary conditions of the D-T experiments, and we discuss the impact of the different assumptions on the modelling of D plasmas extrapolated to D-T plasma mixture. We show the limits of predictive simulations in integrated modelling, and we use the predictive simulations to obtain an estimate of the different fuelling sources in the D-T experiments.